In​ 2012 the 96-year-old Bernard Lewis went on American radio to promote his book Notes on a Century: Reflections of a Middle East Historian. Neal Conan, the host of NPR’s Talk of the Nation, introduced him as a man unafraid of sensitive subjects and invited listeners to call in with questions about ‘taboos’ in Middle Eastern history. Conan asked whether race and slavery were linked in the Islamic world, and Lewis replied: ‘Very much so. Slavery was not an American invention. It was imported from the Old World. When the European and later American slave dealers went to Africa to get slaves, they didn’t go out and hunt for them, they bought them … from local slave dealers. It was already a well-established industry, going back many centuries.’ Conan said that ‘endless documentation’ seemed to be emerging on slavery in the US and Europe, but ‘very little about slavery in the Middle East’. The reason, Lewis told him, was that ‘it’s not politically correct … One must avoid anything that is critical in dealing with the Middle East. You can be critical of anything European or American. You can be critical of anything Christian or Jewish, but not beyond.’ Later in the programme, Lewis – a self-described Orientalist – offered an opinion he claimed was controversial but was in fact representative of much Western writing on slavery in Muslim societies, including his own book Race and Slavery in the Middle East (1990). Arab participation in the slave trade, he observed, had ‘started earlier and continued later’ than in Europe and America. Slavery was, he implied, even more characteristic of Islamic countries than of the West.

Critics of Orientalism didn’t think Lewis was offering a dispassionate corrective to fashionable trends. Edward Said, who was largely responsible for turning the ‘Orientalist’ label into an insult (Lewis called it ‘word pollution’), had dismissed his work as ‘full of condescension and bad faith’ towards Muslims. Said condemned the many scholars of Islam, Lewis among them, who felt it was their duty to ‘mount attacks’ on Islamic societies, accusing them of putting their academic authority to work in the service of Western hegemony. In Covering Islam (1981), Said discussed a Princeton seminar – not involving Lewis – during which ‘much was made of African fear and resentment of Arab Muslims’ and of the claim that they had ‘depopulated’ Africa as part of a system of slavery. As Said saw it, the seminar topic had been chosen as part of a broader effort to undermine relations between African and Arab Muslims, with an eye to backing American foreign policy in the region. In a sense, Said and Lewis agreed that the literature on slavery in Islam tended to be overly political, to the detriment of good scholarship. But they disagreed about which side was doing the politicking.

An important aspect of Lewis’s approach was his comparison of slavery in the Islamic world with the Atlantic slave trade. By claiming that the Muslim history of slavery was worse, Lewis deflected scrutiny from the transatlantic trade. Similar arguments are now made by the European far right. The French journalist and politician Éric Zemmour has claimed that the issue of historical slavery is ‘banal’, since ‘all civilisations’ practised it. Why should Europeans be the only ones to feel guilty? The idea that Muslims did it first and did it worse has been used more broadly to contextualise European colonialism. In 2021 Emmanuel Macron told a group of young Algerians that he was ‘fascinated’ by Turkey’s success in persuading their country that the French were its ‘only colonisers’ – a reference to Ottoman rule over North Africa, which began in the 16th century. Historians were quick to object, arguing that Macron’s comments collapsed distinct historical categories in order to draw a false equivalence between premodern conquest and modern colonialism, with the predictable effect of inproving Europe’s record.

This is not to diminish the use of slavery in the Muslim world. As the historian Ehud Toledano has shown, there has long been a tendency within Islamic society to minimise its violence. According to this view, a ‘softer’ form of slavery prevailed in the Muslim world, one that was domestic, almost familial. These arguments can be traced back to 19th-century Islamic defenders of slavery. Many members of the Ottoman elite, for instance, felt that Europe’s (and especially Britain’s) energetic abolitionism made use of the ‘civilising mission’ rhetoric also employed to justify Western imperialism. To counter this, Ottoman dignitaries adopted what Toledano calls a ‘defensive reaction’. They argued that Islamic slavery was ‘humane’ in comparison to its ‘savage’ American counterpart, and did not depend on and sustain a brutal plantation economy, as it did in the West Indies or the antebellum South. Instead, enslaved people were welcomed as family members in elite Muslim households and frequently manumitted after a few years of service. This myth of benevolent slavery has largely been debunked by scholars such as Toledano, Eve Troutt Powell and Yusuf Hakan Erdem, who have found that physical and sexual violence was common, and that exploitation, not protection, was the rule.

Justin Marozzi’s Captives and Companions offers a ‘history of slavery and the slave trade in the Islamic world’ from the rise of Islam in the seventh century to the present. The comparison with Atlantic slavery appears early on. The number of people enslaved in North Africa and the Middle East since the advent of Islam ‘approximates to the total captured and sold into slavery in the Atlantic slave trade’, Marozzi writes. He wants to underline the equivalence, but is conscious too that the Atlantic trade lasted for four centuries while slavery in the Muslim world lasted for fourteen. Marozzi highlights other important differences between the two forms of slavery. In contrast to the near uniformity of race and status among enslaved people in the Atlantic trade, the category of slave in Muslim contexts encompassed a bewildering variety of ethnic backgrounds and social positions. A powerful Ottoman vizier with slaves of his own and a labourer on a date plantation in Zanzibar could both technically be described as ‘slaves’, but had little else in common. The geographical and chronological span of slavery in Muslim societies accounts for much of this diversity. Marozzi focuses on North Africa, the Eastern Mediterranean and the Middle East, with forays into sub-Saharan Africa, Eastern Europe, the Balkans and the Caucasus. This leaves the Muslim communities of South and South-East Asia outside the scope of the book. Even so, practices of enslavement in eighth-century Baghdad were very different from those in 16th-century Morocco, 19th-century Istanbul or 20th-century Sudan.

Islam did not introduce a new form of slavery. In the seventh century, the Quran codified existing practices of enslavement used in pre-Islamic Arabia and the Greco-Roman world; early Islamic law devised a legal framework that regulated slavery, manumission and set limits on who could be enslaved. In this, as Marozzi observes, Islam was ‘considerably more enlightened’ than its Christian, Jewish and Roman predecessors. One notable innovation concerned the status of enslaved women who bore their master’s children: these women, known as umm walad – ‘mother of the child’ – were not allowed to be sold and were automatically freed on their enslaver’s death. If their father acknowledged paternity, the children of an umm walad were recognised as free, and had the same inheritance rights as the father’s other children.

The new Muslim state expanded rapidly after Muhammad’s death in 632. Conquest brought a steady influx of captives. When Samarkand fell to Muslim forces in 712, the conquerors are said to have demanded ‘thirty thousand healthy slaves of fighting age’ as tribute. These Mamluks, military slaves drawn from the Turkic tribes of Central Asia, were prized for their skill in combat. Uthman, a caliph who ruled barely a decade after Muhammad’s death, is said to have had a thousand Mamluk soldiers in his army.

With territorial expansion came a growing demand for cheap workers. During the Abbasid Caliphate, which began in 750, attempts to restore the irrigated farmlands of southern Iraq led to a need for slave labour on an unprecedented scale. Many of the newly enslaved people were from eastern and south-eastern Africa, and were known collectively as the Zanj. Thousands of them were forced to work in terrible conditions in the salt marshes bordering the Tigris. The tenth-century historian al-Tabari wrote that they worked for long hours in the heat, draining the marshes and scraping off a nitrous layer so that their enslavers might cultivate the land. This was a fully developed system of agricultural slavery, dispelling any notion that Islamic slavery was a benign institution limited to the domestic sphere.

By the ninth century, the Abbasid army had come to depend on slave soldiers from various ethnic backgrounds. Political instability caused by recurring dynastic crises encouraged Abbasid caliphs to recruit Mamluk and Zanj soldiers en masse, binding the fate of the empire to its enslaved population. When the charismatic preacher Ali Ibn Muhammad appeared in the southern Iraqi city of Basra in 869, he found a ready audience among the Zanj. Proclaiming himself sent by God, he urged them to reject their servitude and turn against their enslavers. Within days he was leading the largest slave revolt in Muslim history. The Zanj Rebellion spread rapidly across southern Iraq, shaking the foundations of the caliphate. In 871 Zanj rebels finally captured Basra, burning it to the ground and killing thousands of residents. The social order was reversed: the city’s aristocrats were enslaved by those who had laboured in their fields. The Zanj even founded their own capital, al-Mukhtara (‘the Chosen’), and Ali Ibn Muhammad imagined himself a caliph. The revolt lasted for twelve years before the Abbasid general al-Muwaffaq finally subdued it in 883.

In the case of the Zanj, slavery was fundamentally tied to race. Marozzi catalogues the lengthy ‘roll-call of distinguished, even hugely venerated’ Arab and Persian writers who expressed bigotry towards Black Africans during the first centuries of Islam. Al-Jahiz, the famed Abbasid-era littérateur, described the Zanj as ‘the least intelligent of men’ (even though he was the author of a treatise on ‘the superiority of the Blacks over the Whites’). Ibn Khaldun, five centuries later, echoed the same sentiment. Such views weren’t unique to the Islamic world: Kant, Hume and Voltaire wrote similar things, to say nothing of earlier European scholars. For Marozzi, however, they attest to a persistent and characteristic ‘pattern of Arab racial supremacy’ towards Black Africans. Prejudice, he writes, is ‘deeply ingrained’ in Arab culture from the ninth century down to the present. Arab supremacy was certainly a feature of medieval Muslim literature, but Black Africans weren’t the only target. Ibn Butlan, an 11th-century Baghdadi Christian physician, wrote a treatise on the ‘purchase and management of slaves’, which encompasses many groups we might now consider ‘white’. Similarly, in The Excellence of the Arabs, written in the ninth century, Ibn Qutayba argued vigorously for Arab superiority over the Persians. Despite this, it is these writers’ attitudes towards the Zanj that Marozzi singles out as representative of attitudes he claims were ‘as instinctive to Arabs of this time as breathing’.

In the premodern Muslim world, not everyone who would be considered Black today was seen that way, and vice versa. Both al-Jahiz and Ibn Khaldun considered Sindhis and Indians to be Black; sometimes South-East Asians, such as Malaysians, were also included. Kristina Richardson and Rachel Schine, among other scholars, have shown how fluid these premodern taxonomies were, and how poorly they map onto those inherited from European colonialism. Consider Ali Ibn Muhammad, the leader of the Zanj Rebellion. He was born near Rayy, a village outside of Tehran. His grandmother, from Sindh in modern Pakistan, had been enslaved and was probably classed as Black by her Arab contemporaries. But she was not a Zanj and neither was her grandson. During the Zanj Rebellion, many of Ibn Muhammad’s followers were neither African nor enslaved. Some of the rebels were manumitted Zanj, others were Arabs and Persians, both free and enslaved.

Abdul Sheriff, a Tanzanian historian, estimates that only about 10 per cent of enslaved people in Muhammad’s Arabia were Black as we understand the term. In the following centuries, enslaved people came from an increasingly wide variety of racial and ethnic backgrounds. Nowhere was this more evident than in the Ottoman Empire, which began as a small principality in 1299 and came to rule over the Middle East and North Africa for nearly half a millennium. Greek and Circassian concubines, Balkan and Eastern European janissaries, sub-Saharan African eunuchs and labourers – all of these groups occupied different positions on the spectrum of unfreedom. To capture this, scholars have often resorted to the plural: Abdul Sheriff speaks of ‘slaveries’ in Muslim societies; M’hamed Oualdi of the Muslim ‘worlds’ in which slavery was practised.

In a few cases, enslavement opened the way to wealth and power. These social trajectories were also mediated by race, since they were reserved for the most fair-skinned. Roxelana, the daughter of an Orthodox priest from Ruthenia, in today’s Ukraine, was probably still a teenager when she was abducted during a Crimean slaving raid and taken to Istanbul around 1520. There, as her biographer Leslie Peirce recounts, she was purchased for the harem of the Ottoman sultan Suleyman the Lawgiver (known in the West as the Magnificent). By 1533 Roxelana had become the sultan’s ‘favourite’, a title created for her, and eventually became his wife. Istanbul’s conservative religious elites were dismayed by the sultan’s marriage to his enslaved concubine, but nothing could stand in Roxelana’s way. In the following decades, Hürrem Sultan, as she was called, plotted the assassination of grand viziers, commissioned the construction of mosques and advised her husband on foreign policy. She gave birth to six children, among them Suleyman’s successor, Selim II, who ruled until 1574.

Scores of Ottoman grand viziers started out as Mamluks captured in the Christian peripheries of the empire. Of the first 48 men to hold the post after 1453, only five were native-born Turks. Marozzi tells the story of Edhem Pasha, born a Greek Christian on Chios, captured as a child in 1822 during an Ottoman massacre on the island and purchased by Husrev Pasha, the Ottoman grand admiral (who later became grand vizier). Husrev sent Edhem to be educated in France, where he enrolled at the prestigious École des Mines. He then returned to Istanbul, where he rose swiftly through the ranks: governor, ambassador, interior minister and, in 1877, grand vizier of the empire. His son, Osman Hamdi Bey, would become the empire’s leading archaeologist and painter, training under Jean-Léon Gérôme, the master of French Orientalism. Yet Gérôme’s painting The Slave Market is a lurid portrayal of Muslim slavery; it’s a work that, as Marozzi notes, ‘pits Western moral supremacy against a cruel and depraved East’.

Slavery​ pervaded Ottoman elite society but the empire wasn’t a ‘slave state’ since enslavement ‘never underpinned any system of mass production’ as it did in the Americas. Marozzi nevertheless maintains that slavery inhered in the empire’s ‘very soul’. Captives and Companions confirms that Ottoman slavery did not unfold predominantly in viziers’ mansions and sultans’ harems. The vast majority of Mamluks and concubines did not make it to the pinnacle of Ottoman society; their experience of servitude was one of violence and privation. Over time, military slavery declined in the Ottoman Empire, but domestic slavery remained important and increasingly became the province of Africans. Brought from Wadai, Bornu, Bagirmi, Darfur and Kordofan, enslaved Africans were at the bottom of slavery’s hierarchy. From the 16th century until the early decades of the 20th century, they were forced to travel along perilous Saharan slave routes, crossing the desert on foot, sailing up the Nile to Egypt and across the Red Sea to the Arabian peninsula. Their numbers peaked in the 19th century, when roughly 1.2 million Africans were sold into slavery in Ottoman lands. An estimated 362,000 Africans were taken to Ottoman Egypt alone.

These men and women endured what the historian Orlando Patterson has called social death: the institutionalised negation of personhood through domination. Their voices scarcely appear in the archive, though scholars such as Toledano and Troutt Powell have done much to uncover what’s there. Most of them were women destined for a life of domestic service, working in kitchens and scrubbing floors. By the 19th century, Ottoman slavery was ‘predominantly female, African and domestic’. Some of the most harrowing accounts come from the empire’s final decades. Around 1890, a couple and their young son, Ahmet, were abducted in Kenya and taken to Crete, where they were sold to a local family. While his parents – his father and stepmother – worked on the family farm, Ahmet frequently attempted to escape; his enslavers punished him by locking him in an iron cage. When his father died, the boy, still a teenager, was forced to marry his stepmother, Nuriye, so that the household would still have a male head. Years later, Nuriye was sold to new owners in Istanbul. Ahmet was then compelled to marry again, this time to one of Nuriye’s sisters. The harm and degradation endured by Ahmet, Nuriye and their children stands in for the experiences of many others, mostly unrecorded.

This is to say nothing of the physical violence they and countless other enslaved Africans suffered. Experiences of enslavement in Islamic Africa bore most resemblance to the Atlantic system. During the 19th century, as Paul Lovejoy and Frederick Cooper have shown, a plantation economy sustained by slave labour developed in the West African Sokoto Caliphate and on the coast near Zanzibar, though slavery was not linked to any single agricultural activity and the crops were mostly sold in neighbouring regions rather than being distributed to global markets. For centuries, as Marozzi writes, enslaved Africans took on an ‘enormous variety of roles’ across the continent: mining salt in Niger, harvesting gum arabic in Mauritania, climbing date palms in Saharan oases and fighting in the Moroccan king’s army. Their masters might be Arab, Black or both. They were merchants, landowners or warlords such as Rabih al-Zubayr, the notorious Sudanese slaver who branded his slaves. He died in 1900 fighting the French colonial army.

Marozzi recognises that ‘Islam did not conceive slavery in the Middle East.’ Yet he finds it ‘tempting’ to attribute slavery’s ‘unbroken’ history in the Muslim world to ‘the sheer cultural power’ of Islamic law. Describing the enslavement of Yazidis by Isis militants in the mid-2010s as ‘the revival of an Islamic institution’, he implies that slavery is somehow a Muslim practice, whether in ninth-century Basra or 21st-century Syria. Appealing to scripture in order to lend religious legitimacy to slavery isn’t unique to Islam: the Bible was frequently used to justify Atlantic slavery. But it would be surprising if a modern historian used the Gospels to explain the Atlantic slave trade, or referred to plantations in the American South as a ‘Christian’ institution with an ‘unbroken’ legacy dating back to late antiquity. The Muhammad-to-Isis approach to the study of the Muslim world is always unhelpful, whatever the focus.

It is disappointing that Marozzi fails to engage with untranslated Arabic sources, and I sometimes found his tone jarring. At the start of his chapter on the Saharan slave trade, he recalls riding his ‘favourite camel’ to the Libyan oasis town of Murzuq, along with two other British travellers: ‘morale high, beards full and waistlines diminished’. The scene could have come from Seven Pillars of Wisdom. In fact, Marozzi admits that his venture into the Libyan desert was inspired by the early 19th-century travelogue of a British naval officer called George Francis Lyon, whose account of ‘intrepid explorers’ and ‘treacherous sheikhs’ fascinated him as a teenager. Soon, he felt the ‘irresistible pull of the desert’. Elsewhere, he writes of the ‘quintessentially Arab obsession with bloodlines’ and finds ‘unsurprisingly’ that the Quran’s version of paradise is not a ‘nirvana of gender equality’. These passages distract from his more nuanced arguments.

Also distracting is his deployment of history for political ends. In an opinion piece in the Telegraph last November, Marozzi drew a contrast between conversations about reparations for slavery in Britain and the US, and the absence of ‘public discussion in Arab countries, Turkey or Iran, of the historical practice of slavery, let alone the issue of reparations’. When he lectured on Islamic slavery at the Pharos Foundation in Oxford, he went on, one student accused him of Islamophobia. These arguments echo Bernard Lewis’s idea of a ‘white man’s burden’ of guilt.

That slavery persisted in Muslim societies well into the 20th century is undeniable. More questions need to be asked about its distinctive features, the extent of its connection with the Atlantic trade and the forms of racialisation that underpinned it. Finding answers will require painstaking work in archives across sub-Saharan Africa and the Muslim world, and a command of non-European languages. Equally pressing is the need to confront the memory of enslavement and its connections to anti-Black racism in Muslim countries today, especially in North Africa, where the old slave routes now serve as migrant pathways into Europe via Morocco, Algeria, Tunisia and Libya. But even in these countries a reckoning is underway, often in the face of resistance on the part of local governments. More broadly, there is a growing body of academic works, popular books, NGO reports, novels and movies on the subject, many of them written and produced in Muslim societies. A generation ago the study of slavery in Islam was shrouded in a ‘deafening silence’, as Toledano once put it. This is no longer the case.