As Nadia Maria El Cheikh has said of the harem in western scholarship on the premodern Islamic world, slavery and the slave trade have yet to be fully historicized.Footnote ¹ While scholars of Abbasid-era slavery have made tremendous contributions in this regard over the last two decades, there remain ample opportunities to expand the chronological and geographic scope of their work. What does it mean to historicize slavery in the medieval Islamic Middle East and why is it worth doing?

One pressing task is to uncover the histories of the domestic slaves found across so many domains of premodern Islamic societies, from the individual household to the palace. Popular and scholarly focus has long been drawn to literary representations of the enslaved concubines, courtesans, eunuchs, and soldiers in imperial and elite settings. As Matthew Gordon has noted, far less attention has been paid to the widespread practice of domestic slavery.Footnote ² The historiographical imbalance is in part due to the oblique manner in which domestic slaves appear in literary sources, often as nameless individuals or an undifferentiated mass. Despite this challenge, there are tremendous opportunities to historicize slavery by studying underused documentary records in addition to traditional literary sources.Footnote ³ Greater use of documentary sources permits fine-grained analysis of the diverse forms and practices of slavery in the Islamic world, their change over time, and their relationship to specific local and global contexts.

The study of premodern slavery affords the chance to think about societies across scales—from the microscale of individual interaction and household formation to the macroscale of statecraft, inter-regional trade, and cultural attitudes. Histories of the slave trade in the medieval Islamic world underscore how such macrofactors combined to shape the coerced migration of slaves across Afro-Eurasia. Moreover, due to the particular economic, legal, and social positions of slaves in Islamic societies, the study of them is relevant to broader questions about daily life, individual agency, and social mobility. In turn, such investigations may tell us about other groups, such as peasants and artisans, that are also marginal in the historiography.Footnote ⁴

As a case in point, scholars may juxtapose documentary and literary evidence in order to distinguish the multiple strands that make up what is frequently termed simply “the slave trade.” Chronicles and travel writing, for example, describe very different modes of slaving that operated in distinct ways based on their purpose and access to slaves. In the first decades of the Fatimid Caliphate in Egypt, rulers in Byzantium, Ifriqiya, al-Andalus, and Yemen exchanged gifts of slaves with the caliphs in Cairo as part and parcel of normal diplomacy (on Nubia, see below).Footnote ⁵ Yet diplomatic exchanges of slaves are both intermittent and specific to imperial court contexts and thus not necessarily representative of the overall trade. In Fatimid Egypt, for example, surviving bills of sale for slaves reveal that there were more domestic slaves from Nubia and other regions of sub-Saharan Africa than from all other world regions combined. This pattern contrasts with the pan-Mediterranean slave origins suggested by the aforementioned diplomatic exchanges. This example underscores the need for scholars to parse out the different modes of slave trading in order to understand more fully the composite geography and economy of slaving activities in a given period and region.

Moreover, traders intensively exploited different regions over time in order to satisfy the demand for slaves. Agricultural slavery in 9th-century Iraq drew overwhelmingly from eastern Africa. As mentioned earlier, domestic slaves in Fatimid and Ayyubid Egypt were predominately from black Africa, and the greater Nubian region in particular. During the Mamluk Sultanate (1250–1517), domestic slaves and concubines identified as Nubian became less common compared to those identified as Abyssinian. These historical variations are significant because they were shaped by geopolitical realities, prevailing patterns of trade and communication, and historical contingencies.

The bearing of such macrofactors upon the history of the slave trade is evident in scholarship on the baqṭ, a 7th-century political agreement between Muslim Egypt and Nubia in which payments in slaves played a part. Modern scholars, not to mention medieval jurists, have disagreed over the precise terms and nature of the baqṭ. For present purposes, the baqṭ is noteworthy because through this compact the Nubians and Egyptians mutually agreed to exchange goods and commodities—primarily, the Nubians would supply a number of slaves in return for foodstuffs and textiles from the Muslims. However, as Jay Spaulding has persuasively argued, historians’ interpretations of the baqṭ—as an annual tribute payment due to Muslim Egypt—is inaccurate. One reason for this misinterpretation lies in the medieval sources themselves. For example, the Egyptian chronicler al-Maqrizi presents a version of the baqṭ that elides its original reciprocal nature and portrays the Nubians as a client state. Modern historians have also attempted to quantify slave exports from Nubia to Egypt on the basis of the baqṭ. Yet the Nubians did not themselves understand the baqṭ as a tribute payment to Egypt, and the Nubian state's delivery of slaves is not attested consistently in historical sources over time. Thus the baqṭ should not be taken to represent a continuous and reliable source of slaves for medieval Egypt. Rather, its intermittent practice and changing portrayal in medieval and modern historiography underscore how historicizing slavery and the slave trade can challenge long-standing misconceptions about the relationship between the Islamic world and its peripheries.Footnote ⁶

Greater Nubia did serve as a major source of slaves for Egypt, but the baqṭ was not responsible for most of them. Travelers to Egypt during the 11th and 12th centuries mention kidnapping and organized slave raids in the hinterlands to the south and southeast of Aswan—particularly between this city and the Red Sea port of ʿAydhab. Kidnapping and raiding appear more likely to represent the slaving activities that provided more perennial sources of new, imported slaves in Egyptian markets and beyond than did diplomatic gifts and exchanges. The networks of professional slave traders that used this region as a reservoir of slaves shipped them by river and overland to more northern markets. In literary sources, it is the Beja people, a group of nomadic tribes that lived between the Nile River and the Red Sea, who are most commonly mentioned as the targets of these raids.Footnote ⁷ Yet we do not find (to my knowledge) slaves identified as Beja in the Arabic, Judeo-Arabic, or Aramaic bills of sale from Egypt that are contemporary to these literary sources. Rather, medieval scribes conflated slaves from the entire region south of Aswan as “Nubian,” or more generically as “black.”Footnote ⁸

These medieval slave trades are worth studying because they throw into relief the relationships between different polities and groups around the Mediterranean and the Near East. Slaving practices serve as a window into diplomacy, economic activity, cultural attitudes (that is, what groups were eligible for enslavement and how did medieval authors represent them), and, last but not least, the coerced migration of peoples. One way for scholars to dig deeper into these issues is to differentiate between the multiple strands of the medieval slave trade and to investigate how these different slaving activities comprised systems with their own histories of change and continuity over time, as well as regional specificities.

Enticing prospects for further research lie in corpora of documentary records that can be read productively alongside the literary sources that have long been the bedrock of medieval Islamic history. Scholars such as Yusuf Raghib have uncovered Arabic documents that inform us about slave trading and the lives of slaves with greater specificity than literary sources. S. D. Goitein and Mordechai Friedman have brought to light sources about slavery from the Cairo Geniza, records that come from the Ben Ezra synagogue in Fustat. While the great majority of these documents originate in Egypt (Raghib's corpus contains material from the countryside, whereas the Geniza documents generally have an urban provenance), some come from other parts of the Islamic world, including Khurasan, the Indian Ocean region, and Mamluk Jerusalem.Footnote ⁹

Literary and documentary sources frequently provide very different accounts of the slave trade. While chronicles and travelogues attest to large shipments of slaves in particular times and places across the Islamic world, documentary sources suggest that slave owners had access to decentralized networks for purchasing domestic slaves that may have operated as part of, or parallel to, larger wholesale trades. In particular, Geniza letters from the 11th and 12th centuries show how Jewish merchants and family members in Egypt wrote to each other and to their associates in Yemen and beyond in order to organize the shipment of individual slaves over long distances. These merchants consigned slaves on ships that carried a mixture of cargo. A 13th-century Muslim businessman's letter from the medieval Red Sea port of Qusayr indicates that it was not just Jewish merchants who arranged the shipment of individual slaves in this way. When we consider this practice in light of contemporary ḥisba (market inspector) manuals, it becomes clearer how a decentralized system of slave trading could have been organized on both small and large scales.Footnote ¹⁰ What remains to be understood is why documentary records furnish so many examples of decentralized, small-scale slave trading and so little evidence of the larger wholesale trading that is described or alluded to in literary sources. The good news is that there are still many underused documents that can be explored for what they tell us about the topic—and among the thousands of unedited Arabic documents there are almost certainly more discoveries to be made.

Documents including bills of sale, letters, endowment deeds, and legal records also expand our knowledge of the lives of individual domestic slaves and thus about their households and societies. In some cases it is possible to reconstruct life histories of individual slave women from early childhood until manumission. For instance, in mid-12th-century Egypt, a court scribe recorded in his notebook details about a sale that separated the two-year-old slave girl Wafaʾ from her mother. Wafaʾ later appears as a maidservant in the dowry of a wealthy Jewish bride. Nearly forty years later, a deed of manumission reveals that Wafaʾ gained her freedom when she was either forty or forty-one years old.Footnote ¹¹ While Wafaʾ had been a slave for four decades, other slave women were freed while still within the typical age range of first marriages in the Egyptian Jewish community. As marriage documents indicate, convert freedwomen commonly married Egyptian Jews and bore children.Footnote ¹²

Slaves surface in the documentary record at crucial junctures in the lives of their owners in ways that reveal a great deal about master–slave relationships. For example, impending death due to illness required slave owners to make critical decisions about what to do with their property and how to provide for children they might have. Deathbed declarations and wills indicate how slave owners sometimes viewed their domestic slaves as practical, non-natal kin whom they relied upon more than other relatives to care for them and their children.Footnote ¹³

Another advantage of documents is that they are much more likely to provide intimate, relatively unstylized views of women's lives than are literary sources. Geniza records and Arabic documents demonstrate that free Jewish and Muslim women commonly owned female slaves. While much has been written about how men used slave women for sex and how female slaves served as child-bearing concubines in the Islamic context, the frequency of female slave ownership in the documentary record suggests that we have much more to learn concerning master–slave relationships between women. The ways in which female slave owners used their slaves, and the terms by which they occasionally freed them, demonstrate that both mastery and patronage culture were themselves gendered.Footnote ¹⁴

Documentary sources for the study of the medieval Middle East are increasingly available through digital platforms in addition to numerous print editions.Footnote ¹⁵ The potential of these sources to reveal new information about slavery and slaving across this vast, culturally variegated zone is immense, and continued exploration of this rich history can also challenge historians of slavery in other times and places to ask new questions about their own subjects and sources.