To determine the dimensions of the northern demand for slaves, it is necessary to understand the uses to which slaves were put. The social transformation of Oman into a mercantile state and the expansion of slave-based date production from c. 1700 had created a demand for agricultural slaves. It is extremely difficult to arrive at an estimate of the annual demand for such slaves in Oman. The only clue is the statement in the Omani chronicle that Imam Saif b. Sultan (1692–1711) owned 1,700 slaves and one-third of all the date palms in Oman.7 As the biggest landowner and perhaps the individual most involved in commerce, the Imam probably had a higher proportion of slaves to date palms than the smaller peasants and interior tribesmen. He may also have had a proportionately larger number of domestic slaves, who were presumably included in the total number of slaves owned by him. Assuming, nevertheless, the ratio of 1:3, the slave population of Oman at the beginning of the eighteenth century would have been about 5,000. Assuming further an attrition rate of about 10 per cent8, the annual demand for slaves in Oman would have been about 500. There are no precise quantitative data for the trade except a reference to between two and four large Omani ships which bought slaves from East Africa apparently on royal account, apart from about ‘300 blacks’ exported on ships locally built in East Africa.9 The numbers may have grown during the eighteenth century as the social revolution was consummated. Whatever may have been the absolute figure for the eighteenth century, however, there seems to be little evidence for any significant increase in the demand during the nineteenth century when Oman appears to have been economically stagnant or declining, particularly with the migration of many well-to-do merchants and landowners and the transfer of the capital to Zanzibar.
Map 2.1 The East African slave trade
African slaves were apparently not used in agricultural production in Iran, although they were widely used in maritime activities in the Persian Gulf. Africans were ubiquitous among dhow crews of the western Indian Ocean during the nineteenth century, but there is little quantitative evidence for the number of slaves absorbed in seafaring and in the docks to service Indian Ocean commerce. Africans also constituted a sizeable proportion of pearl divers in the Gulf who were estimated to number between 27,000 and 30,000 in the mid-nineteenth century. An early twentieth-century survey suggests that about one-third of the divers were Africans.10
Slaves were also used in the army. Although warfare in Oman appears to have been based primarily on tribal levies, the secularisation of the Omani state and the weakening of religious and tribal loyalties did suggest the need for a loyal standing army under the Busaidi. Thus Ahmed b. Said purchased 1,000 East African slaves at one time in the 1740s. However, the standing army appears to have been small, and the more prominent component, especially during the nineteenth century, appears to have been Baluchi mercenaries. In 1802 it consisted of 300 slaves and 1,700 Baluchis, Sindhis and Arabs, and in 1809 there were 2,000 mercenaries, but there is no mention of slaves at this time.11
A larger number of unproductive slaves were absorbed in the domestic sector to perform menial domestic chores or as concubines. The demand was for young slaves to be groomed in the household. In 1831 their ages ranged between 10 and 14, and evidence from slave captures suggests a fair balance between the sexes.12 We have no estimates of the size of this population. During the eighteenth century it may have been growing as Omani society was being transformed with the emergence of a wealthy merchant class. The migration of part of this class to Zanzibar after the commercial crash in Oman at the end of the Napoleonic war may have led to stagnation or decline in the annual demand for such slaves.
The rough estimates for each of the uses of slaves are intended to provide only an idea of the magnitude of the demand for slaves against which the scattered numerical estimates by contemporary observers may be judged. We are fortunate in possessing some very precise quantitative data about the slave trade at the receiving end collected by British officials who cannot be accused of trying to minimise the Arab slave trade. According to these officials Muscat and Sur were ‘the principal, if not only primary ports to which all slaves . . . were brought, and whence they are carried into Turkey [i.e. Turkish-controlled Iraq], Persia, Scinde, the Arabian states’, and even India.13 The British Resident in the Gulf reported in 1831 that normally 1,400 to 1,700 slaves were imported into Muscat, although during the preceding season only 1,150 to 1,200 were imported, including 250 to 300 who were smuggled. Of these slaves three-quarters were imported from ‘Sowahel or the coast of [East] Africa’, the remainder being Ethiopians. In 1830 over half of the slaves, i.e. about 500 East African slaves, were re-exported to the Gulf, while the remaining 400 were absorbed within Oman.14
In 1841 the British kept a register at the strategic island of Kharg of all the slave dhows passing to the northern end of the Gulf. Although the Gulf slave traffic normally started from around the beginning of July, the register itself began on the nineteenth of the month. By the end of the slave season 118 boats had passed, carrying a total of 1,217 slaves of whom 640 were females and 577 males. Most of these slaves were re-exports from Muscat and Sur or via third ports, and only one boat with twelve slaves apparently came directly from ‘Sowahel’.15 The register does not make a distinction between East African and Ethiopian slaves, but assuming the proportion given in 1831, and after making allowance for the first eighteen days of July, we get about 1,000 East African slaves reaching the northern end of the Gulf.
A most unfortunate gap remains in the data for the trade of Sur, the other, but smaller, port importing slaves directly from East Africa. However, its re-export to Kharg Island in 1841 was roughly in the proportion of 7:10 compared with Muscat. Assuming that such a proportion held true for the total slave trade, it would appear that the total number of East African slaves imported into Muscat and Sur in 1830 was about 1,500, or ‘normally’ between 1,800 and 2,200. Since these were the ‘only primary ports’ for the slave trade from East Africa to Oman and the Gulf, the dimensions of the Omani slave trade were much more modest than hitherto assumed by colonial and modern historians.
East African slaves were also imported into the Red Sea region. Some were later re-exported to the East. For example, in 1835 three Mahra dhows were intercepted at the Gujarat port of Porbandar carrying seventy-nine slaves.16 Some, however, were locally absorbed, and they formed pockets of African population in southern Arabia. The former headquarters of the National Liberation Front of South Yemen was at a place called Zinjibar which, as the name suggests, was formerly populated by black people. However, the number of slaves imported from East Africa was probably small, partly because the region was closer to Ethiopia, another major source of slaves. In 1840 the French Consul at Jeddah reported 500 East African slaves entering the Red Sea.17
Table 2.1 The northern slave trade, 1831 and 1841
Sources: Wilson to Norris, 28 January 1831, MA, 1/1830–31; Robertson to Willoughby, 4 March 1842, MA, 78/1841–2, pp. 346–62.
The demand for slaves in the more densely populated India, which had its own fairly large poor population, was probably very small. In 1838 only twenty-six slaves were imported into the main Kutch port of Mandvi which, before being superseded by Bombay, was the Indian port with the largest trade with East Africa. They were for the most part domestic