Slavery in the Ottoman Empire

Abstract

This chapter analyzes Ottoman slavery as an institution in order to provide readers with insights into how Ottoman society both benefited from slavery and became largely defined by it. The topic will be examined by illuminating the aspects of entry into slavery, extraction of slave labor, and exits from slavery, arguing that these three phases were cyclical rather than gradually evolutionist.

Introduction

The Ottoman Empire was initially a small state formed as a warrior polity in Northwestern Anatolia in the late thirteenth century. During its expansion phase, slavery and slave raiding greatly contributed to the enrichment of the military, state officials, mercenaries, and some merchants and private owners. Beginning in the mid-fourteenth century, however, slave raiding gradually became a rather marginal practice with the establishment of a sophisticated bureaucracy that extracted fiscal revenues and taxes from a vast array of territories, activities, institutions, and actors.

This chapter explores the phenomenon of slavery in the early modern Ottoman Empire through the phases of entry, labor extraction, and exit with a view to the slave–master relationship, drawing on Marcel van der Linden’s tripartite analytical scope.¹ Albeit useful as a framework, van der Linden’s take on this particular form of coercive labor relations does not suffice to fully explain the historical dynamics and rationale at stake in the Ottoman society. The Ottomans of the early modern era had a “populationist” motivation for maintaining a yearly influx of thousands of slaves into their territories despite not maintaining production systems that relied exclusively on servile labor. The empire therefore never constituted what Moses Finley has called a “slave society”—one in which production and the social order depended in large parts on slavery, and where enslaved people made up a substantial portion of society. The demographic proportion of the slave population in the Ottoman Empire was roughly between one and ten percent in major cities and towns, with the more or less constant trafficking in slaves somewhat offset by the regular practice of manumission. This particularity is best epitomized by a fifteenth-century chronicler: “Every year more or less fifty thousand male and female infidels are taken from the abode of war as captives; those become Muslim, and their progeny join the rank of the faithful until the day of resurrection.”² The forced migration of scores of “infidels” each year (by way of conquest, abduction, slave raiding, and human trafficking, but also often through sale by their own families) therefore cannot be explained solely with the exigencies of economic growth, but more so with the ruling elite’s desire to forcibly recruit future Ottoman subjects, loyal taxpayers, and new Muslim believers—thus essentially employing slaving as a strategy of bolstering the ranks of the empire and its dominant religion.

Entries into Slavery

Who could become a slave in the Ottoman Empire, and how? Firstly, it is important to note that non-Muslim foreigners were most commonly (and legally) considered enslaveable. Simply being a non-Muslim and non-Ottoman subject was not a uniform category, however; citizens or subjects of a state that had signed a peace treaty with the Porte (such as the Republic of Venice outside the main periods of war against the Ottomans, or the Kingdom of France from the 1530s onwards), for example, were considered off limits and under protection, at least in legal and diplomatic terms. By contrast, people hailing from a polity at war with the Ottoman Empire were deemed fit for enslavement. Non-Muslims could therefore become enslaved through direct capture in warfare—but they could also be born into slavery, acquired via trade networks (implicating merchants and actors outside the Ottoman realms), or enslaved illegally.

Capture in the course of war best illustrates the ways in which Islamic law shaped the institution of slavery over many centuries. Legal slavery was entirely exogenous by principle.³ Compared to putting prisoners of war to death, slavery was presented as a “lesser evil” that nevertheless constituted a way of imposing the victor’s law upon the vanquished (to paraphrase Carl von Clausewitz). The age, gender, beauty, talent, and social status of captives helped their captors decide whether to enslave, ransom, or kill them. As a structuring element of most societies in the Mediterranean, the implementation of slavery was based on juridical, political, and religious criteria along with military circumstances and simple misfortune.⁴

A discrepancy existed between the legal tenets of the “abode of Islam” (dâr al-islâm) and the “abode of war” (dâr al-harb), however. In theory, the populations of countries having concluded truce or peace treaties could not be enslaved—in other words, enslavement could not be applied to people who had already recognized the political supremacy of an Islamic state or a Muslim sovereign. This often amounted to little more than a legal fiction, however—and one that did not correspond to the Realpolitik of Ottoman imperial expansion. In practice, military strategy and material gain were the determining factors for slave raids on warring parties, even after a peace treaty had been concluded.⁵

The empire’s expansion brought diverse peoples into slavery directly at the hands of Ottoman military forces (officials, conscripted subjects, mercenaries, and the like)—especially from many of the territories conquered between the fourteenth and seventeenth centuries: Greeks, Bulgarians, Moldavians, Serbs, Albanians, Georgians, Hungarians, Germans, Spanish, and Italians, among others. Independently of this imperial expansion, however, corsairs also raided for slaves along the maritime borders of the Ottoman Empire and beyond. Marauders from the autonomous Maghreb provinces (Algiers, Tunis, Tripoli), for example, ravaged northwestern Mediterranean shores, sometimes venturing well beyond the Ottoman sphere of influence and raiding as far north as Iceland during the seventeenth century.⁶ Engaging Christian military orders like the Knights Hospitaller of St. John in Malta or the Order of San Stefano in Tuscany—who in turn took scores of Muslims from the Maghreb as well as Ottoman subjects of all religions from the eastern Mediterranean between the sixteenth and eighteenth centuries—corsairs participated in a trans-Mediterranean network of captive ransoming and slave trading that impacted coastal communities throughout the region.⁷

The ubiquity of the phenomenon dissipates any illusion that enslavement and slave raiding were purely motivated by ideological and religious convictions. The pirates, corsairs, mercenaries, and their associates conducting these regular but relatively sudden and quick operations were often Muslim converts turned marauders, so-called “renegades” who subsequently set out to raid their former fellow Christians. Oranian corsairs originally hailing from Corsica or Calabria, for example, were newly converted Muslims who returned to Christian communities as their aggressors. Explanations for this behavior should be sought not in the sincerity of religious conversion but rather in the potential material gains to be achieved in such a frontier society, where the Catalan, Ligurian, Sardinian, and Provençal coasts were easily looted.⁸

Since slave raiding was a common maritime activity in the Mediterranean, and considering the fact that it was virtually impossible to control and patrol the official borders of the empire, commoners in coastal villages constituted easy targets for pirates and corsairs—even when the capturer and the captured were both Ottoman subjects. From the sixteenth century onwards, many sultans and their imperial governments tried to intervene to prevent the illegal enslavement of their own subjects, for example in the Aegean Sea. A practical problem arose here, however: By the time a collective petition reached the Porte, it had already travelled for at least a couple of weeks. In the meantime, even though officials could ascertain the circumstances and detect the profiles and activities of the enslavers, the victims of illegal enslavement were dozens or even hundreds of kilometers away.⁹ Even worse, the act of sale—officially recognized by way of a fiscal stamp—sealed the legal slave status, notwithstanding the originally illegal enslavement. People hailing from distant towns and provinces were able to prove their origins and thus the illegality of their enslavement in numerous “freedom suits” throughout the early modern era thanks to their knowledge of the workings of Ottoman justice, the furnishing of witnesses who testified in their favor, and their own perseverance.¹⁰

An interesting case in point that corresponds to both entry and exit phases at once is the Ottoman conquest of Venetian Cyprus in 1570–1571 during the war between the two realms in 1570–1573. Establishing their own rule on the island in 1487, Venetian settlers and administrators had played an important role in maintaining and tightening the servitude of Greek Orthodox peasants at the hands of Latin masters in a form combining aspects of both chattel slavery and feudal serfdom. The military success of the Ottoman troops and navy in 1570–1571 led to several vicious massacres of Venetians as well as to the enslavement of the men and women who were spared. The Greek Orthodox peasants who had formerly been subdued by their Catholic masters not only became free tax-paying Ottoman subjects (re ‘âyâ), but some of them even went on to own Venetian slaves in a spectacular reversal of this specific form of asymmetrical domination, within the Ottoman imperial order.

A tributary state and key military ally of the Ottomans, the Tatar Khanate of Crimea, played a major role in the constant supply of slaves to the Ottoman Empire from its annexation into the “well-protected domains” around 1475 until 1699, when the Treaty of Karlowitz was signed with the anti-Ottoman European coalition following the severest Ottoman defeat in Europe until then. This regular source and network of supply developed incrementally and at times exponentially, especially with the slowing of the Ottoman military conquest machine from the mid-sixteenth century onwards. Besides the formal conquests (in which Crimean troops also participated in), regular raiding operations into Hungary, Poland-Lithuania, Muscovy, the Caucasus kingdoms, and sometimes even into friendly realms (such as other tributary states, especially the Romanian principalities of Transylvania, Moldavia, and Wallachia) were organized annually. The latter represented a problem for the Porte since its demand for yearly tribute from these territories was in exchange for autonomy in internal affairs and the protection of the physical and juridical integrity of the respective principalities’ subjects.¹¹ The Black Sea slave trade supplying entire Mediterranean networks from the macrocephalic Istanbul and various ports on the southern shores of the Black Sea had ancient roots, of course. Essentially, the power and volume of the lucrative slave trade in the Ottoman Empire were owed to the fact that the same polity—having expulsed medieval Italian colonies from Caffa—reunited all these old routes, markets, and demanding urban centers of the intercontinental slave trade in the same way the Roman, Abbasid, or Byzantine Empires (to which the Ottoman sultans considered themselves the legitimate successors) once had.

Some comments regarding the juridical tenets of entry into an Ottoman’s possession and service as a slave are in order at this point. The religion of the potential buyer/owner as well as that of the captured/enslaved were important in this context. Simply put, Muslims had the right to own slaves of both Muslim and non-Muslim obedience, whereas protected, free, non-Muslim taxpayers of the empire, the zimmîs, officially only had the right to possess other non-Muslims as slaves—they could not own Muslim slaves, not even as percentage holders. To be clear, although a Muslim as such could not be legally enslaved by other Muslims, a legally enslaved person serving in the Ottoman realms could convert to Islam while remaining in their owner’s possession (and in fact many did). Studies on executive orders from sixteenth- and seventeenth-century Istanbul and its Jewish communities show that not only did many non-Muslim Ottomans possess Muslim slaves, but some masters even made their slaves convert to their own religion, with frustrated local and imperial authorities unable to prevent this phenomenon despite the sovereign’s displeasure. Community leaders were however called upon to compile lists of slaveholders with the names and religious denominations of their slaves. When an Ottoman official was able to efficiently intervene in such a case, he was to effect the sale of the newly Islamized slave to a Muslim master.¹²

When female slaves owned by Ottomans gave birth to a child that was not recognized by the male owner, the offspring would automatically have servile status. In addition, men were not legally allowed to have sexual relations with female slaves owned by their wives; fatwa compilations contain a plethora of rulings stating that even the wife’s explicit authorization would not render such intercourse legal. This meant that a husband had to buy his wife’s female slave in such a case so as to be formally irreproachable from a legal standpoint. If a child born to an enslaved woman was recognized by the male slave owner as his own, however, the newborn was considered free and its mother transitioned from enslavement to being an ümm-i veled (“mother of the child,” a status discussed in more detail in the section on exits from slavery below). From a broader perspective, while slave children were certainly born in different places and situations within the sultan’s realms, slavery by birth was far from being a common form of obtaining new slaves. It was not an internal reproduction system that sustained the Ottoman Empire’s servile population but rather a combination of war plunder, slave raiding at the frontiers, and of course the massive imports via the Crimean peninsula—augmented to a modest degree from the Caucasus as well as Sahelian and sub-Saharan Africa, and to a much lesser extent via the Indian Ocean and Central Asia.

Entry into private slaveholding was not possible through penal condemnation or debt bondage. This does not mean, however, that penal condemnation and unpaid debts could not lead to forced labor or coercive methods of extracting work and other benefits from affected individuals. Simply put, different legal statuses existed within the same professions and workplaces. All slavery was not forced labor, and all forced labor was not slavery. What is legally certain in this context is that insolvent persons could not be commodified.

Before delving into phenomena of extraction, the exception of the devşirme should be mentioned. Under the Pax Ottomanica, zimmîs were normally exempt from all attempts at enslavement within the Ottoman society. And yet, a peculiar sultanic custom that presumably began toward the end of the fourteenth century consisted of organizing “levies” within the Christian communities of the Balkans (which would later be implemented in Anatolia as well), coercing teenaged boys into compulsory service to the sovereign. They were enslaved, forcefully converted to Islam, and circumcised before undergoing various intellectual, military, physical, artisanal, and agricultural formation phases to eventually become a member of the prestigious and elite Janissary corps. As personal slaves (kuls) of the sultan, the most brilliant and beautiful of them could also be employed at the imperial palace and eventually become imperial officials or advisors, with the highest attainable position being that of the grand vizier, second only to the sultan himself. The particularity of being the sultan’s slave also suspended existing shariatic norms:

• A kul of the sultan could possess slaves of their own, whereas this was not possible for ordinary kuls of private slave owners.

• While the sultan had the arbitrary power of life and death over his kuls, he was not allowed to have his free subjects executed without a judicial decision. Nor were ordinary slave owners legally permitted to kill their own slaves.

Extraction Possibilities (Labor and Other)

There were possibilities of extraction and exploitation in many areas: lucrative, labor, services, prestige and domination, sexual favors, and the production of heirs. The use of slaves in the Ottoman Empire thus served many other purposes beyond pure labor extraction calculated according to a rationale of optimized and profitable work through intensified production methods. This was especially true for the richest slave owners, since they did not depend on rationalized and profitable labor extraction producing financial revenue through the exploitation of slaves over a calculated period of time before having to let them go either by sale or manumission. In other words, they had the means to lose some money by pursuing different objectives or gains with their slave ownership. We also find that slaves were present in virtually every sector, profession, activity, and function. This implies several things:

• No profession or form of production was reserved exclusively for slaves. Even sexual labor that could be imposed on female slaves in probably the direst circumstances was not an exclusive domain of servitude, with the quasi-merchandising of brides through dowry and ordinary prostitution being the main counterexamples, which could simply be other forms of human trafficking. They were not always necessarily a trade in humans—as opposed to slavery, which always was.

• Workers of different legal status with similar or identical tasks coexisted in the same workshops, fields, bazars, shops, households, and so on.

• A clear dichotomy between free and unfree labor is too abstract and thus meaningless in the context of the Ottoman Empire.

• While present in many if not all work sites, slaves usually did not represent the majority.

The rationale for slave ownership that did not primarily follow economic goals consisted of integrating enslaved persons into a family, political faction, or group of trustees in a private and/or semi-public sphere. Beyond this motivation, broader considerations of growing the political, fiscal, and religious communities along with the logic of depriving the enemy of “human resources” played a role as well, in line with the “populationist” argument cited above.

Enumerating the key sectors of activity in which slaves could be found alongside workers with other statuses (wage earners, apprentices, Janissary trainees, convicts, ransomable captives, volunteers, requisitioned re ‘âyâ, conscripts, corporation members, and so on), we can cite pre-industrial textile workshops, agriculture, sexual, domestic, and menial labor, bureaucracy, intellectual activities, commerce, arsenals, mining, public works and the construction sector, and galleys.

In the mining sector, similar tendencies can be discerned in the rationale for workforce deployment as in the well-documented eastern Mediterranean sugar production from the fourteenth century onwards, with high mortality rates due to plague epidemics on a regular basis.¹³ The driving factor in this and other extraction and manufacturing sectors was certainly the need of the state and its business partners to maintain a constant flow of key raw materials and lucrative goods. The Ottoman state based its approach to mining on the economic value and strategic importance of ores, implementing fiscal stimulus policies to attract voluntary and free tax-paying subjects of the sultan in particular. Despite being less noble than silver, copper was a crucial metal for the Ottoman state, which established a monopoly on its extraction—most notably for the fabrication of artillery and other military equipment. Furthermore, exportation of and trade in copper was prohibited throughout the empire, a proscription enforced in particular on the ores from Küre in the Kastamonu province in northern Anatolia (on the southern Black Sea shore), which had been active since at least Byzantine and Seljukid times.

Although the state managed to attract free and salaried miners to Küre from the fifteenth century onwards by offering fiscal exemption, corrupt local officials frequently prevented the systematic application of these benefits to the miners. In addition, the state’s demand for the constant output was regularly interrupted by numerous instances of earthquakes, floods, and epidemics (notably the Black Death). The natural catastrophes and poor hygiene that decimated the Küre miners caused the Ottoman state to rely on slave labor as a last recourse. Archival sources show that the imperial bureaucracy treated the issue as a purely budgetary one, but also that the state was not the sole owner of slaves extracting ore in Küre: Private investors and contractors were likewise involved. State archives (internal correspondence, account books, fiscal records) and contemporary chronicles (such as those of Evliyâ Çelebi and Kâtib Çelebi in the seventeenth century) provide us with specific and valuable data on the population of Küre that was involved in mining, as well as on the bureaucratic and commercial world revolving around the mining town.

A document from the year 1682 indicates a devastating plague outbreak in the mines of Kastamonu that left almost no survivors among the slaves. Local officials urgently called for the recruitment of new laborers in order to resume mining (BOA, İbnülemin, Me ‘adin 2/105, dated 18 July 1682). As compensation, 400,000 akçes—roughly equivalent to 1333 Venetian ducats at the time—were paid out to the mines by the central government. To provide a sense of the scale of this sum, we can consider that a luxury slave cost approximately 5000 akçes in the sixteenth to seventeenth centuries, while cheaper menial slaves were valued at around 200 akçes. The government handout thus allowed the purchase of roughly 2000 slaves at the lowest price. Simultaneously, the document also evidences the direct involvement of the central government in the assignment of the workforce. Another record from the same mines in 1690 confirms this modus operandi in terms of recruitment, with the allocation of 9000 Dutch lion dollars for the purchase of new slaves who were presumed to be worked to their deaths in the non-stop extraction business (BOA, İbnülemin, Maliye 30/2947, dated 9 June 1690). This sum would have allowed 180–240 slaves to be bought at average period prices.

Mining work was arguably the most difficult task with the highest mortality rate to be performed by slaves in the Ottoman Empire. Other instances of slave use as a business model were radically different: Boutique owners and merchants dealing in goods of international commerce could own slaves to whom they taught the tricks of the trade, using them as reliable and mobile agents “authorized” (mezûn) to complete various high-stakes transactions on their behalf. Some mezûn slaves even traveled abroad for financial operations respectively for import–export trading using the capital they were entrusted with, returning loyally to their patrons despite having the opportunity to escape with considerable amounts of money and/or merchandise. Some did in fact abscond, which in turn could cause Ottoman authorities to solicit their Venetian, Polish, or Muscovite counterparts for assistance, sometimes to no avail.¹⁴ This “authorization” can be viewed as a pure delegation of authority, with the possibility of the mezûn overseeing other slaves of the same owner. As agents of their masters, mezûn slaves determined the outcome of their status by constantly reassuring their owner with their competence and trustworthiness. By doing so, they could eventually succeed in exiting the slave status altogether by being enfranchised—though the nature of the labor relationship between an issuer of orders and proprietor of means on the one hand and the competent but subordinate secretary or valet on the other meant they could never truly become equal business partners despite their achievements. We might say there was a continuum in the labor relation, and that enslavement was an important element of this relation—but nevertheless only an element, not a defining factor for the overall dynamics of such business setups.

Appellations like “domestic” slaves, or slaves tout court, by way of different synonyms in Ottoman Turkish (kul, köle, câriye, etc.) do not precisely indicate their role and function in a private setting. This has led many scholars of the Ottoman world to speak about them as performing “various tasks, menial tasks” as unspecialized workers subject to their masters’ and mistresses’ will, doing anything that was demanded of them. It is true that most documents do not clearly state the nature of the daily work of domestic slaves. In court records, however, we sometimes find indications of individuals working as washers, cooks, gardeners, blacksmiths, wetnurses, or the like, rather than generic and vague phrasings like “Ahmed’s Bosnian slave” or “my tall female slave with golden hair.” I would argue that these men and women performing everyday household work were anything but “unskilled”: On the contrary, they were constantly under pressure of having to multitask, to remain flexible and focused, and most of all to specialize in essential activities whose value was underappreciated. The same mentality is still demanded of cleaning staff, caretakers, merchandise handlers, warehouse workers, and other professions in our modern Western societies today. In other words, domestic slaves in the Ottoman Empire were skilled in ways that were not recognized institutionally or socially by guilds or employers—but this naturally does not mean they lacked skills altogether. On a slightly provocative note, one might even suggest that when historians emphasize the lack of skill of these slaves, this may translate as an inconscient bourgeois disdaining menial work while feeling contempt or pity for those who have nothing but their manual labor to make a living.

Although slavery as practiced by the Ottomans was not an institution exclusively related to labor, some form of benefit or gain was always a consideration—be it in terms of prestige, social standing, sexual gratification, or the satisfaction of ruling over someone (albeit with limitations). Access to slaves depended on the flows of the human trafficking networks and the geographic location of potential buyers as well as on market availability, the budget of the buyer, and the profit margin of the seller. An idealized form of ethnic labor division can be found in physiognomy and ethics treatises, which certainly had an effect on representations and clichés without necessarily having a day-to-day implementation. For instance, whereas sixteenth-century chronicler, bureaucrat, and gentleman Mustafa Ali of Gallipoli suggested Latin Western Europeans or Hungarians as the best and most intelligent personal servitors, Ottomans in need of personal servants generally had to purchase whatever ethnicity was available and affordable at their local market at the time.

Sociological and economic criteria were paramount in the context of slave ownership in the Ottoman Empire. The wealth of a master or mistress determined the use of their slaves and the duration of this use, which in turn defined the possibilities and ways of exiting and continuing the work relation under different circumstances. Was a prospective buyer looking for a new family member, a business partner, or a permanent servant? It all depended on their social standing, along with their available means and the purposes they envisioned for their future human property.

Exits

Exit from slavery cannot be likened to the termination of any other labor relation, since the very ways in which an exit was (im)possible were determined by enslavement itself, whereas the slaveholding per se did not necessarily entirely define the extraction phase, which merely translated the means and objectives of employers in general.

Combining canonical Sunni jurisprudence with various customary elements, Ottoman law, and social practice offered several ways of exiting slavery. Unconditional, direct, and absolute manumission was always a possibility, although not widely practiced. For religious, moral, or any other reasons, owners could set a slave free by way of a simple formula declaring the liberation of the individual, who in turn could claim his or her new status. A written attestation or credible witnesses represented further guarantees for the enfranchised.

Another possible form was conditional manumission (tedbîr): In such cases, the owner defined a more or less predictable—and at times quite random—condition for their slave to attain freedom. This condition could be a specific occurrence: “if I survive this illness,” “if I do not return for more than 60 days after my departure,” “if I come back alive from my hajj to Mecca” or similar. However, stipulations were also often unsystematic and aleatory. Technically, testamentary manumission falls into this category as well, since the main condition for enfranchisement was the master’s death. The heirs had the possibility of formally annulling a testamentary manumission if the slave’s value was superior to a third of the whole inheritance.

Manumission contracts are well documented in the court registers of virtually any given Ottoman judicial district (kazâ). They correspond to the third and arguably primary legal manner of exit from slavery, which once again reveals the key logic of slaveholding. In general, such contracts were conditioned on duration of service, on the payment of a lump sum, or at times on the requirement that the slave produce a certain quantity of goods before being set free. In some cases, a manumission contract (mükâtebe) represented a combination of two or all three of these aspects. These written contracts allow us to determine the average duration of a slave’s service in a private household as 8–12 years, with extremes in the range of 4–20 years. Lifetime service was never excluded, of course, and could result from a slave’s failure to respect the terms of their contract. Beyond the average service duration, manumission contracts also enable us to estimate the profit that slave owners expected to make. In this context, we must consider the estimated purchase price of the individual slave, the nature and duration of extraction, and of course, the sum of money the potentially manumitted was required to pay for their freedom. Incidentally, we can conclude from the mükâtebe that while indebtedness did not lead to slavery in the Ottoman Empire, enslavement could in turn easily lead to debt for those wishing to exit the status via contractual manumission. It was considered morally and conventionally inappropriate to manumit individuals deemed too old to provide for themselves. Ethical ideals suggested that owners assume the responsibility for material upkeep until their aged slaves’ death.

The exit phase could also be a form of extraction in itself. Walâ’, the Arabic term indicating the patron-client relationship, was a secular institution originating in pre-Islamic times, specifically in Roman law and in part in customs of the Arabian Peninsula. Once the former slave master became a patron and the former slave a free person (and client) by way of manumission, heirs on both sides could potentially perpetuate this relation. If there were no heirs on the client’s side, the patron’s family would inherit everything the client or their childless descendants had possessed. Patronage also entailed various obligations and duties, along with a sense of fidelity which often made it so that the work relation did not change significantly beyond its juridical nature. The agnatization of clients and the patrons’ right of preemption regarding their inheritance means that the value and meaning of the widespread and canonical formula used in manumission declarations before the kadi’s court can be seriously questioned. The legal formula translating as “s/he shall be free from now on like every other person of free birth (hurr al-asl)” was apparently an optimistic portrayal and more of a legal fiction than a truthful rendering of the reality of the situation.

The power of a patron over their slave-turned-client naturally depended on the social standing along with the means and objectives in the new phase of the relationship. This means that not all theoretically established walâ’ frameworks necessarily led to the perpetuation of patronage over many generations—and thus that not all manumissions can be considered transitions to a new form of extraction of privileges or as a further step in a never-ending power relation.

The primary counterexamples of this type of all-dominating power of the patron over the client were instances where clients or their descendants made use of their agnatic affiliation to the patron’s family, above all in the client’s own interest as an underdog. In a groundbreaking study on Damascene society in Ottoman times, Karl K. Barbir examined biographical dictionaries of the sixteenth and seventeenth centuries. Darwîsh Muhammad (d. 1605/06) was the son of a holder of timâr (fiscal revenue for a caste of privileged military staff) in the Great Syria region and a manumitted female slave who had been owned by the notable Damascene family of the Ibn Tâlû prior to her enfranchisement and marriage. Darwîsh grew up in poverty owing to his father’s fall from grace and flight from Damascus. Despite this adversity, he managed to become a renowned member of the ulema class of legal-religious scholars. His fame extended from Damascus to Istanbul, and he was known under the patronym Tâlûzâde—derived not from his father’s name but instead from that of his mother’s former owner. In the Damascene society of the time, an Ottoman bureaucrat or soldier and thereby member of the fiscally exempted class of state officials (‘askerî) could thus obtain a new patronym by marrying a free, enslaved, or manumitted woman. This could also provide an official hailing from another province with a local affiliation. Darwîsh’s mother had never been his father’s slave; the patron–client relations clearly indicate that she had been owned by the Ibn Tâlû family, who manumitted her and allowed her to marry Darwîsh’s timariot father. This matrimonial contract allowed the Ibn Tâlû family to establish a new network within the Ottoman public service via a public servant who had recently arrived in the Damascus district. As the disgrace of his military father meant the end of his salary and his fiscal and real estate benefits, however, Darwîsh—now left in poverty—had to define a new “fictitious kinship” made possible by the slave past of his mother. He associated quickly with his mother’s proprietors-turned-patrons. Instead of being a burden, his mother’s past enslavement thus became a source of prestige for a man who had lost all social and symbolic capital he could have had, thanks to his father. He undertook considerable efforts to study in madrasas and become a member of the ulema in his own right, but it was above all the juridical, social, and perhaps also the economic heritage of his enfranchised mother that contributed to his success. These circumstances caused Darwîsh to choose a matrilinear filiation which, while fictitious, was juridically and socially solid. The new kinship between Darwîsh and the Ibn Tâlû created obligations for the latter in favor of the former. In this case, too, the economic and political power of slave owners influenced the possibilities open to the manumitted and their descendants over several generations.¹⁵

As an exit strategy, conversion to Islam or whichever was the owner’s religion could work in some cases, but such intentions were mostly recognized as insufficiently sincere—and besides, the agendas and calendars of the respective parties usually did not match. As mentioned above with regard to the criteria for becoming a slaveholder, a non-Muslim slave belonging to a non-Muslim owner could only hope to be manumitted upon converting to Islam. This outcome was anything but certain, however; in short, the possibilities were:

• Being sold to a Muslim and continuing to be a slave for some time.

• Being illegally detained by the same owner.

• Being enfranchised (in the knowledge that this was legally problematic since it established a patron–client relationship between a non-Muslim and a Muslim with the former being the superior party in the relation, which was deemed undesirable by the norms of the ruling class of the Ottoman Empire).

A brief mention should also be made of the umm al-walad, Arabic for “mother of the child.” When a female slave belonging to a male owner became pregnant and the owner recognized his paternity, the offspring would automatically be considered free at birth. While remaining a slave, the mother would obtain the special ümm-i veled (Turkish pronunciation) status, which rendered her inalienable. An ümm-i veled could no longer be sold or transmitted through legacy. If the owner did not choose to manumit her while he was still alive, she automatically became free upon her owner’s death. Many slaveholders nevertheless attempted to sell or bequeath an ümm-i veled, and there are numerous court records of litigations where such women claimed their rights, sometimes successfully.

Although the master’s social standing, power, economic means, and intentions for his or her slave(s) were key factors, this did not mean everything was inevitably determined by the slaveholder’s will—as stipulated in the legal norms demanding total obedience of slaves to their owners. The simple reason—and the final point of this contribution—is that there were also illegal exits from slavery. Such illegal exits by way of abscondence occurred by the thousands in the Ottoman Empire. Mentions of fugitive slaves are abundant in the court registers—more precisely, it is captured fugitives who were most frequently registered, compared to the far less numerous cases where owners unilaterally seized the kadi’s court to declare their human property missing. Some fugitives committed theft (material resources) and/or murder (as revenge and/or in order to make good their escape) before their flight. The chances of success of such getaways were scarce but not nonexistent. Yavacıs, a corps of semi-professional and prize-winning fugitive hunters (who also searched for lost livestock¹⁶) working closely with judicial authorities in many districts captured men and women suspected of being escaped slaves in the peripheries of towns, on major travel routes, and the like. Many of those captured in this way claimed to be free individuals or manumitted former slaves. Some managed to prove their assertion through documents and witnesses, while those who could not were detained by the kadi’s court for a customary period of three months or a hundred days. During this period, which was considered sufficient to allow owners to claim their absconded slaves, the court would provide food, shelter, and clothing for the captives. A retrieving owner could take their human merchandise back upon verification and after reimbursing the court for the expenses incurred for the fugitive’s material upkeep. If the owner did not show up by the end of the customary period, the court sold the respective detainee at auction, thereby recouping its expenses and perhaps realizing a small margin of profit (knowing that the price would be below market, since a fugitive was a “faulty” and thus less desirable slave). Although successful escapes did occur, the explicitly recorded cases are almost exclusively those in which absconders were recaptured and returned to slavery.¹⁷ To this day, we know of no tangible networks of solidarity between fugitives or autonomous escapee communities in remote areas comparable to the spectacular and inspiring examples in the Caribbean, for example in Jamaica. This takes us back to the very beginning: Illegal exits by abscondence were still exits, though not always as efficient as the legal ones like manumission. Since the method was perilous for the escapees (and their recapture lucrative for fugitive hunters), these attempts could take slaves hoping to find their freedom back to square one—and “illegitimate” victims of the hunters could be illegally enslaved despite having been free Ottoman subjects by birth or through manumission. Therefore, an illegal exit could become a re-entry, and a presumed illicit exit an entirely new entry into slavery—and an unlawful one at that.

Notes

• 1. Marcel van der Linden, “Dissecting Coerced Labor,” in On Coerced Labor: Work and Compulsion after Chattel Slavery, ed. Marcel van der Linden and Magaly Rodriguez Garcia (Leiden: Brill, 2016), 293–322.

• 2. Muhammed b. Mahmûd-ı Şirvânî, Tuhfe-i Murâdî [1429], ed. M. Argunşah (Ankara: Türk Dil Kurumu, 1999), 73. Excerpt cited and translated into English by Cemal Kafadar, “A Rome of One’s Own: Reflections on Cultural Geography and Identity in the Lands of Rum,” Muqarnas XXIV (2007): 13–14.

• 3. Madeline C. Zilfi, Women and Slavery in the Late Ottoman Empire: The Design of Difference (New York: Cambridge University Press, 2010), 211.

• 4. Colin Imber, “Fiqh for Beginners: An Anatolian Text on Jihād [2000],” in Warfare, Law and Pseudo-History (Istanbul: The Isis Press, 2011), 65–66; Christine E. Sears, “‘In Algiers, the City of Bondage’: Urban Slavery in Comparative Context,” in New Directions in Slavery Studies: Commodification, Community, and Comparison, ed. Christine E. Sears and Jeff Forret (Baton Rouge: Louisiana State University Press, 2015), 201. See also Rainer Oßwald, Das islamische Sklavenrecht (Würzburg: Ergon Verlag, 2017).

• 5. Majid Khadduri, War and Peace in the Law of Islam (Clark, NJ: The Lawbook Exchange, 2006 [1955]), 130–32; Claude Cahen, “Note sur l’esclavage musulman et le devshirme ottoman: à propos de travaux récents,” Journal of the Economic and Social History of the Orient XIII, no. 2 (1970): 211–18.

• 6. Olafur Egilsson, The Travels of Reverend Olafur Egilsson: The Story of the Barbary Corsair Raid on Iceland in 1627, trans. K. S. Hreinsson and A. Nichols (Reykjavik: Fjölvi, 2008).

• 7. Salvatore Bono, Schiavi musulmani nell’Italia moderna: Galeotti, vu’ cumprà, domestici (Naples: Edizione Scientifiche Italiane, 1999); Godfrey Wettinger, Slavery in the Islands of Malta and Gozo ca. 1000–1812 (San Gwann, Malta: Publishers Enterprises Group, 2002).

• 8. See among others the works by Bartolomé and Lucile Bennassar, Guillaume Calafat, Francesca Trivellato, and Nathalie Rothman on trans-imperial subjects and intermediaries of servile condition (at some point in their life) between Islam and Christendom.

• 9. Nicolas Vatin, “Une affaire interne: Le sort et la libération des personnes de condition libre illégalement retenues en esclavage sur le territoire ottoman (XVI^e siècle),” Turcica XXXIII (2001): 149–90.

• 10. Joshua M. White, Piracy and Law in the Ottoman Mediterranean (Stanford: Stanford University Press, 2017).

• 11. Christoph Witzenrath, ed., Eurasian Slavery, Ransom and Abolition in World History, 1200–1860 (Farnham: Ashgate, 2015).

• 12. See most notably the monographs by Yaron Ben Naeh (2008) and Minna Rozen (2002) on Ottoman Jewish communities.

• 13. J. H. Galloway, “Mediterranean Sugar Industry,” Geographical Review LXVII, no. 2 (1977): 190–92 and passim. On plague epidemics and their impact on the Ottoman Empire, see the monographs by Nükhet Varlık (2015), Birsen Bulmuş (2012), and Yaron Ayalon (2015).

• 14. Halil İnalcık, “Capital Formation in the Ottoman Empire,” The Journal of Economic History XXIX, no. 1 (1969): 109.

• 15. Karl K. Barbir, “From Pasha to Efendi: The Assimilation of Ottomans into Damascene Society, 1516–1783,” International Journal of Turkish Studies I, no. 1 (1979–1980): 68–81.

• 16. The primary term used in Ottoman Turkish (yava) for fugitives applies both to slaves and livestock, similar to the word “maroon” derived from the Spanish cimarrón.

• 17. Yvonne Seng, “Fugitives and Factotums: Slaves in Early Sixteenth-Century Istanbul,” Journal of the Economic and Social History of the Orient XXXIX, no. 2 (1996): 136–69.