Rebellion, Sovereignty, and Islamic Law in the Ottoman Age of Revolutions

Abstract

This article traces how Ottoman interpretations and uses of Islamic law adapted to deal with non-Muslim rebellions between roughly 1769 and 1830. Drawing on Ottoman fatwas (legal opinions), bureaucratic registers, sultanic decrees, and chronicles, as well as British diplomatic records, it argues that the Ottoman state actively reinterpreted its commitment to the Islamic legal tradition in order to forge the law of rebellion into a weapon against both foreign and domestic enemies. The Ottomans first used declarations of rebellion to mobilize military forces against its imperial rivals, then to discourage and suppress domestic dissent, and finally to deny rebels’ sovereignty under international law. In doing so, the Sublime Porte effectively redefined sovereignty by interleaving Islamic and international concepts. The article places this development in global context, arguing that it resembled legal moves made by Atlantic states as they dealt with the Age of Revolutions. At the same time, this story shows how the Ottoman state’s commitment to Islamic law, and its employment of officially authorized scholars to interpret that law, could both license and constrain state policy. The Islamic legal tradition was flexible, but also meaningful. Interpretation occurred at different levels, and was done by different actors. While the state had significant latitude, it faced limits arising from the substance of the law and the process of interpretation itself.

In 1822, in the midst of the Greek rebellions that would eventually lead to independence from Ottoman rule, Sultan Mahmud II dispatched a fleet to the island of Chios (also known as Scio, or Sakız). By the time the fleet arrived, most of the rebels had departed from the island—leaving only the local population, which had not supported them. Nevertheless, the Ottoman force killed many and enslaved others, especially women and children. The sight of these white, Christian captives being sold openly on the streets of Istanbul shocked European diplomats, and the incident became a cause célèbre in Europe, as, for example, in Eugène Delacroix's painting The Massacre at Chios.¹

When the British Ambassador, Lord Strangford, protested, the Ottoman reisülküttap (chief scribe/foreign minister), Canib Efendi, pointed to Ottoman interpretations of Islamic law. “[T]he Captives taken at Scio,” he declared, “were condemned to slavery by the Mussulman Laws and Religion—which not only permitted, but enjoined such a disposal of the wives and children of their Enemies.”² Indeed, the Ottoman Empire's chief jurist, the Şeyhülislam Abdülvehhab, had issued several legal opinions (fatwas) to this effect.³ Canib went on to contrast Ottoman adherence to the law with “the Christian Powers of Europe.” These nations (to quote Strangford's paraphrase of Canib's words), “without having the plea of Law or Religion . . . had for ages tolerated Slavery—not because their Messiah commanded it, but because it was a source of gain—that it was true, England had abolished it, but that it was only of late years that we had found out that it was wrong—and that half of Europe still differed from our opinion on the subject.”⁴

This exchange is intriguing in itself, especially for Canib's arch reference to the Atlantic slave trade, but it also points to a larger story, one that can illuminate key questions in both global legal history, and the history of the Middle East. Throughout the late eighteenth and early nineteenth centuries—the Atlantic “Age of Revolutions”⁵—domestic rebels increasingly challenged the Ottoman state, from modern-day Romania to Serbia to Greece. Many of these rebellions have been explored as part of national historiographies, but they have seldom been put together into a coherent story—let alone one framed around law. This article, drawing on Ottoman legal opinions, imperial orders, diplomatic correspondence, chronicles, and travel narratives, aims to trace state responses to those rebels, particularly when they were Christians.

I argue that between roughly 1769 and 1822, the Sublime Porte (the Ottoman central government) developed and deployed existing rules of Islamic law as a weapon against non-Muslim unrest. In doing so, the Ottomans redefined sovereignty, and demonstrated both the possibilities and the limitations implied by their commitment to the Islamic legal tradition. An analysis of that development offers important lessons about law and rebellion in the Age of Revolution, about the Ottomans’ place in that history, and about Ottoman state engagement with the Islamic legal tradition.

Through this period, the state used the resources of the Islamic legal tradition in order to define the empire's sovereignty, to threaten rebellious communities with death or enslavement, and to motivate Ottoman troops. Those who resisted the sultan, the empire insisted, were subject to his retribution under Islamic law, even though they could never escape his sovereignty under international law. This reflected a changing Ottoman conception of sovereignty, which was in turn shaped by the empire's geopolitical position. In earlier centuries, the Ottoman state, and its legal interpreters, had been content to regard sovereignty as a matter of possession: the sultan's sovereignty reached as far as his army and navy. Now, as the Ottomans lost more wars than they won, they retreated to a more abstract concept of sovereignty as a way to assert rights over territory that they could no longer control militarily.⁶ In human terms, this meant that rebellious populations were subject to enslavement, and the proceeds of their sale could enrich the militias sent to suppress them.

Although European observers, like Strangford, were shocked by some of the results, the Ottomans’ legal position was not so unusual. Powers such as the United States and France resorted to a similar dual definition of sovereignty: they argued that rebels were subject to attack as if they were co-equal belligerents in a war, even though they would not recognize those same rebels as independent or sovereign. We should therefore see Ottoman responses to rebellion, including in the famous Serbian and Greek revolts of the early nineteenth century, as part of a broader global history.

At the same time, this story also helps us understand a long-standing question in Ottoman historiography: how much did the Islamic legal tradition, or sharia, constrain the Porte? Ottoman diplomats—like Canib—certainly found it useful to claim, in the face of growing European power, that the sharia tied their hands on one issue or another.⁷ Both supporters and opponents of the sweeping Ottoman reforms of the nineteenth century, and the autocratic sultans who imposed those reforms, appealed to Islamic law.⁸ European contemporaries, too, were invested in this tension: some believed that the sharia, and the jurists who interpreted it, malignantly constrained modernization. Others saw it as evidence of “Oriental despotism”—or, conversely, as a restraint on the sultan that prevented despotism.⁹ These debates echo down to the present: historians of the Islamic legal tradition have suggested that it was a safeguard against authoritarianism, one that functioned to some degree until centralizing modern states seized control of the law and disempowered the scholars who made it and safeguarded it.¹⁰ Baki Tezcan has even suggested that Ottoman scholars helped provide a sort of constitution for the early modern empire.¹¹

The story of how the Ottomans responded to rebellions in the Age of Revolution shows that while statesmen had some flexibility to maneuver within the Islamic legal tradition, the process of legal interpretation could still shape and channel their actions. The production of legal opinions helped legitimize imperial action, but at the same time it forced statesmen to reckon with the legality of their preferred policies. The process of interpretation afforded the Porte several ways in which it could, in good faith, justify its actions, but at the same time, the categories and logic of the tradition limited and structured the options from which statesmen could choose.

Rebellion and Enslavement

Like any state, the Ottoman Empire faced both foreign and domestic enemies throughout its history. The end of the eighteenth century and the beginning of the nineteenth saw particularly severe outbreaks of brigandage and banditry, due in large part to Ottoman military policies. The eighteenth-century Porte relied largely on irregular forces, rather than the (once) disciplined janissaries who had once terrified Christian Europe. These irregulars were enlisted for one campaign at a time, and paid in cash—or sometimes left unpaid.¹² Some turned to plunder in the field, or to brigandage during and after wars, and local populations often took up arms to resist them.¹³ At times the resulting unrest, coupled with Ottoman elites’ attempts to take advantage of it, rose to the level of outright rebellion: most famously in Serbia in 1804–16 and in Greece in 1821–30, but also in 1769–70 in Moldavia and Wallachia, in 1770 in the Morea/Peloponnesus, in 1788–90 in Moldavia and Wallachia again, and in 1821 in Moldavia yet again. Many, but not all, of these revolts coincided with the Ottomans’ frequent wars against Russia.¹⁴

Ottoman soldiers, like those of many empires before them, had long sold foreign captives into slavery, and this was an even more important source of income as the soldiers were left underpaid.¹⁵ Both Muslims and Christians believed it was acceptable to enslave each other, and this provided a license for Ottoman forces fighting against rival Christian empires like Russia, Austria, and Venice. By the late eighteenth century the Porte took captured enemy soldiers into its own custody as prisoners of war, but civilians remained fair game for private enslavement.¹⁶ This did not apply to the residents of friendly states, with whom the Ottomans had agreements of friendship and commerce (the famous Capitulations).¹⁷

When campaigning within Ottoman territory, matters were more complex. Ottoman Muslims were immune to enslavement, although they could be executed or punished for crimes. Christian subjects were officially considered zimmîs (from the Arabic dhimmi), protected against enslavement in exchange for paying the poll tax (Arabic jizya/Turkish cizye). Announcing and enforcing that immunity was an important part of the Ottoman state's efforts to assert sovereignty.¹⁸ However, the enslavement of zimmîs could become legal if they lost their zimmet (the status of protection; from Arabic dhimma).¹⁹

Legal Interpretation

How precisely was that protection lost? This article will argue that the changing answers to that question illuminate the relationship between the Ottoman state and Islamic law. But first we need to understand the legal tradition within which the Sublime Porte operated. “Islamic law,” as used in this article, refers to jurisprudence: the rules produced by scholars following the Islamic legal tradition. These rulings were derived from the Qur'an, the hadith (sayings attributed to the prophet), and reasoning based on analogy and scholarly consensus. Within Sunni Islam, there are four main schools of legal interpretation: the Hanafi, Hanbali, Maliki, and Shafi'i, of which the Ottoman state preferred the Hanafis.²⁰

The Ottomans were far from the first Muslim state to encounter questions about rebellion and enslavement. The eighth-century Islamic jurist Muhammad Shaybānî, who was very influential in the Ottomans’ preferred Hanafi school, discussed when disobedience allowed enslavement. Rebellion alone was not sufficient; non-Muslims living under Muslim rule who took over their own city but left Muslims unmolested could not be enslaved (although they could be reconquered).²¹ The key issue was whether the non-Muslims “violated their covenant and fought the Muslims.”²² Shaybānî repeatedly associated this type of disobedience with “going over” to the Abode of War (meaning all the lands ruled by non-Muslims, which were presumed to be hostile). This might mean that rebellious non-Muslims physically moved to the Abode of War, or it could mean that they brought the Abode of War to them by taking over Muslim territory such that “there was no Muslim population between them and the inhabitants of the territory of war.”²³ For Shaybānî, it seems, there was little difference between non-Muslims rebelling, and non-Muslims joining with outside enemies.

Shaybānî lived centuries before the Ottoman dynasty arose, but the empire officially followed the same Hanafi school of jurisprudence. Legal interpretation in the Islamic tradition was the job of scholars, who might be funded by rulers but were not controlled by them. As a result, some scholars have argued, Islamic law had significant autonomy from political power, and could serve as a site of resistance to it.²⁴ The day-to-day work of legal interpretation was done by muftis, jurists whose popular recognition and education entitled them to issue answers to particular questions about law and ethics. Such answers, or fatwas, were formally non-binding but could be very persuasive to individuals, to the state, or even to judges. Typically a mufti, or his staff, would translate an applicant's specific question into more abstract terms, in order to emphasize the particular legal point in question. He would then answer with a simple “yes” or “no,” and might add more explanation or citations to earlier authorities.²⁵

In the Ottoman Empire, one mufti was particularly important. Early Ottoman sultans created the office of şeyhülislâm, or chief mufti. This scholar, appointed by the sultan, held a particularly lofty status, so much so that he rarely cited other authorities in his opinions. Anyone could petition the şeyhülislâm for a fatwa, but the Ottoman government particularly relied on him to legitimate state actions. Richard Repp, in a study of the office, portrayed the şeyhülislâm merely as a rubber stamp, “corrupted” and “subordinated to the secular government.”²⁶ More recently Abdurrahman Atçıl's study of the sixteenth century, while emphasizing the şeyhülislâm's ties to the state as the “top official in the hierarchy of scholar-bureaucrats,” argues that the sultan did not fully control access to the juristic profession, or, perhaps, to the substance of the law.²⁷

Probably the most famous Ottoman şeyhülislâm, who served under the even more famous Sultan Süleyman (r. 1521–66), was Ebussuud Muhammed (d. 1574). He shaped the institution and its practices, and is often said to have reconciled sultanic and customary law (kanun) with the sharia. Among many other legal opinions, Ebussuud addressed the question of non-Muslim rebellion.²⁸ Ottoman şeyhülislâms’ fatwas typically did not cite their sources, but Ebusuud surely drew on the Hanafi tradition, and perhaps on Shaybānî himself, in answering this inquiry:²⁹

Question: Does the [sharia] permit the enslavement and buying and selling of those zimmîs who rebelled against the Padişah of Islam, and who were suppressed by his order, like other harbîs [infidel inhabitants of the Abode of War]?

Answer: If their rebellion necessitates the abrogation of the zimmet, yes. [But] the zimmet cannot be abrogated when they decline to pay taxes, turn out to be thieves and robbers, or take refuge in the mountains because of the timariots’ [semi-feudal local officials] oppression …Unless they join the Abode of War or make war with us, they cannot be enslaved.³⁰

A parallel fatwa specifically noted that only if non-Muslims, as a community, joined with the Abode of War, or were annexed by the Abode of War, would they lose their zimmet and become subject to enslavement.³¹ In a sense, it was leaving—whether in body or in spirit—that made the difference.³²

At the same time, Ebusuud's reasoning implied that crimes such as tax evasion, theft, robbery, or flight, while they might merit punishment, did not rise to the level of abrogating zimmet and thus allowing enslavement. He made this clear in another fatwa, explaining that even banditry did not cast communities out of the Abode of Islam. Bandits (presumptively male) themselves, he found, were subject to different treatment than their female and minor relatives. “It is lawful for the thieves to be killed,” but “their wi[ves] and children would not be enslaved.”³³ Criminals—even non-Muslim ones—were to be dealt with through individualized punishment, whereas enemies were subject to communal enslavement. This was not just a matter of doctrine, but was also applied, for example in revolts in the Aegean in the 1570s.³⁴

Ebusuud's opinion did not create a binding precedent as a common-law judicial decision would, but it was common for other muftis to follow earlier şeyhülislams’ rulings.³⁵ But more importantly, Ebusuud's fatwa suggests how scholars understood the law, and the options available within it, in the sixteenth century. Similarly, the seventeenth-century şeyhülislâm Feyzullah Efendi, whose rulings were collected in an influential compendium, issued several fatwas reiterating that zimmis “broke the pact” when they joined with the Abode of War.³⁶ It is not clear that this rule was applied on a large scale in the seventeenth and early eighteenth centuries, but in 1716, the Porte did authorize the enslavement of Moldavians and Wallachians accused of joining the Austrians during wartime.³⁷ From time to time, Ottoman Christians who rebelled or enlisted with foreign pirates also seem to have been enslaved on Ottoman galleys; they are listed in surviving record books alongside other non-Muslims captured from the Abode of War.³⁸ The same record books, however, separately list other zimmîs who were convicted of ordinary crimes and sentenced to hard labor. The latter category were explicitly labeled as still being zimmîs. The distinction shows how the Porte maintained the boundaries between criminals, who incurred just punishment but did not forfeit the protections of zimmet, and rebels or enemies, who were outside the Abode of War and subject to enslavement.

It seems that this distinction between criminals and enemies was the important one for Ottoman jurists.³⁹ Within the latter category, there was no sharp line between “foreign” and “domestic” enemies. In either case, the sultan faced an organized group of antagonistic non-Muslims. Whether or not those enemies had once been under Ottoman rule, they were now in the Abode of War, and needed to be (re)conquered. This approach fit the broader military and legal realities of the time. In a world of empires, sovereignty was divided and layered, rather than monolithic and territorial in the modern sense.⁴⁰ The Ottomans and their imperial rivals—Russia, Austria, Venice, and Persia—traded territories after every war, and in contrast to the twentieth and twenty-first centuries, conquest was a widespread and accepted war aim.⁴¹ If a territory was lost to rebels, then, there was little reason not to consider it part of the Abode of War, temporarily lost and awaiting re-conquest. Moreover, large-scale non-Muslim unrest in the Ottoman Empire was most common along its borders with Christian rivals, so it was natural to assume that rebellion and imperial warfare went together.

Weaponizing the Law of Rebellion

Thus the law stood in 1768, when the Porte launched a new war with Russia. It went badly, and the humiliating 1774 Treaty of Küçük Kaynarca handed Russia considerable territorial, commercial, and propaganda gains. Ottoman defeats stemmed largely from the nature of the irregular troops upon whom it had to rely. The untrained, often undisciplined forces were paid or dragooned into service and all too often went home after military setbacks or when their meager cash salaries were in arrears. Probably more than ever, the troops now sought plunder and slaves to replace the lack of pay. But defeats by the Russian army made it impossible to reach enemy soil. In the words of the chronicler Câvid, “it was difficult to take Muscovite slave girl[s],” as Russia was “a far away country,” and well guarded.⁴²

Poland was more accessible. The Ottomans had gone to war largely because they felt threatened by Russian military intervention in Poland. But now, the Porte decided that the very presence of Russian troops on Polish soil meant that some residents of that territory had abrogated their friendship with the Ottoman Empire. They were therefore fully subject to enslavement as enemy residents of the Abode of War.⁴³ The British ambassador to Constantinople, John Murray, laid bare the Porte's cynicism: “these Excuses for entering into Poland,” he wrote, “seem to me to be entirely groundless: but there is little Plunder for the Troops elsewhere.”⁴⁴ So, in the summer of 1769, an Ottoman force enslaved between 1,000 and 5,000 Poles, who were sold in the army camp for 30 piasters each—or at a bargain, five for 50.⁴⁵

Other soldiers, without authorization, turned against Christians within the Ottoman Empire itself. The Porte condemned this illegal behavior, and in fact executed an insubordinate officer, Kahraman Pasha, in August 1769 partly because “he had plundered all the villages where he passed in Ottoman Moldavia.”⁴⁶

Yet the actions that cost Kahraman his life in August became official policy by December. Moldavia was a mostly Christian principality whose ruler was appointed by, and loyal to, the sultan. It became one of the first battlefields of the 1768 war, and Ottoman leaders worried that its inhabitants would aid the Russians. (They had done so half a century earlier, when Peter the Great launched an abortive invasion of Ottoman territory.)⁴⁷ So in October 1769, the prince, nobles, priests, and elders of Moldavia were sent imperial orders warning of their families’ enslavement, among other consequences, if their people continued to aid the Russians.⁴⁸

Two months later, the chronicler Şemdanizade reports, the Porte carried out this threat, securing a fatwa from Şeyhülislâm Said Efendi:⁴⁹

[Question:] ‘If the subjects of some villages, among the tribute-paying people in the lands of Moldavia, rebel and refuse to pay the cizye and ally with Muscovy and attack the lands of Islam and fight against the people of Islam, is it lawful for their lands to be attacked and for them to be enslaved?’

The answer [was] ‘It is lawful’ and accordingly, in [the month of] Sha‘ban an imperial order was issued.⁵⁰

This query seems designed to meet Ebussuud's standards, listing rebellion and alliance with a foreign enemy (alongside failure to pay the cizye, which did not necessarily make one an enemy). Therefore Said's opinion followed logically from the question; whether or not the failure to pay the cizye was legally determinative, the other factors were sufficient. Most notably, the allegation that the Moldavians had “all[ied] with Muscovy” clearly meant that they had joined the Abode of War.

What is intriguing, though, is that the query itself involved prior legal interpretation, whether by the Divan (the sultan's imperial council), the şeyhülislâm, or his office. The question would have been prepared based on information (and perhaps a draft question) submitted by the Divan or high-ranking officials—even though that information might “beg the real question.”⁵¹ This might mean that the Divan presented an unsuspecting şeyhülislâm with a carefully manicured set of facts, or it might mean (as we will see later) that the şeyhülislâm himself was intimately involved in advising which facts should be presented in order to compel a certain conclusion.⁵²

This was not necessarily a matter of dishonesty or omission—it was intrinsic to the form. Fatwas, by definition, answer questions of law based on a given set of facts. But as Kim Lane Scheppele has argued in the Anglo-American judicial context, the distinction between law and fact is not fixed.⁵³ Any description of the facts that is legally useful “will contain terms invoking legal categories, which mediate between the facts of ordinary description and legal judgments.”⁵⁴ The bare, physical facts must be “processed” to place them into legally relevant categories, before they can be invoked as objects of further legal analysis.⁵⁵ This is precisely what the document did, as it claimed that certain Moldavians had “rebel[led],” “all[ied] with Muscovy,” and “attack[ed]” Muslims. After this first level of interpretation, these findings became the object for Said's own interpretation, moving from what the non-Muslims had done, to what the consequences were.

Not long after the Moldavians were declared rebels, a similar order, again justified by a fatwa, was issued for Moldavia's sister principality of Wallachia. Enslavement was probably not simply a logical consequence of their rebellion, but was actually the Porte's goal. Indeed, Şemdanizade refers to both orders simply as “for the enslavement of the subjects of Wallachia and Moldavia.”⁵⁶ The true intent was, in the words of the British ambassador, “to make Slaves and lay waste the Country.”⁵⁷ It seems likely that he was right: the Porte might have regretted the disorder caused by Kahraman's depredations earlier in the year, but it came to see such plunder and enslavement as the lesser of two evils, once it became clear that enslavement would both deter rebellion and provide sustenance for the troops.

The following year, some inhabitants of the Morea/Peloponnesus also rose against the Porte. This 1770 uprising became known as the “Orlov Revolt,” because it was aided and instigated by the Russian fleet in the Mediterranean, under Count Aleksei Orlov. The largely Albanian forces sent to the peninsula were directed to kill the (male) inhabitants, and to enslave “their women and children.”⁵⁸ This was done, and enslaved Greeks flowed into the markets of Thessaloniki and Sarajevo.⁵⁹

Thus the Porte had weaponized the şeyhülislam’s abstract restatement of the Islamic law of non-Muslim rebellion. Enslavement had been a recognized legal response to rebellion since at least the time of Ebusuud, but it was now used on a much more frequent, large-scale basis, and for new purposes. The declaration of rebellion was now intended and perceived not only as a way to intimidate non-Muslim subjects, but also to facilitate military mobilization by allowing enslavement. Legal interpretation and imperial policy turned unrest into an opportunity, and resistance into a resource.

Honing the Weapon

In 1787, war broke out again between the Ottomans and Russians, and the following year, the Habsburg Empire joined the Russians. Once again, the Ottomans began to see domestic Christian outlaws as having joined with their foreign enemies. In Bosnia, the Porte was troubled by a bandit chieftain named Koçu, and this may have led to orders, issued in May 1788, declaring that the Christian inhabitants of the district of Novibazar (Yenipazar) were in revolt, and could be enslaved if they did not submit.⁶⁰ This was a conveniently timed decision, as the area became the focus of fighting against the Habsburgs.

A similar, but better-documented incident occurred in Moldavia and Wallachia. Ottoman forces, and those of their tributary princes, found that despite prewar prohibitions, Moldavian peasants had enlisted in the Russian and Habsburg armies, with some fighting as regulars and others as guerillas.⁶¹ Ottoman archival and manuscript sources describe these fighters, whom they bitterly blamed for defeats, as “Volunteers” (volonter).⁶² Fighting them presented the classic conundrum of counterinsurgency. The Divan reported to Sultan Abdülhamid I (r. 1774–89) that, “it is doubtless that if we attack them, by [them] taking refuge in Moldavia, it will be impossible to distinguish them from the obedient subjects, and if we do not attack them, they will lay waste the countryside.”⁶³ As Mao would write a century and a half later, the guerrillas were like fish, swimming in a sea of the people.⁶⁴

Abdülhamid and his Divan agreed that they should consult the şeyhülislam, Hamidizade Mustafa Efendi. The sultan himself saw a parallel with the 1770 Greek uprising, noting in the margin of a report that “in the time of [my] late brother [Sultan Mustafa III (r. 1757–74)], the subjects of the Morea also revolted.”⁶⁵ These open-ended deliberations suggest that the Divan was not firmly committed to seeing the Volunteers either as criminals or rebels.

In January 1789, the Divan hardened its position—and solicited a fatwa. A “group” (taife) of Moldavians and Wallachians, “known as Volunteers,” had rebelled and broken the pact and joined with the Muscovites and Austrians. Thus, the fatwa confirmed, they could be enslaved.⁶⁶ Here, again, the Volunteers were identified as enemies by affiliating them with enemy states and the Abode of War.

It was not enough, however, for the Divan to satisfy itself that enslavement was illegal. Officials felt that they also needed to convince the sultan's subjects. If the captured Volunteers were sent to the Ottoman shipyards in Istanbul as state slaves, they might claim that they had been enslaved illegally—since they had after all been born as protected Ottoman subjects. Thus “a crowd of commoners,” probably meaning non-Muslim Istanbulites, “who do not know the meaning of ‘Volunteer’ will probably talk idly, saying ‘are they [the government] enslaving [non-Muslim] Ottoman subjects?’” The Divan therefore decided to execute a few of the leaders in the field, and to make the others row on the vessels of the Danube flotilla, safely out of view of the Istanbul crowd. Fatwas were again requested to confirm the legality of this decision. This was probably not a difficult legal issue—rebels could be executed, and if a given person could be enslaved, then surely the Porte could decide where he would be made to work.⁶⁷

Commanders in the field, however, still saw “difficulties” in the orders—likely because it was ambiguous who was included in the “group” of Volunteers. Discussions had focused on the combatants themselves, but Hamidizade Mustafa, it seems, had not made clear whether his fatwa applied to others. Here it may be that Hamidizade or the Divan preferred not to do much first-level interpretation, simply applying the Hanafi tradition, as stated in Ebussuud's fatwas, in order to find that only the male combatants, not their relatives and neighbors, were subject to enslavement.

The Divan itself—apparently without a formal fatwa, but probably in informal consultations with Hamidizade Mustafa—soon went further. By February, they believed that “most” Moldavian men had become Volunteers. In April, the Divan raised the stakes, reporting to the newly enthroned Sultan Selim III (r. 1789–1807) that because “all” of the Moldavians had rebelled, new orders had been sent out summarizing the fatwas and allowing enslavement.⁶⁸ Hamidizade Mustafa officially agreed, and the orders eventually issued to the regional military commander, the Serasker Hasan Pasha, allowed the enslavement of all those who “broke the pact” (nakz-ı ahd).⁶⁹

Now rebellion was defined communally, rather than individually. Women and children, too, could be enslaved. Here, then, we see the Divan, directly doing what we can only glimpse it having done in 1769: using its own first-level interpretation to find that certain incidents amounted to a legal “rebellion” by “all” of the Moldavians. This was an important threshold to cross, as it opened up far more zimmis to enslavement. That risked a wider rebellion (as Abdülhaimd had feared), but it gave Ottoman commanders more options to motivate their troops. Crossing this threshold did not necessarily require any new facts—it required the Porte to describe the situation in a new way.

At about the same time, Hamidizade Mustafa issued another fatwa, giving a separate rationale for enslaving Moldavians.

[Question]: And the people of the Abode of War occupying (müstevlî) and conquering a kingdom of the Abode of Friendship [here the term refers to friendly territory like Moldavia and Wallachia] which is adjacent to the Abode of War, if the infidels of the kingdom are conquered and defeated by subjugation by the people of the Abode of War, that kingdom being subject to the Abode of War and in the administration of the Abode of War, is the killing and enslaving of the people acceptable?

The Answer: Yes.⁷⁰

The Porte applied this opinion to Moldavia, which was indeed now largely under Russian control. Taken together with the application of the previous fatwa, the Porte's intention to authorize widespread enslavement in Moldavia becomes clear.

This would have made sense militarily: as the spring of 1789 approached, it appeared that the main imperial army would march to face the Russians, who were advancing southward after capturing the fortress of Ochakov in December⁷¹—meaning that the campaign would be fought in Moldavia. As in previous fatwas, Ottoman officials had “processed” the bare facts in Moldavia to reach preliminary legal findings, which the şeyhülislam's fatwa then took as the basis for a further conclusion. The result was, almost certainly, what the Divan had wanted. Any lingering concerns about the reactions of Christians in Istanbul to Moldavians’ enslavement had apparently been overcome by military necessity.

The Porte had thus honed the law of rebellion into a weapon that could be systematically deployed against its own non-Muslim subjects when useful. Fatwas from various şeyhülislams had repeatedly confirmed that such activities caused the loss of zimmet, and hence enslavement. The critical question, however, remained: who was a rebel? The Divan and şeyhülislams gave different answers at different points during the 1787 war, applying the label of “rebel” sometimes to individuals, and sometimes to whole communities or principalities. It is also significant that legal interpretation was not monopolized by the state—the Divan feared that even non-Muslim subjects could interpret the law, and could mobilize based on those interpretations.

Turning the Weapon Inward

The period between roughly 1790 and 1830 was extraordinarily violent, as Sultans Selim III and Mahmud II attempted to gain control over both local notables and bandits—who, as Tolga Esmer has shown, were often intertwined. Leaders chosen to march against them could make their fortunes, either through victory or collusion, while troops could benefit from plunder.⁷²

When the bandits or rebels were non-Muslim, the law of rebellion remained a useful legal and military weapon. But now it moved beyond its eighteenth-century role as an adjunct to foreign wars, to become a domestic counterinsurgency strategy in response to rebellions that had little to do with foreign empires.

This began in Serbia. An uprising there broke out in 1804, and Selim issued orders to suppress the rebels. He allowed the enslavement of combatants themselves (assumed, in the orders, to be male). The orders specifically banned the enslavement of Serbian women and children.⁷³ This distinction echoed the Porte's attitude toward the Volunteers in 1789, as well as Ebussuud's sixteenth-century fatwa: Male combatants were liable for punishment, but women and children were not.

It is notable, however, that Ebussuud had referred to killing bandits—something which was legal even if the bandits were zimmîs or Muslims—rather than enslaving them. Now, the Ottomans mixed genres, speaking the language of war (and thus enslavement) rather than of crime (and thus execution or other punishment). Yet whether killing or enslavement was the answer, the question was still one of individualized rebellion. The offense and punishment might be conceived in “enemy” terms, but it was seen as an individual matter, not a communal one.

Ottoman commanders in the field, however, wanted more. They sought to treat entire Serbian communities as enemies, so that they could enslave Serbian noncombatants, especially women and children. As early as March 1806, they requested the Porte to declare that the entire Serbian community had renounced its pact of zimmet.⁷⁴ Selim seems to have demurred at first, but by early 1807, he accepted reports that the Serbs were enslaving Muslim women and children as sufficient reason to allow their enslavement. Declaring that the Serbs had “become Muscovites,” Selim allowed that they were liable to enslavement. Selim likely consulted with at least one of the şeyhülislams who held office in this period, and obtained a fatwa.⁷⁵ It was probably not a coincidence that by 1807, the Ottomans had gone to war with Russia. Thus the Serbs’ enslavement could be justified by associating them with an enemy empire, the “Muscovites,” just as rebels had been in past wars.

But this association was weaker than it had once been. Even in 1806, a fatwa issued by the Şeyhülislam Esad omitted any reference to the Serbs allying with Russia.⁷⁶ The Porte's language shifted further when orders against the Serbs were renewed in 1813, and again in 1815, because the Ottomans had now made peace with Russia.⁷⁷ The Serbs were now described as being “in” the Abode of War, but not as having joined it. What justified their enslavement was that they had attacked and fought with Muslims and become enemies.⁷⁸ The Serbs were still associated with the hostile non-Muslim world, but—unlike in previous decades—not with a particular enemy state. The language even implied that they might be their own entity in the Abode of War, not part of a foreign empire. The implicit legal assumption that rebels had to be allied with an outside empire, which had long marked fatwas and orders like this, was weakening.

The Porte went even further in the 1820s, as Ottoman Greeks rose up in revolt.⁷⁹ The first uprising, in Moldavia in early 1821, failed to take hold and was quickly suppressed, but no sooner had this revolt failed than resistance erupted far to the south, in the Peloponnesus. This resulted in a decade-long war that led to Canib's confrontation with Strangford, discussed earlier, and eventually to a clash between the Ottomans and the Christian powers at Navarino in 1827.⁸⁰

When the revolt initially erupted, Ottoman authorities were perplexed, partly because it did not seem to be backed by Russia, even though the insurgents had crossed into Ottoman territory from there.⁸¹ (While Russia was sympathetic to the Greeks’ plight, Tsar Alexander I had helped defeat Napoleon, and was now committed to a pro-monarchical, anti-revolutionary policy—including in Greece.) The Divan could not reach a consensus, and legal guidance did not help them. In the initial meeting, the şeyhülislam, Abdülvehhab, declined to issue a fatwa.⁸² He was a particularly loyal supporter and companion of Sultan Mahmud II, and later served as ambassador to Persia. Abdülvehhab also had no sympathy for the Greek rebels, announcing in to the sultan's ministers in 1821 that he would happily take up a musket and fight himself.⁸³ So it seems unlikely that his initial failure to issue a fatwa was a matter of insubordination. More likely, the Divan simply could not settle on a coherent and unified military and legal strategy. They looked to Abdülvehhab for guidance, but the law alone could not solve the political problem.

This initial uncertainty is reflected in Ottoman administrative documents. In July 1821, after a victory in Moldavia, the Ottoman governor of Silistria requested instructions. He reported that some of his forty-six captives, taken in battle, had wives and children in Wallachia. Could he send men to enslave these noncombatants?⁸⁴ Mahmud II and the Divan seemed unsure about what to do, in terms of both law and policy.

For the law, they consulted Abdülvehhab, who now did issue a fatwa. The query began by noting that some zimmîs had broken the pact, and had joined with other enemy non-Muslims to fight against Muslims. Then, the Divan asked, “if this [rebellious male] infidel is taken, his enslavement is permissible by the sharia, but is the enslavement of this infidel's wife and children who are in the Abode of Truce [i.e. Ottoman territory]⁸⁵ also permissible?” Abdülvehhab's answer was “definitely not.”⁸⁶ This followed the approach initially taken in Wallachia in 1789, and in Serbia in 1804–6.

On policy grounds, Mahmud and the Divan may have welcomed this answer as a form of guidance. As Hakan Erdem has shown, there were clear reasons both for and against declaring rebellion and widening enslavement. On the one hand, as noted, the Porte had learned over the past half-century that allowing enslavement was a powerful military tool for recruiting and motivating irregular forces. Erdem describes the “Machiavellian tone” of the Divan as it discussed declaring rebellion in order to motivate Muslims to spend their own resources in suppressing the revolt, in return for the promise of plunder.⁸⁷ On the other hand, as early as March 1821, the Divan recognized that declaring a rebellion might be a self-fulfilling prophecy, making otherwise innocent Christian subjects fearful and angry enough to revolt.⁸⁸

This was a similar, although not identical, dilemma to that which Abdülhamid I had faced in 1788, when deciding whether to send forces against the Volunteers. For Abdülhamid, the choice had been whether to send forces; for Mahmud, the Ottomans’ military weakness meant that sending forces at all required declaring rebellion. But that declaration might have even worse effects. In both cases, the sultans turned to their şeyhülislams, seeking legal solutions for military quandaries. Their requests, both formal and informal, offered little first-level interpretation. This marked a sharp difference from requests in which the Porte clearly desired a certain result.

However, first-level interpretation soon returned, as Mahmud, like Abdülhamid before him, found it better to widen the scope of enslavement. Most notably, in the late summer of 1821, Abdülvehhab issued a new fatwa that did authorize the enslavement of one particular rebel's wife and children.⁸⁹ This again responded to a request from the governor of Silistria, but the Ottoman chronicler Şâni-zâde indicated that officials felt that the previous fatwa issued at his request “did not include” these particular Greeks. Doctrinally, Abdülvehhab justified his new conclusion by noting that in this case, the woman and children had been brought along when rebels fled to a new place and continued to fight.

However, a new political context was probably at least as important as this factual difference: the Porte was moving toward a wider policy of enslavement. As the rebellion went on, the Porte secured fatwas through requests that assumed that whole communities were in revolt. As early as the spring of 1821, the Ottomans began issuing orders, usually with attached fatwas from Abdülvehhab, declaring that individual communities along the Aegean coast were in revolt. These orders specifically authorized enslavement, as did others for parts of the Peloponnesus. Over and over again the Porte declared particular localities as being legally in revolt, and that thus that Christian women and children living in those localities could be enslaved. (Adult men, for their part, had less value on the slave market and were often simply killed).⁹⁰ The 1822 sack of Chios was among the largest of these incidents, and it may have been the most visible to Europeans, but it was far from the only one.

Asserting Sovereignty

The enslavement of the Chiotes did prompt widespread European protests, as we saw at the beginning of this article. However, those diplomatic protests also suggest a shift in how the Porte defined rebellion and justified enslavement: instead of emphasizing the rebels’ ties to foreign powers, as in previous decades, the Porte's legal strategy now relied more on such ties’ absence. Targeted Greek communities had, of course, lost their zimmet, but this was a result of their own actions, not of joining with outsiders from the Abode of War. The fatwas themselves suggest this: Abdülvehhab's opinions on the wives and children of rebels made only a vague reference to the offenders “joining with other infidels.” This was probably a reference to volunteers from the Russian Empire, who accompanied the initial rebels into Ottoman territory. Later fatwas, and the imperial orders based on them, noted simply that Greek communities had rebelled or attacked Muslims.⁹¹ There was far less emphasis than before on rebels directly joining with the Abode of War. This was not a break with the Hanafî tradition—after all, Shaybānî had said that enslavement followed when zimmîs “violated their covenant and fought the Muslims,” and Ebusuud had agreed.⁹² But both of them, and other Ottoman jurists, had customarily portrayed such acts as intimately associated with, and almost synonymous with, joining foreign enemies from the Abode of War. During the Greek War of Independence, that association was weakened, although perhaps not entirely broken.

This made sense in the context of inter-imperial politics. In previous decades, when rebellions had overlapped with wars against Austria and Russia, the Porte had no qualms about admitting that territories had been lost and would have to be re-conquered (which is what would allow enslavement). In 1789, for example, the Şehülislâm Hamidizade Mustafa did precisely this for Moldavia already conquered by Russia. When the Porte was already at war with a foreign empire, lost territories would have to be re-taken anyway, whether or not the Porte admitted their loss. So there was little harm in making that admission.

However, the power dynamics of the 1820s, and the changing conceptions of inter-imperial law, made this position much less attractive. The Ottomans were at peace with their neighbors and no foreign power initially backed the Greek revolt (although the Porte suspected Russian involvement).⁹³ Indeed, avoiding foreign intervention was the key objective of Ottoman diplomacy, since the Porte was unlikely to win a war against Russia, Britain, or France. The main legal tool that the Ottomans had in forestalling such intervention was the idea of sovereignty under contemporary notions of international law.⁹⁴ As scholars like Emer de Vattel had laid out in eighteenth-century treatises, sovereign states did not interfere in each other's internal relations and could not aid rebels unless the rebels could be recognized as independent.⁹⁵ Thus the Ottomans insisted that no European power had the right to aid the Greeks if they were not at war with the sultan. In this context, an outright Ottoman declaration that some of the sultan's territory had been conquered or joined the Abode of War would have been catastrophic. What would then stop European empires from supporting Greece, or even annexing it? Earlier Ottoman jurists could rhetorically cast rebellious zimmîs out of the Abode of Islam, confident that this was temporary, and that they would be reconquered. But Abdülvehhab lived in a different world, both politically and legally. He could not so blithely release the Porte's legal grip on any of its subjects.

Thus the Reisülküttap Canib told the British Ambassador Strangford, in the 1822 discussion that began this article, that the Porte “was an independent Government,” with “a right to act as she pleased toward her own subjects, except where Treaties interfered.” Indeed, he continued, no foreign power—not even Russia—had ever sought to “impose upon Turkey the general principle that she was not entitled to make slaves of her own subjects whenever she chose to do so.” The Divan insisted that “[n]o one's interference will be tolerated in bringing [the Porte's] own subjects back into obedience.”⁹⁶ Many European diplomats agreed: in 1821 Strangford had said that he and other ambassadors “had no right to obtrude their opinions upon the Government to which they are accredited, and with the internal concerns of which, I presume, we are not entitled to interfere,” while the Austrian Chancellor Prince Metternich said that the Christian powers could not “base on principles of public law the support that they would lend to the cause of Greek independence” because “all Europe is in relations of peace with the Sultan.”⁹⁷

The Porte thus had its cake and ate it too: the Greeks were subject to enslavement because they had rejected the pact of zimmet. But they were not eligible for foreign aid because their territory was still under the sultan's legitimate rule. In other words, they were outside the sultan's sovereignty for the purposes of Islamic law, but still within his sovereignty for the purposes of international law.

These claims could only be sustained simultaneously because the Porte had de-emphasized the link between rebellion and joining the Abode of War. Now, rebels need not necessarily have left the Abode of War to qualify for enslavement. The Porte had therefore subtly shifted the nature of the sultan's authority over the Abode of Islam, into something more like sovereignty in modern terms: it covered not just the territories he actually controlled, but those that he legitimately claimed to control, and these claims deserved recognition from other states.⁹⁸

Of course, the Ottomans could not hold off intervention forever. Eventually Britain, France, and Russia imposed a blockade and then destroyed the Ottoman fleet in the 1827 Battle of Navarino, leading to war between the Ottomans and Russia. In the wake of this conflict, the Porte recognized Greek independence, on the authority of another fatwa from Abdülvehhab.⁹⁹ But the Ottomans’ defense of their legal sovereignty helped hold off European involvement for 6 years, largely because their arguments resonated in Europe and France, and Britain feared the geopolitical consequences if they endorsed a Russian war with the Ottomans.

Sovereignty and the Global Context

The Ottomans’ arguments may have worked in part because they were familiar to the Europeans. As the states of the Atlantic world dealt with uprisings in the eighteenth and nineteenth centuries, they too adopted an approach that claimed the rights of both a sovereign and a belligerent.¹⁰⁰ This was more than just a coincidence: such ideas echoed back and forth across the Atlantic, from the Greek War of Independence, to the pages of the New York Times, to a famous United States Supreme Court case of the Civil War era that itself looked back to the Haitian Revolution.

To see this, we can begin once again in the 1820s. The British government, like the Ottomans, also wrestled with the status of the Greek rebels, although for slightly different reasons. They did not want to recognize Greece as an independent sovereign state,¹⁰¹ but they also did not want to regard the sailors of the Greek navy, who sometimes attacked merchant ships trading with the Ottomans, as pirates. The British secretary of state for foreign affairs solved the problem by declaring that the Greeks could be legitimate belligerents, possessing all the rights of a state at war (including the right to attack enemy commerce) even though they were not legitimately sovereign.¹⁰²

Nearly 40 years later, Canning's move echoed in the memory of the Prussian-American legal scholar Francis Lieber. Lieber had himself briefly fought alongside the Greeks as a philhellenic volunteer in the 1820s, before coming the United States, where he advised Abraham Lincoln on the laws of war and wrote a code that became the foundation of modern jus in bello.¹⁰³ In August 1861, Lieber wrote a pseudonymous letter to The New York Times arguing that the United States should adopt the same approach in dealing with the Confederacy. It was a “rule of international law,” he argued, that “when an insurrection had reached a certain height the insurgents were entitled to be treated as belligerents,” so the British had been correct to recognize “the belligerent rights of the Greeks in their insurrection against the Turks.”¹⁰⁴ The Union could do the same to the Confederacy. But, “by conceding to our adversary the character of a belligerent, we admit nothing more than that the insurrection has reached a certain height, and do not in any respect admit to its legitimacy.”¹⁰⁵ The Ottomans had in a sense done the same thing when they allowed the enslavement of Greek women and children: they recognized (whether or not this was justified) that the insurrection had “reached a certain height” such that it was no longer a matter of criminal individuals, but of enemy communities.

This view, which Stephen Neff calls a “dualistic approach,”¹⁰⁶ became the Union's official theory of the war, allowing the federal government to impose measures like a naval blockade, which was normally only legal during a war against a sovereign nation, without recognizing the Confederacy's sovereignty. When Lieber wrote a code of rules for the Union Army (a code that is the basis for much of the modern law of war), he explicitly codified this understanding: “When humanity induces the adoption of the rules of regular war to ward rebels . . . it does in no way whatever imply a partial or complete acknowledgement of their government, if they have set up one, or of them, as an independent and sovereign power.”¹⁰⁷ The United States Supreme Court upheld that theory in the 1863 Prize Cases, with Justice Robert Cooper Grier writing for the majority that “[i]n internal wars, it is competent for the sovereign to exercise belligerent powers generally,”¹⁰⁸ and “[i]t is not necessary that the independence of the revolted province or State be acknowledged in order to constitute it a party belligerent in a war according to the law of nations.”¹⁰⁹

The former statement relied partly on early modern jurists, but more importantly on legal precedents from the Atlantic Age of Revolution. Scholars have emphasized that unrest against centralized rule—whether in North and South America, Haiti, or the Balkans—took many forms during this era. Not all uprisings aimed at achieving national independence, but many ended up there. From an international perspective, the Age of Revolution repeatedly raised the question of the rebels deserving recognition as legitimate sovereigns (or at least, legitimate belligerents).¹¹⁰ This question became particularly pressing when insurgents sponsored privateers (or as the empires called them, “pirates”) and the prize courts of third parties had to adjudicate the legality of their seizures. Thus Grier particularly cited United States Supreme Court cases stemming from the Haitian Revolution against France and Latin American wars of independence from Spain and Portugal.¹¹¹ In all such cases, he found, courts had held that for a sovereign, or for third parties, to treat rebels as belligerents under the law of war, with the right to commission privateers, did not imply recognizing their sovereignty.

This was the legal move that the British made when dealing with Greece. The United States did the same with Spain's breakaway Latin American colonies in 1819, and several decades later with its own rebellious Southern states.¹¹² In its own way the Sublime Porte made the same move, interleaving its different legal commitments so as to separate rebels’ belligerency from their sovereignty.

Conclusion

By tracing this history, we can see Canib's debate with Strangford as the end point of a decades-long process by which the Sublime Porte had crafted long-standing principles of the Islamic legal tradition into a weapon. First the law was tested, then honed, then deployed ever more widely as a counterinsurgency tool. At every step, this was done through conscious choices by the Ottoman government, and at every step the law was adapted to the Porte's political needs. Nevertheless, it was both shaped and limited by the Islamic legal tradition.

First, the declaration of draconian penalties for Christians who sided with Russia in the 1768 War allowed enslavement and plunder to keep the Ottoman army funded. As more widespread revolts, backed by Russia, erupted in the Morea in 1770, the use of such fatwas provided a means to mobilize irregulars to suppress these. Ottoman officials and even Sultan Abdülhamid I recognized such fatwas and orders as a normal tool of governance by the time war began again in 1787. The law of rebellion provided a ready solution to a new problem, that of Moldavians and Wallachians not only rebelling, but enlisting in Russian and Austrian auxiliary units. As the Age of Revolution dawned, bringing with it peacetime revolts aimed at renegotiating the relationship between the center and peripheral populations, it was again legally useful to declare these as being connected to foreign empires whenever plausible. When that was not plausible, however, Ottoman interpretations could shift to turn the law inward, making domestic unrest grounds for enslavement. In doing so, the Porte continued to wield enslavement as a threat against would-be rebels, while also holding out the promise of plunder to motivate disorganized irregulars. But the shifting emphasis of these fatwas also adapted to the Porte's new arguments about sovereignty, which were tailored to foreign and especially European audiences. The fatwas’ new wording allowed the Ottomans to treat rebels as wartime enemies, even while denying the legitimacy of foreign intervention. And, finally, the fatwas allowed Canib to claim that enslavement was not a choice; it was “enjoined” by the supposedly inflexible Islamic legal tradition.

Yet this very history shows that in the hands of the Ottoman Divan and şeyhülislâms, that tradition was flexible enough to accommodate a variety of changing state goals. The Islamic legal tradition was, as Lauren Benton and Benjamin Straumann have argued about Roman law and colonialism, “more resource than road map.”¹¹³ All the actors involved engaged in legal interpretation, and they shaped the law as they did so.

The Ottomans’ interpretive choices were not predetermined. For example, by following Ebussuud's views more strictly, the Porte could have classified unrest more often as banditry than as war. Thus only combatants themselves, not their families or communities, would have lost their zimmet. Or, the Ottomans could have continued to insist that for unrest to become war, the enemies must join the Abode of War by allying with outside enemies.

These choices turned on how one assigned facts to legal categories, and on who was doing the assigning.¹¹⁴ When şeyhülislâms took the lead in interpretation (for example, Ebussuud and, early in the Greek war, Abdülvehhab), they seemed to prefer these more limited approaches. But when the Divan took the lead, they were more likely to construct questions that took for granted the elements of a real rebellion, thus removing zimmet from whole communities. Increasingly as time went on, that was the Porte's preferred approach.

At the same time, the Porte did not have total freedom. Its commitments to the Islamic legal tradition, and to supporting its orders with fatwas from the şeyhülislâm, did impose some constraints. Ottoman statesmen were themselves Muslims, with ethical and religious commitments to the legal tradition—indeed, Ottoman elites customarily saw themselves as equally, and harmoniously, committed to the din ü devlet, the “religion and state.”¹¹⁵ They also had to maintain legitimacy before the Muslim population and the Muslim scholars, of the empire. Şeyhülislâms’ fatwas, and the facts that they were based on, would be scrutinized by several audiences. And as legal process scholars have shown in other contexts, the need to produce compelling legal justifications can itself be an important means of channeling and shaping state action.¹¹⁶ Even if the şeyhülislâm himself was chosen and paid by the sultan, other learned and respected jurists were not, and they might criticize unfounded fatwas. Ottoman officers and Muslim commoners also had to take official legal opinions seriously. And the Porte even worried what non-Muslim subjects would think, as we see from the Divan's 1789 comment about the “crowd of commoners” in Istanbul. European diplomats, too, might value the existence of fatwas (if not their facts, reasoning, or conclusions).¹¹⁷ To convince all of these audiences, as well as to satisfy their own ethical and professional concerns, the şeyhülislam and Divan alike needed to produce convincing legal interpretations, even if their political goals were clear to everyone.

Thus the authorities did not control the law, did not create it, and could neither ignore it nor remake it from scratch. There were hard limits on the Porte's legal autonomy. One concrete and remarkable example of this has been recently noted by Yusuf Karabıçak. At least once in the 1820s, Mahmud demanded to execute all Greek Christians in Istanbul (and perhaps outside it). However, the Şeyhülislâm Hacı Halil demurred, and did not issue a fatwa allowing this. Halil lost his position—perhaps the price of refusal—but Mahmud still felt restrained, noting that he could not move forward without a fatwa since “our sublime state is a Muhammedan state” and people obeyed him as “the leader of the Muslims.”¹¹⁸ Mahmud demanded that a fatwa be issued to clarify the matter, and declared that he himself would act “in accordance with God's word and our Prophet's law.”¹¹⁹ Here the law, and Mahmud's sense of his own duties and legitimacy, clearly did restrain state action.

The law's constraints are also apparent in the Porte's differing reaction to rebellions by subjects of different religions. When rebellion came from non-Muslims, the Porte—as we have seen—became quite adept at arguing that the rebels had lost their zimmet and could be enslaved. But this strategy was simply unavailable against Muslims. In the Islamic legal tradition, Muslims, including the sultan, could not enslave other believers under any circumstances. Regardless of their actions, Sunni Muslims could not join the Abode of War and become enemies; they could only be, in essence, criminals.¹²⁰ No fatwa could change this, however cleverly written. For example, in the early 1800s, Wahhabi rebels in the Arabian Peninsula, led by the bin Saud family (the forerunners of today's Saudi dynasty), resisted Ottoman power. When Ottoman Egyptian troops defeated them in 1818, Wahhabi leaders were brought to Istanbul and executed. But neither they nor their followers were enslaved.¹²¹ The categories of Islamic law gave the Porte legal options against non-Muslim resistance that were unavailable against Muslim dissidents.¹²²

The changing use of the law of non-Muslim rebellion, then, shows both the flexibility and the strength of Islamic law in the Ottoman Empire. It offered resources that allowed the Porte to adapt new legal policies as an age of conflicts with relatively equal imperial rivals gave way to an era of secessionist insurgencies. At the same time, though, the Porte's commitment to the Islamic legal tradition channeled and shaped its actions in both form and substance. This story illuminates the role of Islamic law in pre-modern Muslim governance. But it also points to a story of continuities with other states, both those that resisted the Age of Revolution and, more broadly, those that base their legitimacy on the rule of law.