and their loads came to a halt en route, never reaching their destination. What survived of them is impressive. Along with approximately two thousand standard-sized rods of copper and copper alloy were several containers filled with cowry shells. The site, called Ma’den Ijafen, is believed to date to the twelfth century (see map 1.1).16
It is not known precisely when and where cowries first began to circulate south of the Sahara as a form of currency. The first mention of a cowry currency in the Arabic sources comes from al-Umari, who wrote in 1337/8 about the Mali Empire and what was remembered thirteen years after the pilgrimage to Mecca of its leader, Mansa Musa. Among his informants was one Ibn Amir Hajib, a governing official of Cairo, who had met and befriended Musa and claimed to have had many informative conversations with him. Hajib recounted to al-Umari some of what Mansa Musa had told him about Mali’s history and culture. Gold was apparently collected as tribute but exported, and the main currency circulating in Mali at that time was imported cowry shells. He added that merchants who were specialists in the business of supplying cowry imports profited handsomely from it.17 Ibn Battuta, a Moroccan traveler in West Africa from February 1352 to December 1353, provides a brief but very instructive eyewitness description of cowry currency in his written account of his travels. He noted that cowries were a currency there, specifying Mali and Gao in particular. Most precisely, he stated that in both places he had seen them changing hands at the same rate of 1,150 shells per gold dinar.18 Cowries amounted to small change and were especially useful for making minor transactions. As fixed and durable units, they could circulate by count in single- or multiple-shell units or, if necessary, in larger quantities by weight. Creating standard gold-cowry conversion rates thus provided greater flexibility and social reach to West African commerce by enabling a Muslim gold-weight system to operate in regular and reliable contact with shell-counting systems of polytheistic and rural peoples.19
A growing trade in cotton textiles set off stunning cultural and economic changes in West Africa, first by way of the high prestige and popularity of imported cotton clothing, especially among Muslims, and then with widespread production of it locally. Cotton had been known and probably also produced and woven in the Nile valley in northeastern Africa very early on between the third century BCE and fourth century CE, but it is unclear whether cotton and spinning, weaving, and sewing technologies traveled into other parts of sub-Saharan Africa before the Muslim era in the later first millennium.20 Cotton was a favored fiber for clothing in the Muslim world, and wearing high-quality cotton garments was a sign of elegance and modest good taste. Muslims promoted these values, thereby generating a “cotton culture” that extended beyond Islamic circles and over much of West Africa. Places in West Africa where cotton textile production became well-established—cultivating cotton fiber, spinning cotton yarn, weaving cotton cloth, and tailoring cotton garments—also became producers of cotton currencies that circulated widely.
Cotton technology spread into new areas of West Africa especially during the dry period of ca. 1100 to ca. 1500. Linguistic evidence in the form of words borrowed from Saharan Arabic for key aspects of this cotton culture demonstrates that the dissemination of cotton south of the Sahara owed much to these (presumably) Muslim merchants and their networks. Two separate groups of speakers of the language passed cotton goods and also the knowledge of cotton technology on to peoples living along the desert’s southern shore and lent them also their word for cotton fiber, kútan and gótun, dialectal variants of the Arabic qutn. The general geographical locations of these two groups of Arabic speakers and the geographical locations of sub-Saharan speech communities who borrowed these Arabic words for cotton both match well with the known geographical locations of the main western and central trans-Saharan trade routes. A very general time frame for when this transfer of knowledge could have occurred can be estimated based on the growing prominence of Muslims in the towns, cities, and courts of tropical West Africa early in the second millennium. Hence these events could have been set in motion in the tenth or eleventh century. Convincing evidence in support of this time frame comes from archaeology. Material remains in the form of spinning tools and locally made cotton textiles and tailored garments together corroborate the linguistic evidence by dating from a broad period spanning the tenth to seventeenth centuries.21 A superb example is this remarkably well-preserved cotton tunic, illustrated in figure 1.1, which comes from one of the cave burials at Sanga, an archaeological site located in the Bandiagara escarpment in what is now Mali. It is made of narrow cotton strips sewn together and tailored, and like many of the roughly five hundred other garment fragments discovered at the site, it displays a repeat pattern woven with white and indigo-dyed thread. Skeletal remains associated with this tunic date to the eleventh or twelfth century.22
FIGURE 1.1 Cotton tunic, eleventh or twelfth c. CE, Cave C, Sanga, Bandiagara escarpment, Mali. Nationaal Museum van Wereldculturen, The Netherlands, RV-B237–755.
However, precisely when and where major cotton production centers arose is very incompletely known, especially for the period before 1500. The earliest written information about cotton and cotton textile manufacture in sub-Saharan West Africa comes from al-Bakri’s eleventh-century compilation in which he states that the currencies of Sila (in the lower Senegal River valley) consisted of sorghum, salt, copper rings, and lengths of finely woven cotton. Weavers produced the cloth in a neighboring town where many households grew cotton on a small scale, apparently as a perennial. Al-Umari noted in the fourteenth century that one of the currencies in Kanem, in the vicinity of Lake Chad well to the east, was a locally woven cotton strip cloth, and that Mali, far to the west, was reportedly cultivating much cotton and producing high-quality cotton textiles.23 By the time of the Guinea trade, European explorers and merchants took note of places such as Senegambia on the Upper Guinea Coast and the Bight of Benin in the Gulf of Guinea where cotton textiles were available for export to other coastal markets. As Euro-African trade expanded, locally made cotton currencies remained necessary for purchasing provisions and paying for labor. They thus continued to be produced in competition with the higher-priced cotton imports from overseas.24
West African cotton currencies circulated as narrow strips or breadths of various widths—woven on local handlooms and calculated as standard units based on the number of lengths that would make a finished cloth or item of clothing. Al-Bakri’s report about the currencies in the lower Senegal River valley noted the cotton currency there circulated in lengths he referred to as izar, the Arabic word for mantle, the most basic and versatile of garments. Kanem’s cotton currency as described by al-Umari circulated in lengths of ten cubits (the forearm length from the middle finger tip to the elbow bottom), and purchases could be made using fractions from one cubit upwards. Imported cowries, beads, copper pieces, and silver coins also circulated there as currency, but their values were calculated in terms of the local cotton cloth units. Al-Umari also noted that the excellent white woven cotton of Mali was called kamisiya, from the Arabic qamis, a generic term for shirt.25
These cotton currency units, based on linear measures of cloth or sometimes in reference to a garment, continued to be standard into and throughout the era of Atlantic trade. Álvares de Almada described a former time on the Upper Guinea Coast, probably when he lived there in the mid-sixteenth century, when Luso-African merchants could safely lodge with local nobles and purchase slaves very cheaply. They paid for them in cows and cloths called sigas, which he described as a fixed length of the cotton currency called teada. Major cotton currencies along the Guinea Coast were known in many West African languages by the vernacular names for wrapper-size cloths.26
The passage of more goods and commodity currencies through the Sahara, the Sahel, and the savannas, rainforests, and coastal regions depended on the variations of seasonal working patterns, the frequencies and locations of trans-Saharan caravan arrivals, and the fluctuating intensities of local mining, processing, and workshop production. The West Africa that European merchants encountered was much more than the “Land of Gold” they dreamed of. Though not a fully integrated economic system or single market, it was a dynamic multicentric