Race and Slavery in the Middle East: An Historical Enquiry

by Bernard Lewis

  There is a fair amount of information on slave prices, most of it too heterogeneous in date and provenance to provide more than a general impression. The best-documented data come from medieval Egypt and show a remarkable consistency in price levels. Slave girls averaged twenty dinars (gold pieces), corresponding, at the rate of gold to silver current at that time, to 266 dirhams (silver pieces). Other medieval data show somewhat higher prices. Black slaves seem to have cost from two to three hundred dirhams; black eunuchs, at least two or three times as much. Female black slaves were sold at five hundred dirhams or so; trained singing girls or other performers, at ten or even twenty thousand. White slaves, mainly for military purposes, were more expensive. Prices of three hundred dirhams are quoted for Turks near the source in Central Asia, and much higher prices elsewhere. In Baghdad they fetched four to five hundred dirhams, while a white slave girl could be sold for a thousand dinars or more.43 The mid-nineteenth-century German report from Turkey quotes prices of four thousand to five thousand piasters, or two hundred to three hundred dollars, as the current price in Istanbul for a "trained, strong, black slave," while "for white slave girls of special beauty, fifty thousand piasters and more are paid.i44 In general, eunuchs fetched higher prices than other males, younger slaves were worth more than older slaves, and slave women, whether for work or pleasure, were more expensive than males. Olufr Eigilsson, an Icelandic Lutheran pastor who was carried off to captivity with his family and many of his flock when his native village was raided by Barbary Corsairs in 1627 and who wrote an account of his adventures, notes that his young maidservant was sold for seven hundred dollars 4` and later resold for a thousand.

  Slaves were employed in a number of functions-in the home and the shop, in agriculture and industry, in the military, as well as in specialized tasks. The Islamic world did not operate on a slave system of production, as is said of classical antiquity, but slavery was not entirely domestic either. Slave laborers of various kinds were of some importance in medieval times, especially where large-scale enterprises were involved, and they continued to be into the nineteenth century. The most important slaves, however, those of whom we have the fullest information, were domestic and commercial, and it is they who were the characteristic slaves of the Muslim world. They seem to have been mainly blacks, with some Indians, and some whites. In later times, for which we have more detailed evidence, it would seem that while the slaves often suffered appalling privations from the moment of their capture until their arrival at their final destination, once they were placed with a family they were reasonably well treated and accepted in some degree as members of the household. In commerce, slaves were often apprenticed to their masters, sometimes as assistants, sometimes advancing to become agents or even business partners.

  The slave and also the liberated ex-slave played an important part in domestic life. Eunuchs were required for the protection and maintenance of harems, as confidential servants, as palace staff, and also as custodians of mosques, tombs, and other sacred places. Slave women were required mainly as concubines and as menials. A Muslim slaveowner was entitled by law to the sexual enjoyment of his slave women. While free women might own male slaves, they had of course no equivalent right.

  The economic exploitation of slaves, apart from some construction work, took place mainly in the countryside, away from the cities, and like almost everything else about rural life is sparsely documented. The medieval Islamic world was a civilization of cities. Both its law and its literature deal almost entirely with townspeople, their lives and problems, and remarkably little information has come down to us concerning life in the villages and the countryside. Sometimes a dramatic event like the revolt of the Zanj in southern Iraq or an occasional passing reference in travel literature sheds a sudden light on life in the countryside. Otherwise, we remain ignorant of what was happening outside the cities until the sixteenth century, when for the first time the surviving Ottoman archives make it possible to follow in some detail the life and activities of rural populations-and the exploration of this material has still barely begun. The common view of Islamic slavery as primarily domestic and military may therefore reflect the bias of our documentation rather than the reality. There are occasional references, however, to large gangs of slaves, mostly black, employed in agriculture, in the mines, and in such special tasks as the drainage of marshes. Some, less fortunate, were hired out by their owners for piecework. These working slaves had a much harder life. The most unfortunate of all were those engaged in agricultural and other manual work and large-scale enterprises, such as for example the Zanj slaves used to drain the salt flats of southern Iraq, and the blacks employed in the salt mines of the Sahara and the gold mines of Nubia. These were herded in large settlements and worked in gangs. Large landowners, or crown lands, often employed thousands of such slaves. While domestic and commercial slaves were relatively well-off, these lived and died in wretchedness. Of the Saharan salt mines it is said that no slave lived there for more than five years. The cultivation of cotton and sugar, which the Arabs brought from the East across North Africa and into Spain, most probably entailed some kind of plantation system. Certainly, the earliest relevant Ottoman records show the extensive use of slave labor in the state-maintained rice plantations.41 Some such system, for cultivation of cotton and sugar, was taken across North Africa into Spain and perhaps beyond. While economic slave labor was mainly male, slave women were sometimes also exploited economically. The pre-Islamic practice of hiring out female slaves as prostitutes is expressly forbidden by Islamic law but appears to have survived nonetheless. 47

  The military slaves were in a sense the aristocrats of the slave population. By far the most important among these were the Turks imported from the Eurasian steppe, from Central Asia, and from what is now Chinese Turkistan. A similar position was occupied by Slavs in medieval Muslim Spain and North Africa and, later, by slaves of Balkan and Caucasian origin in the Ottoman Empire. Black slaves were occasionally employed as soldiers, but this was not common and was usually of brief duration.

  Certainly the most privileged of slaves were the performers. Both slave boys and slave girls who revealed some talent received musical, literary, and artistic education. In medieval times most singers, dancers, and musical performers were, at least in origin, slaves. Perhaps the most famous was Ziryab, a Persian slave at the court of Baghdad who later went to Spain, where he became an arbiter of taste and is credited with having introduced asparagus to Europe. Not a few slaves and freedmen have left their names in Arabic poetry and history.

  In a society where positions of military command and political power were routinely held by men of slave origin or even status and where a significant proportion of the free population were born to slave mothers, prejudice against the slave as such, of the Roman or American type, could hardly develop. Where such prejudice and hostility appear-and they are often expressed in literature and other evidence-they must be attributed to racial more than to social distinction. The developing pattern of racial specialization in the use of slaves must surely have contributed greatly to the growth of such prejudice.

  During the last half century or so, the word "race," in most Western languages, has undergone substantial and significant changes of meaning. Much confusion and misunderstanding have been caused by the failure to recognize these changes, still more by the survival of earlier meanings when new ones have already been generally accepted. As late as the midcentury, the word "race" was still commonly used in Europe, and occasionally in the United States, to designate what we would nowadays call an ethnic group, that is to say, a group defined primarily by linguistic and other cultural, historical, and in some sense geographical criteria. In Britain the word was generally, even officially, used to designate the four components that made up the common British nationality.' Similarly, India was inhabited by a great variety of socalled races, speaking different but closely related languages and sharing a common civilization. Sometimes, the term "race" was used in a b
roader and looser sense, to denote a group of peoples, speaking related languages. It was in this sense that philologists and ancient historians spoke of the Semites, the Indo-Europeans, and other linguistically defined families of peoples.

  As so often happens, social scientists took a word of common but imprecise usage and gave it a precise technical meaning. For the anthropologist, a race was a group of people sharing certain visible and measurable characteristics, such as hair, pigmentation, skull measurements, height, and other physical features. Races thus consisted of such categories as whites, blacks, Mongols, and the like. These might be sub-divided-thus, for example, whites could be classed as Nordic, Alpine, or Mediterranean. This kind of race, though obviously overlapping to some extent with ethnic groupings, was independent of ethnic features. Different races could share a culture. Different cultures could divide a race. By the strictly physical definition, even members of the same family, with different genes, could belong to different races.

  In current American usage, which has now spread to most other countries, the word "race" is used exclusively to denote such major divisions as white, black, Mongolian, and the like.' It is no longer applied to national, ethnic, or cultural entities, such as the English or the Irish, the Germans or the Slavs, or even the Japanese, who are now seen as being part of a much larger racial grouping found in East Asia.

  In this modern sense, race was of minor importance in antiquity. Where modern scholarship has discerned racial tension and hostility, it has been in the earlier sense of race as an ethnic or national group, such as the Egyptians, the Assyrians, the Israelites, and others defined by language, culture, and religion. Though the ancient civilizations of the Middle East show considerable diversity, there is no great racial difference between their peoples. In friezes and other pictorial representations, aliens are distinguished by their garb, their hair, their beards, and their accoutrements, rather than by physical features. The nose alone-used rather in the manner of a modern cartoonist-seems to have provided the ancient artist with a physical symbol of national identity.' No doubt there were differences of predominant physical type between, say, Egyptians and Assyrians, but these were no greater than the differences between the different peoples of Europe. In anthropological terms, the major peoples of the Middle East who have left their mark in history-the Egyptians, the Sumerians and Akkadians, the Israelites, the Aramaeans, the Hittites, the Medes, and the Persians and even, later, the Greeks and the Romans-exhibited no marked contrasts of racial type.

  Like every other society known to human history, the ancient Middle Eastern peoples harbored all kinds of prejudices and hostilities against those whom they regarded as "other." But the "other" was primarily someone who spoke another language (the prototypal barbarian) or professed another religion (the Gentile or heathen or-in Christian and Islamic language-the infidel). There are many hostile references to the "others"-among Jews about Gentiles and heathens, among Greeks about barbarians, among Romans about almost everybody. It would be easy to assemble a fine collection of ethnic slurs from Greek and Latin literature-but they are ethnic, not racial, slurs. When Juvenal, irked by the Syrian presence in Rome, complains that the Orontes had overflowed into the Tiber or when Ammianus Marcellinus, who was himself a Syrian, said of the Saracens, meaning the Bedouin, that he did not find them desirable either as friends or as enemies, they were making cultural, not racial, statements.' This and other similar anti-Arab remarks, and the attitude which they express, did not prevent an Arab chieftain from becoming the Roman Emperor Philip or a Syrian local priest from becoming the Emperor Heliogabalus.

  Other races were of course known. The ancient Egyptians were closely acquainted with their black southern neighbors and sometimes portray them, in words or pictures, with characteristic Negroid features. But there is no evidence that they regarded them as inferior for that reason. The much-cited inscription of Pharaoh Sesostris III, in the nineteenth century B.C., barring, or rather restricting, access by blacks to Egyptian territory, is a normal security precaution on a vulnerable frontier, where many wars had been fought. It is no more a sign of racial prejudice than are numerous similar restrictions on numerous other frontiers.' The Persians, Greeks, and later Romans had some occasional contacts with China, and rather more with Ethiopia, which was a known part of the civilized world even in biblical times. But these countries were very remote, and contacts with them were few. Ethiopia and China were both respected, and there is no real evidence in Jewish, Greek, or Roman sources of lower esteem for darker skins or higher esteem for lighter complexions.' Nor were there slave races. Foreigners, especially if barbarians, were enslavable. The ancients, like the rest of humanity, believed foreigners to be inferior. Conquest confirmed that belief and, through the universal rule of enslaving the conquered, provided it with practical application. Classical writers, from Aristotle onward, stated the general principle that there are races suited by nature to slavery, but although there are occasional references to this or that people as fitting this description, these are only passing examples of wit or spite, in no sense amounting to any kind of scientific or philosophical statement.'

  The advent of Islam created an entirely new situation in race relations. All the ancient civilizations of the Middle East and of Asia had been local, or at most regional. Even the Roman Empire, despite its relatively larger extent, was essentially a Mediterranean society. Islam for the first time created a truly universal civilization, extending from Southern Europe to Central Africa, from the Atlantic Ocean to India and China. By conquest and by conversion, the Muslims brought within the bounds of a single imperial system and a common religious culture peoples as diverse as the Chinese, the Indians, the peoples of the Middle East and North Africa, black Africans, and white Europeans. Nor was this coming together of races limited to a single rule and a single faith. The Muslim obligation of pilgrimage, which requires that every adult Muslim, at least once in his lifetime, must go on a journey to the holy places in Mecca and Medina, brought travelers from the remotest corners of the Muslim world, covering vast distances, to join with their fellow believers in common rites and rituals at the very center of the Islamic faith and world. The pilgrimage, probably the most important factor of individual, personal mobility in pre-modern history, combined with the better-known forces of conquest, commerce, and concubinage to bring about a great meeting and mixing of peoples from Asia, Europe, and Africa.

  At different times and places, Muslims have responded to the challenge of racial encounter and cohabitation in a variety of ways. These responses are reflected, in sometimes striking contrast, in both old and recent literature. One view of Muslim racial attitudes, widely accepted in the modern West, is expressed in a famous passage in Arnold Toynbee's Study of History, documented, like so much in that massive work, with a personal experience:

  For instance, the Primitive Arabs who were the ruling element in the Umayyad Caliphate called themselves "the swarthy people," with a connotation of racial superiority, and their Persian and Turkish subjects "the ruddy people," with a connotation of racial inferiority: that is to say, they drew the same distinction that we draw between blondes and brunettes but reversed the values which we assign to the two shades of white. Gentlemen may prefer blondes; but brunettes are the first choice of Allah's "Chosen people." Moreover, the Arabs and all other White Muslims, whether brunettes or blondes, have always been free from colour-prejudice vis-a-vis the non-White races; and, at the present day, Muslims still make that dichotomy of the human family which Western Christians used to make in the Middle Ages. They divide Mankind into Believers and Unbelievers who are all potentially Believers; and this division cuts across every difference of Physical Race. This liberality is more remarkable in White Muslims today than it was in the White Western Christians in our Middle Ages; for our medieval forefathers had little or no contact with peoples of a different colour, whereas the White Muslims were in contact with the Negroes of Africa and with the dark-skinned peoples of India from the be
ginning and have increased that contact steadily, until nowadays Whites and Blacks are intermingled, under the aegis of Islam, throughout the length and breadth of the Indian and the African Continent. Under this searching test, the White Muslims have demonstrated their freedom and race-feeling by the most convincing of all proofs: they have given their daughters to black Muslims in marriage."

  The Arabs, that is to say, as swarthy whites, felt superior to the fairerskinned peoples to the north of them but were entirely free from any feeling of color prejudice directed against their darker southern neighbors. Prejudice against those of fairer skin, clearly, is felt to be no more than an amusing paradox. What counts is prejudice against those of darker skin, and since this is lacking, the Arabs and Islam may be pronounced free from infection.

  Similar views are expressed in a number of other writings, and date back, it would seem, to the nineteenth century, and more especially to the American Civil War, which brought the linked issues of race and slavery into sharp focus before world opinion. The freedom of the Islamic world-as opposed to Western Christendom-from racial prejudice and discrimination rapidly became commonplace.

  The Middle East is an ancient land of myths in which the mythopoeic faculty-the ability to create myths, to believe in them, and to make others believe-has by no means died out. It would be wise to subject any widely held assumption regarding this area to critical scrutiny.