Map 1.2 The monsoons
The spatial extent and the differential pattern of the monsoons thus helped define the normal radius of action of Indian Ocean dhows. They tended to favour the northern part of the East African coast which enjoyed a longer trading season between the two monsoons, especially for Indian dhows. The southern coast, on the other hand, is at the extreme periphery of the monsoon system, and it thus experienced a shorter trading season, forcing dhows to ‘winter’ in East African waters and sail back with the tailend of the south-west monsoon in August or September. This required a more elaborate entrepôt system for effective exploitation of the commercial resources of the southern coast, especially that beyond Cape Delgado.
On the one hand, therefore, the narrow coastal belt failed to provide an adequate productive base for the economies of the city-states. They were thus dependent on the transit trade between the African interior and the regions beyond the Indian Ocean. On the other hand, the longitudinal alignment of the coast accessible to the monsoon, even if access was differential, provided the various ports with potentially independent commercial bases. This imprinted on the coastal economy and politics a generally acephalous and even fissiparous tendency, and a spirit of political independence despite, or because of, the individual economic dependence of the city-states on international trade.
The mercantile civilisation of the Swahili coast
From the beginning of the Christian era African social formations along the East African coast, which were initially geared to production of use values for direct consumption or barter, were induced by international trade to produce surplus products for exchange. A surplus of food grains and mangrove poles was exported to the food- and timber-deficient south Arabian coast and the Persian Gulf at various times, though these mundane staples are hardly ever mentioned in the few available sources before the sixteenth century. External demand, however, also induced the production of certain luxuries for which there was probably no local use. While new sources of wealth were thus opened up, they may have entailed the diversion of labour from other essential economic activities. Early Greek and Roman demand for ivory induced south Arabian traders to extend their commercial activities down the East African coast as early as the second century BC. After the collapse of the Roman empire its place was taken by India and China which remained the main markets until the beginning of the nineteenth century. Gold was procurable in sufficient quantities only from Zimbabwe. As demand in Asia increased at the beginning of the second millennium AD, greater amounts of labour, including that of young females, were mobilised to mine gold. Coastal shipping was extended south from Kilwa to Sofala where it met the land routes from Zimbabwe. The medieval glory of Kilwa was directly dependent on this entrepôt trade.3
International trade induced not only the diversion of labour from one economic activity to another within the same social formation but, at various times, even the physical transfer of that labour. The Periplus of the Erythraean Sea, which was probably written in the late first or early second century AD, mentions the export of slaves, but only from the Horn of Africa. From the seventh to the ninth century, however, there was a massive demand for slave labour to reclaim the marshlands of southern Iraq. Severe exploitation and oppression of a large number of slaves concentrated near Basra led to a series of slave revolts from the end of the seventh century, and culminated in the famous ‘Zanji rebellion’ in the ninth century when the rebels controlled southern Iraq for fourteen years (868–83). The revolt was suppressed but, by their resistance, the slaves had ensured the failure of one of the few cases of agricultural exploitation based on slave labour in Muslim history. The subsequent economic decline of the Middle East meant that the slave trade did not again attain such massive proportions until the eighteenth century.4
In return for African products merchant capital made available to the East African social formations manufactured goods and luxuries which helped expand the sphere of circulation. During the first century AD, imports included metal tools and weapons, but also wine and wheat ‘to gain the goodwill of the barbarians’. By the thirteenth century they included beads, Chinese porcelain and cloth, part of which was consumed by the affluent merchant classes in the city-states themselves. Such imports may at times have offered stiff competition to local bead and textile industries. These industries, which apparently flourished at Mogadishu, Pate, Kilwa and elsewhere, showed signs of decline, although at Kilwa they were killed only with the coming of the Portuguese.5
While merchant capital thus helped to expand the production and circulation of commodities, it may also have helped to impart to East African economic structures a certain lopsided character, with overdeveloped commercial, mining and hunting sectors, and a more stunted industrial sector. Moreover, international trade appears to have been conducted on the initiative of foreign traders to supply the needs of their own homelands, and generally in their own ships. As late as the eleventh century, al-Idrisi commented that ‘the Zanj have no ships to voyage in, but use vessels from Oman and other countries’.6 This suggests inequality in the level of technological and socio-economic development between the trading partners. Under these circumstances there may have been an imposition of a pattern of development on the East African social formations that made them dependent on international trade, and that was more beneficial to the more developed social formations across the ocean. This tendency, however, should not be exaggerated for the period before the rise of capitalism since commodity production, largely of luxuries, played only a limited role in total production.
International trade stimulated the growth of market towns, some of which may have been initially established by the indigenous people themselves. In these market towns trade provided a base for the emergence of a ruling merchant class which appropriated its middleman’s profit from commodities passing through its hands. It ruled over a coastal population that was no longer undifferentiated. In the first century AD the Periplus talks of ‘men of the greatest stature, who are pirates, inhabit the whole coast and at each place have set up chiefs’.7 Commerce may have begun to undermine the tribal constitution, paving the way for the emergence of a class society and a state. By the second century, Ptolemy talks of Rhapta, one of the market towns, as a ‘metropolis’, which Gervase Mathew suggests meant the capital of a state. A thirteenth-century Chinese source describes a stratified society along the coast consisting of the ‘bareheaded and barefooted’ commoners who lived in ‘huts made of palm leaves’, and a ruling class whose stone habitations are alone visible among the coastal ruins today. The populations of some of these city-states are said to have numbered several thousand, and yet none of the ruined cities seem to have more than about fifty stone-built houses.8
In these coastal city-states flourished a civilisation which was prosperous but not self-reliant. The cultural level attained surprised the early European visitors, one of whom was moved to compare the mosque at Kilwa with that at Cordova. A modern archaeologist has been impressed by ‘the complexity, luxury, variety and sensitivity’ of design of the Husuni Kubwa palace at Kilwa. The merchant class, according to the early sixteenth-century Portuguese factor, Duarte Barbosa, exhibited great opulence and luxury:
The kings of these isles [Pemba, Mafia and Unguja] live in great luxury; they are clad in very fine silk and cotton garments, which they purchase at Mombaca from the Cambaya merchants. The women of these Moors go bravely decked, they wear many jewels of fine Çofala gold, silver too in plenty, earrings, necklaces, bangles, and bracelets, and they go clad in good silk garments.9
But that civilisation was mercantile and dependent, as evidenced by the storage rooms which backed the Husuni Kubwa palace. Its dependence on the international connection was not confined to the economy, but permeated the whole range of cultural and social life of the Swahili city-states. Religion and the fundamental bases of culture generally came from abroad, though they were gradually indigenised. The standing architecture appeared on the coast ‘fully fledged’ in the earliest known buildings,