Ironworking, Metallurgy and Material Science

This file reports what instruments have measured. That is a deliberate restriction, because claims about pre-colonial African technology attract both inflation and dismissal, and the laboratory record is the ground on which neither survives. The archaeometallurgy and the glass chemistry are unusually good evidence: they involve counted samples, named analytical techniques and published compositions, and the strongest single result in the whole of Yoruba material science is a case where analysis overturned a century of confident scholarly assumption in the tradition's favour.

The craft and social dimensions of these trades, the smith and Ògún, the guilds, the dye pits and the adìrẹ industry, are 07-arts/07-metal-pottery-calabash-leather and 07-arts/05-textiles. This file owns the chemistry and the analytical evidence.

Confidence here is marked high because the central claims rest on published instrumental analysis rather than on report.

Iron: what the slag shows

The relevant evidence is from Modakẹ́kẹ́ at Ifẹ̀, analysed by Akin Ige and Thilo Rehren and published in 2003 1. Fourteen fayalitic bloomery slags, together with possible ore and furnace wall samples, were examined by chemical and mineralogical techniques 1.

The two findings that matter.

A skilful operation. The authors conclude that the slags derive from a skilful bloomery smelting operation extracting the maximum possible amount of iron from the ore 1. That is a technical judgement made from slag chemistry: the composition tells you how much iron was left behind, and little was.

A deliberately mixed ore charge. Most of the slags show exceptionally high titanium oxide, up to around ten weight percent, indicating the use of a mixed ore comprising limonite iron stone together with ilmenite-rich black sand 1.

The second finding is the more interesting one and it deserves unpacking. Mixing two ore sources of different character is a fluxing decision. A bloomery furnace has to produce a slag that is fluid at the working temperature so that it separates from the metallic bloom and can be tapped or drained; get the chemistry wrong and the slag is too viscous, traps iron, and the smelt fails. Adding a titanium- and iron-rich beach sand to a limonite ore changes the slag chemistry in a way that a smelter who did not understand what he was doing would not have arrived at and could not have sustained. This is empirical process control, arrived at without any theory of oxides, and it is one of the more solid pieces of evidence for indigenous Yoruba metallurgical competence.

What bloomery smelting is. Iron ore is reduced with charcoal at temperatures below the melting point of iron, producing a spongy solid bloom of metal mixed with slag rather than liquid cast iron. The bloom is then repeatedly heated and hammered to expel slag and consolidate the metal. It is labour-intensive, produces wrought iron rather than cast, and was the universal method everywhere before the blast furnace.

Dating and scale, stated carefully. The Modakẹ́kẹ́ material establishes competence, not chronology; the publication does not supply a secure date for the smelting operation. Estimates of the scale of Ifẹ̀'s iron industry in the historical period, such as the figure of some 500 smelters, smiths and other iron workers living within a radius of the city, come from documentary rather than archaeological sources and should be treated as approximate. Iron production in the region declined after the destruction of Old Ọ̀yọ́ around 1830 and further with the importation of European iron.

What must not be transferred. The very early West African iron dates that circulate, around 2000 BCE at Lejja and 750 BCE at Opi, are from Igbo sites in the Nsukka area, not Yoruba ones, and the earliest is contested within African archaeometallurgy. They are not evidence about Yorubaland and applying them there is a common and serious error. 07-arts/07 makes the same point.

Glass: the Igbo Olokun result

This is the strongest single result in Yoruba material science and it has a satisfying shape, because it is a case where the received scholarly view was wrong and the analysis corrected it.

The problem. Igbo Olókun, on the northern periphery of Ilé-Ifẹ̀, has been recognised as a glass-working site for over a century; Leo Frobenius described glass-encrusted crucibles from his pit shafts there in 1913 2. For most of the twentieth century the material was read as evidence of secondary processing: imported glass, remelted locally and worked into beads 2. Davison's neutron activation and X-ray fluorescence analyses in 1972 found a dominant group with low soda and significant potassium, which suggested wood-ash flux, and since wood-ash glass dominated medieval glassmaking north of the Alps she concluded that the beads most likely originated in medieval Europe 2. For three decades after that, all Ifẹ̀ glass was presumed imported 2.

The correction. Only after comparative databases on glass composition had been built did the significance of the very high lime and alumina levels, averaging around 15 and 13 weight percent respectively, become apparent as a defining and unique signature 2. Lankton, Ige and Rehren proposed in 2006 that this high lime, high alumina composition, HLHA, represented a glassmaking tradition unique to West Africa and possibly to southern Nigeria 2.

The excavation and the analysis. Babalola excavated at Igbo Olókun in 2011 and 2012, in four one-by-three metre units within or adjacent to the fenced reserve, recovering almost 13,000 glass beads screened through 1.2 mm mesh, over 800 crucible fragments, almost three kilograms of glass waste and cullet, and approximately 14,000 potsherds 2. Fifty-two beads from the excavated assemblage were analysed by laser ablation inductively coupled plasma mass spectrometry, LA-ICP-MS, and by scanning electron microscopy with energy-dispersive X-ray spectroscopy, SEM-EDS 2.

The results. The analyses confirmed the prevalence of HLHA glass and provided firm evidence of a second compositional group characterised by low lime and high alumina, LLHA. No imported soda-lime glass beads were present among the analysed samples at all 2. Crucible evidence indicates that LLHA glass was worked together with HLHA glass at the site and may have been made locally as part of the same technological tradition 2.

The conclusion, in the authors' own terms. They state that they are now confident beyond reasonable doubt that both the HLHA and LLHA groups represent glass produced in early Ilé-Ifẹ̀ using local recipes, raw materials and technology, and that these glasses are distinct from the soda-lime glass of the Islamic world, from the wood-ash or potash glass of medieval Europe and from Indian high-alumina glass 2.

The raw materials. Most likely granitic sand, with or without added calcium carbonate, was used to produce both types, with colorants rich in MnO, Fe2O3, CuO and CoO intentionally added 2. Freestone had suggested that the high alumina could come from feldspar and the lime from limestone or shell 2. Work by Ogundiran and Ige at Ọ̀ṣogbo, a seventeenth to eighteenth century site about 40 km north of Ilé-Ifẹ̀, recovered HLHA beads and production waste and argues for a continuing manufacturing tradition using snail shells and local sand derived from pegmatite 2.

The dating. Radiocarbon determinations from the excavation include 840 ± 30 BP (1058-1264 cal AD) at 0.60 m depth, 570 ± 30 BP (1304-1423 cal AD) on basal deposits in unit IO-C, and 610 ± 30 BP (1295-1404 cal AD) from the lowest levels of unit OO-A 2. These combine to support a chronology of the eleventh to fifteenth centuries for the glass beads and the workshop, consistent with the eleventh to fifteenth century dates from Garlake's excavations at Woye Asiri and Obalara's Land and with thirteenth to fifteenth century thermoluminescence dates on the fired clay cores of cast brass sculptures from Ita Yemoo 2. Most scholars identify the eleventh to fifteenth centuries as the period of Ilé-Ifẹ̀'s florescence 2. The authors state that available compositional and archaeological data strongly indicate that HLHA glass beads were being made and used at Ilé-Ifẹ̀ at least by the eleventh century AD 2.

The distribution. HLHA beads have been identified at Igbo-Ukwu in a ninth to twelfth century context, Essouk in a twelfth to fourteenth century context, Gao Ancien in a tenth to twelfth century context, and at Kissi, Bura, Diouboye and Kumbi Saleh 2. In every case they appear alongside beads from at least one other production area 2. The authors describe the scale of production at Igbo Olókun as unique in West Africa 2.

Why this matters and what it does not show. It establishes primary glass production, meaning glass made from raw materials rather than imported glass remelted, in West Africa from at least the eleventh century, using a recipe found nowhere else in the Old World, at a scale that supplied a distribution network reaching the Middle Niger. That is a first-order finding about indigenous technology and it required nothing but careful analysis to establish.

It does not show that Ifẹ̀ invented glassmaking, that Ifẹ̀ glass is older than glass elsewhere, or that the technology was independent of any external stimulus, and the paper claims none of these. The soda-lime beads found at Orun Oba Ado and Ita Yemoo show Ifẹ̀ was connected to long-distance networks carrying glass from the Middle East, Egypt and Syria-Palestine 2. A local tradition with a unique recipe developed in a connected place is what the evidence shows, and it is more interesting than isolation would be.

Brass

Brass casting at Ifẹ̀ is the best-known Yoruba technology and the sculpture belongs to 07-arts/02-ife-sculpture. Three technical points belong here.

The method is lost-wax casting, cire perdue, in which a wax model is invested in clay, the wax melted out and metal poured into the void, destroying the mould at extraction. The thermoluminescence dates on the fired clay cores of the Ita Yemoo castings fall in the thirteenth to fifteenth centuries 2.

The alloys are variable and the terminology is loose. Copper alloyed with zinc gives brass; with tin gives bronze; West African castings use varying proportions of copper, zinc, tin and lead, so "bronze" as commonly applied to the Ifẹ̀ and Benin material is frequently inaccurate.

The copper is not local. There is no significant copper source in Yorubaland and the metal arrived through trade, with the trans-Saharan routes the usual attribution. The technical achievement is therefore in the casting rather than in the metal production, which is the opposite of the iron case and is worth keeping straight.

The ẹdan Ògbóni, brass cast over an iron core, is treated in 07-arts/07; the material argument it embodies, non-rusting brass fused to cutting iron, is a case where the metallurgy is the meaning.

Indigo: the chemistry of a reduction vat

Yoruba indigo dyeing is a redox process and the practitioners controlled it without any theory of oxidation. The craft, the adìrẹ industry and its collapse are 07-arts/05-textiles; the chemistry is here.

The plant. Yoruba indigo is èlú, Philenoptera cyanescens, formerly Lonchocarpus cyanescens, a West African vine of the Fabaceae, and not Indigofera tinctoria, the Asian indigo of the world trade 3. The substitution is routinely made in popular sources and is wrong.

The problem the process solves. Indigotin, the blue pigment, is insoluble in water. A pigment that will not dissolve cannot penetrate a fibre, so it cannot dye anything as it stands. The dyer therefore has to convert it to a soluble form, get it into the fibre, and then convert it back.

How the traditional process does it. Leaves and young shoots are bruised to a pulp, formed into balls of about ten to twelve centimetres, dried and traded in the markets 3. The balls are steeped in water with an alkali, and the alkali is wood-ash lye, obtained by leaching burnt hardwood ash with water to give a potassium carbonate solution 3. The vat is left to ferment, in accounts running from about six to eight days up to three weeks 4.

The chemistry. Fermentation in the vat is anaerobic bacterial action, and it is a reducing environment. In alkaline reducing conditions insoluble indigotin is converted to leuco-indigo, the soluble reduced form, which is yellow-green. The fibre absorbs the leuco form from solution. When the cloth is lifted into the air, atmospheric oxygen reoxidises the leuco-indigo back to insoluble indigotin, now trapped inside the fibre. This is why cloth comes out of a Yoruba dye pot yellow-green and turns blue as it hangs, and it is why depth of colour is built by repeated dip-and-oxidise cycles rather than by a single long immersion.

The control problem. Both the alkalinity and the reduction have to be right, and the dyer has no instruments. What the sources record is that an experienced dyer judges readiness from the colour of the foam on the surface of the vat 4. That is a trained observational proxy for a redox state, and it is a real analytical skill.

Why the industry died. Caustic soda replaced ash lye because it works in hours rather than the days or weeks a fermentation vat takes, and it is cheaper; it is also a far harsher alkali that degrades cotton fibre. Combined with synthetic indigo it produced cloth that looked acceptable on the stall and failed in use, and the resulting quality-signalling collapse destroyed the Abẹ́òkúta industry between 1937 and 1939. This is treated fully in 07-arts/05-textiles and rests on Byfield's The Bluest Hands.

Chemical characterisation. Analysis of traditional Yoruba indigo-dyed material reports indigotin together with 2-hydroxynaphthoquinone and mineral ions including Al³⁺, S²⁻, Na⁺ and Fe³⁺ 4.

Soap: saponification with a potash lye

Ọṣẹ dúdú, black soap, from ọṣẹ, soap, and dúdú, black, originates in Yoruba communities of southwestern Nigeria 5. Its manufacture is a controlled chemical synthesis and it is worth describing precisely, because it is the clearest example in Yoruba practice of a two-stage process with an intermediate the makers cannot see.

Stage one, making the alkali. Plant material, characteristically plantain peel, palm bunch, cocoa pod or hardwood, is dried and then burnt to ash. The ash is leached with water and the solution filtered. What this produces is a lye rich in potassium carbonate and related alkaline salts 5. The choice of feedstock matters to the strength of the lye: plantain peel ash yields a more strongly alkaline lye than cocoa pod ash 5.

Stage two, saponification. The lye is combined with plant oils, characteristically palm kernel oil, palm oil, shea butter or coconut oil, and heated. Saponification is the alkaline hydrolysis of the triglycerides in the oil, cleaving them into glycerol and the potassium salts of the fatty acids, which are soap 5. A potassium-based lye gives a soft soap, which is why traditional black soap is characteristically soft or paste-like rather than hard-milled.

What is impressive about it. The maker cannot see the alkali, cannot measure its concentration, and is running a reaction with a real failure mode: too little alkali leaves unsaponified oil and the soap does not work; too much leaves free alkali and the soap burns skin. The traditional process controls this by fixed ratios of specified feedstocks prepared in a specified way, which is exactly how a pre-instrumental chemistry has to work. The medicated soap ọṣẹ used in Yoruba medicine, mentioned in file 01, is this process with plant material incorporated.

Pottery and other materials

Pottery is women's work, built without a wheel, and is treated in 07-arts/07. The technical points relevant here are that the fired clay crucibles used in the glass industry at Igbo Olókun are a pottery product, which links the two industries directly, and that over 800 crucible fragments were recovered in the 2011-2012 excavation alone 2. A glass industry requires a refractory ceramics industry capable of making vessels that survive repeated heating with molten glass in them, and the crucible evidence is therefore evidence about ceramic competence as well as about glass.

What can be claimed

Stated at the level the instruments support: the Yoruba smelted iron competently, controlling slag chemistry by deliberate ore mixing 1. They made glass from raw materials, not merely worked imported glass, using a recipe unique in the Old World, from at least the eleventh century, at a scale unique in West Africa, and traded the product across the western Sudan 2. They ran an alkaline reduction vat and a saponification reaction by trained observation without instruments 345. They cast brass by lost wax to a standard that has never been in dispute, using imported metal.

None of that requires the word "advanced" or any comparison to Europe. It is a description of what the analysis found.