CHAPTER 7

OTTOMAN EUROPE

An Ancient Fracture

THROUGHOUT MUCH of the modern age, a large part of Europe—a quarter or a third of the continent—was under the political domination of Islam. That fracture within the continent was not new. To a large extent, Ottoman Europe simply covered the pars orientalis of the continent, the religious and cultural sphere of influence of Constantinople, in opposition to that of Rome. Hence the Ottoman conquest ultimately followed a much more ancient cleavage, though, for some 150 years, between 1541 and the very end of the seventeenth century, it also crossed over that line, especially in Hungary.

The Ottoman Europe was far from homogeneous, however, and Istanbul’s power was hardly of the same nature and the same strength everywhere. In actual fact—if not in the sultan’s discourse—three circles could be distinguished. They did not always correspond to present-day national divisions.

THE THREE CIRCLES OF OTTOMAN DOMINATION IN EUROPE

Hungary, the Romanian Countries

The outer circle, the farthest from the capital and the most difficult to control, comprised the countries located north of the Danube and the Sava.

Moldavia and Wallachia were only tributary countries (kharajgüzar) of the sultan. They preserved their own social organization, dominated by the boyar aristocracy, and their own institutions, beginning with their princes, the voivodes, and their religious hierarchy. Their respective territories (after the successive amputations performed by the Turks for strategic reasons) were entirely closed off to the official Ottoman presence, whether it took the form of government agents (especially tax agents), garrisons, or representatives of the Muslim faith. The Ottomans did exert their influence, however: these countries annually paid the sultan a tribute, which grew in size over time and was supplemented by contributions called “gifts” (pishkesh). In addition, they were regularly called on to provide auxiliary troops for the Ottoman campaigns and certain quantities of goods and raw materials constituting their principal wealth (salt, cattle, wool). The voivodes ruled only with the sultan’s approval, whether granted a priori or a posteriori, and only so long as it pleased him. They were chosen from among the offspring of the country’s great families. These children spent their youth and received their training in the sultan’s capital, where they were held hostage. They were therefore more or less “Ottomanized” before their reigns began. Ottomanization from above developed further in the eighteenth century when, after the defection of Dimitrie Cantemir, prince of Moldavia, the Ottomans stopped appointing voivodes from the Romanian aristocracy, replacing them with Phanariots, that is, the important Greek or Hellenized families of the Phanar district of Istanbul.

As of 1541, Transylvania was also a tributary state, but its tribute was lighter and the situation unusual. In addition to Ottoman suzerainty, Transylvania recognized that of the Habsburgs with the Treaty of Szatmár (Satu Mare) and the Treaty of Speyer (Spire) of 1570, signed by the voivode John Sigismond Zápolya. Moving back and forth between these two vassalages, the voivodes sometimes sought to become autonomous, for example, during the Thirty Years’ War. The population was organized into three “recognized” nations (the Hungarians, the Saxons, and the Székelys, or Sezklers, Hungarian-speaking peoples distinct from the Hungarians). The Romanians, despite their large numbers, were not recognized as a nation. Multidenominationalism was officially established, with four different “accepted” religions: the Hungarians were Catholics or Calvinists; the Saxons were Lutherans; and the Székelys were Unitarians, with the creation of that new church by Francis Davis, bishop of Kolozsvar (Cluj). The Orthodox faith was tolerated but not “accepted.”

Central Hungary, as well as the Banat of Temesvár, Slavonia (the countries between the Sava and Drava), and certain parts of Croatia, were in principle part of the empire. These were Ottoman provinces that had administrators and military staff representing the central power and that possessed the institutions specific to the empire. But the region displayed very distinctive features, associated with its distance from the center, its relatively late integration (which would also be limited in time), and its lasting situation as a border zone (serhadd). There the Muslim element was reduced to a narrow stratum of administrators, soldiers, merchants, and artisans, confined to a few chief cities (Buda, Pest, Pécs, Székesfehérvár, Szeged). Furthermore, these Muslims were often not Turks but Islamized Bosnians. The rural areas and a good part of the cities remained wholly Christian and largely autonomous. A peculiarity of that situation was the dual tax system, set in place not only on the borders between Royal and Ottoman Hungary, but in localities within Ottoman Hungary itself. In addition to the Ottoman system, feudal lords, now living in the Habsburg part, continued to levy taxes on their subjects and even to enforce their own laws.

Greece, Serbia, Montenegro, Bosnia-Herzegovina, and Albania

Inside the outer circle just described was a transitional zone, bordered on the north by the Danube and Sava rivers, and on the east by northern Bulgaria and the Vardar Valley. These lands were Ottoman but still remained fairly remote from the center of the empire, and they had common borders with Venetian and Habsburg possessions. The Muslim population there was confined to certain cities and towns located on the former routes of Turkish penetration or on old border fronts. In addition, the proportion of Christian converts was higher than among Turkish settlers. Belonging to the zone thus defined were continental and Aegean Greece, Serbia, Montenegro, Albania, Bosnia-Herzegovina, plus a little appendix, perfectly autonomous but still a tributary of the sultan and in communication with the Christian world: the republic of Dubrovnik (Ragusa). That intermediate zone included many other enclaves, which were largely autonomous because of their location and natural conditions. These regions were not easily accessible and were the source of only mediocre revenues. Installing the timar regime of land grants in compensation for military service was therefore out of the question there. The ancient tribal systems continued to operate. That was the case for Montenegro, where the traditional cadres were under the authority of the vladika, the Orthodox bishop residing in Cetinje. It was also true of the northern mountains of Albania. For the most part, and increasingly from the sixteenth century on, the sultan drew contingents of warriors from these regions.

Greece also included mountainous districts, remote from Turkish authority and influence, such as the Mani Peninsula south of Peloponnesus, the district of Suli in Epirus, and that of Agrapha in the Pindus Mountains. In the Aegean Islands as well, and in the monastic republic of Mount Athos, on the Chalcidice Peninsula, various modes of self-administration occupied a large place.

Also associated with that second circle—because of their distance from the center and the absence of revenue distribution in the form of timars—were the provinces north of the Black Sea, neighbors to the khanate of Crimea and the Tatar steppes, the sanjak (and later the eylet) of Kefe and Akkerman. Süleyman the Magnificent summed up the situation in June 1560 when he wrote to the khan of Crimea: “When the troops bringing victory are sent over there, they will encounter serious obstacles, since there are enormous rivers to cross and traverse.” And he continued: “Given the distances, when troops are sent there, difficulties of all sorts emerge.”¹

Bulgaria, Thrace, Thessaly, Macedonia, and Dobruja

Finally, the innermost circle of Ottoman possessions in Europe included Bulgaria, Thrace, Thessaly, Macedonia, and Dobruja. The lineages of akinji beys, whose ancestors were the true conquerors of these zones when the Ottomans first came to live there, retained strong local prestige, as well as a secure and profound importance via ancient pious foundations. These were the Evrenos Oullari in Macedonia; the Mihal Oullari in northeast Bulgaria; the Turahan Oullari in Thessaly; and the descendants of Ishak Bey in Skopje. But the heirs of these dynasties had become the sultan’s loyal servants, like his other provincial governors.

This circle included the provinces that were conquered first and that were closest to the two successive capitals, Edirne and Istanbul. This was Rumelia in the strict sense, the part of Europe most firmly rooted in the Ottoman Empire, with no common borders with other European countries. It was only there that the Muslim population, whether converts such as the Pomaks of Bulgaria and Greece, or Turkish settlers from Anatolia, was of a considerable size, at least in some cities, such as Skopje, Nibolu (Nikopol), Kyustendil, and Trikkala.

A MULTIFAITH EUROPE

A major feature of that Ottoman Europe was that Islam remained in the minority in terms of numbers, even in the parts most under Ottoman control. The dark predictions made in the letter that John Hunyadi sent to Pope Nicholas V on September 17, 1448 (and in other texts)—a formal epistle drafted by the Hungarian humanist János Vitéz—did not come true. In it, the hero of the anti-Ottoman struggle declares: “If my memory does not fail me, the spiteful weapons of the Turks have been lurking around Europe for a hundred years now. They conquered Greece, Macedonia, Bulgaria, Albania in quick succession … enslaving them, depriving them of their religion, forcing onto them foreign face, foreign morals, foreign laws and the language of the infidels. They showed no mercy either to the rights of the people or to those of God.”²

These lines evoke the threat of systematic religious and cultural assimilation, combined with the desire to incite the recipient to energetic action; but that assimilation did not take place. In other words, what had occurred in Asia Minor between the end of the eleventh century and the thirteenth century did not occur in the Balkans of the fourteenth to fifteenth centuries.³ In Asia Minor, even though Christian (Orthodox, Gregorian, Nestorian) and Jewish elements survived, the bulk of the population had gone through a rapid religious and cultural transformation—both Islamization and Turkification. The former Byzantine territory had become a single “Turkey,” even though, at a profound level, heterogeneous substrata long persisted beneath that apparent unification.

We must not underestimate the enormous changes produced by the Ottoman conquest in the zones of Europe in question. They affected the countries annexed to the empire but also, to a lesser degree, the countries that were merely vassalized. New institutions, a new political and social configuration, replaced the prior order. The ethnic map shifted after the wars, as a result of spontaneous or forced displacements of the populations, but also because of the integration of these regions into a much vaster economic unit. A place must also be made for Eastern cultural contributions in areas such as architecture, furniture, clothing, and cuisine, along with the vocabulary to designate these new realities.

Nevertheless, these changes went hand in hand with strong ethnic and religious continuities, which prevent us from imputing a radical break to the Ottoman conquest. Alongside Islam, that newcomer in the conqueror’s appurtenances, the previous faiths remained, with their respective beliefs, rites, and clergies. Although the Ottoman occupation established an Islamic political regime and haughtily asserted the superiority of Islam over other religions—as illustrated, for example, by the transformation of some of the churches into mosques—it also organized religious pluralism. The contrast between the two Europes can in no way be reduced to a simplistic opposition between a Christian Europe and a Muslim Europe. On one hand, there was a multifaith Europe where Islam was institutionally (but not demographically) dominant; on the other, a Europe with universalist religious aims, whose unity was dramatically shattered by the Reformation, and within which Judaism itself incarnated an alterity that was never totally accepted.

It is also not possible to claim that, within the coexistence of religions instituted by the Ottoman framework, each remained entirely closed on itself. On the contrary, it is very clear that, up to a certain point, that situation favored mutual influences, a cross-contamination of popular practices and beliefs. Saints, pilgrimage sites, and ceremonies were sometimes shared by several religions. But these phenomena of osmosis were not accompanied by shifting identities. The respective clergies made sure of that: everyone knew to which community he belonged, so long, at least, as he did not take the leap of conversion.

Let us consider a few statistics, insofar as the sources allow, about the persistent majority status of Christianity in Ottoman Europe. Ömer Lutfi Barkan’s studies, based on the Ottoman census records from the sixteenth century, show that, in the years 1520–1535, the beylerbeyilik of Rumeli counted a total of 862,707 Christian households—most of them Orthodox—compared to 194,958 Muslim households, which therefore represented only 18 percent of the total population.⁴ Three centuries later, if we are to judge by the census conducted on more modern foundations in 1831, that proportion had significantly increased. But it is still remarkable that, despite the shrinking of Ottoman Europe at the time and the immigration into this territory of Muslims fleeing the lost provinces, Muslims were still in the minority. At that time, out of a total of 1,334,691 adult males, there were 833,994 Christians and 500,697 Muslims, who therefore represented 37.5 percent of the population.⁵

Both before and after the Ottoman conquest, Europe under Turkish domination remained Orthodox in the majority. Although it had communities of Roman Catholics, called “Latins” or “Franks,” these were established in border regions in Hungary, Croatia, and Albania, or in the Aegean world, where they were a legacy of medieval Latin colonization. In addition, there were missionaries from the Western religious orders, which became increasingly numerous from the seventeenth century on, especially with the creation of the Roman Congregation for the Propagation of the Faith in 1621. Western merchants residing in the large commercial centers of the empire also grew continuously in number; they were granted the status of musta’min and were under the protection of their respective consuls.

Compared to the Christians and the Muslims, the Jews were only small minorities, apart from the famous exception of Salonika, “city of the Jews,” which had a Jewish majority between the early sixteenth century and the end of the Ottoman era. One of the consequences of the Ottoman occupation of the Balkans, however, was a proliferation in that zone of Jewish communities that were very diverse in their origins, languages, and rites. In fact, the Hellenophonic Jews (known as “Romaniotes”) who remained from the Byzantine period—Sultan Mehmed II had deported most of them to Istanbul in the interest of populating his new capital—were later joined by Italiote, Askenazi, and especially Sephardic elements, as waves of persecution and expulsion measures followed one upon another in most of the European states. That last wave of immigration resulted from the far-reaching banishment decrees of the Iberian sovereigns in the late fifteenth century and, by contrast, the relatively welcoming policy of Sultan Bayezid II and his successors. These Sephardic Jews, and the conversos who succeeded them throughout the sixteenth century, settled primarily in a few large and small cities of southeastern Europe (Salonika, but also Vlorë, Patras, Trikkala, Nibolu, Sofia, Skoplje, Serres, Kavala, Kastoria, Volos, Larissa, Sarajevo, Rustchuk, Brila, and others).

THE LIMITS OF TURKISH COLONIZATION IN EUROPE

Two reasons may account for the relatively weak Islamization of Ottoman Europe. First, the conquest of eastern Europe was not on the whole accompanied by a significant emigration of Anatolians. A true colonization movement, composed both of voluntary emigration and of systematically organized deportations, existed only in the early part of the conquest, until about the mid-fifteenth century. In that phase, peasants from western Anatolia and nomads (the yürük) were deliberately relocated along the major roads in the principal strategic zones: the east-west penetration route leading to the Adriatic through Thrace and Macedonia, and all along the valleys of the Maritsa and the Tundza in the direction of the Danube. In addition, nomads were installed in the mountainous parts of the Balkan Peninsula, thereby creating Turkish villages distinct from the Christian ones. Following Barkan, some historians have emphasized the essential role that religious groups of dervishes played in the creation of these villages. Monastery-lodges (zaviye), coupled with attached farming operations, usually constituted their initial core.

Turks were also installed in the conquered fortified cities, which had a strategic value for the new state. Those strongholds that initially resisted the conqueror were emptied of a large part of their previous Christian population, becoming majority Muslim. The Christians who remained were relocated to segregated neighborhoods. Cities that followed that pattern included Nibolu (Nikopol), Köstendil (Kyustendil), Vidin, and Silistra in Bulgaria; Tirhala (Trikala) in the Thessaly region of northern Greece; and Skopje (Üsküp) in Macedonia. For example, Skopje, conquered in 1391, had twenty-two Muslim neighborhoods (mahalle) some sixty-five years later, in 1455, compared to eight Christian neighborhoods. Other cities, those that had negotiated their capitulation, remained majority Christian.

After the mid-fifteenth century, by contrast, the newly conquered territories beyond the Rhodope Mountains and the Balkans gave rise to far less extensive colonization. Emigration was limited to state-ordered deportations to a few military centers on the new borders.

That reduction in the rate of displacement from Asia Minor to Europe has been associated with a weakening of the ethnic Turkish pool in Anatolia, itself attributable to the obstruction upstream, linked to the political situation, of communications between central Asia and Anatolia via Iran, a hostile Shiite zone.

The yürük registries studied by Tayyib Gökbilgin⁶ provide a few notions about the number, at least approximate, of Turkish nomads and seminomads (the yürük) traveling from Anatolia to Europe who belonged to paramilitary organizations. In 1543, 1305 units (ojak) were counted, corresponding to some 160,000 individuals. Although the figures for the seventeenth century provided by later sources are higher (190,000 to 220,000 individuals), we must take into account the fact that, at the time, the yürük organization was supplementing its inadequate labor force by recruiting elements of various backgrounds (Tatars, Islamized Balkans, Gypsies, and others). The bylaws of the yürük of Kojajik, for example, mention emancipated yürük slaves, available elements unattached to a timar, who came from other regions or from Anatolia. It cannot be ruled out, however, that relatively large-scale emigration movements from Anatolia to the Balkans occurred after the mid-fifteenth century. There is information, for example, on a current—in all likelihood, limited in number—of kizilbash elements, that is, Turkomans from Anatolia who were considered heretics and were deported to Peloponnesus in the early sixteenth century. Other operations may have taken place in much later eras, though that history remains obscure. It has been argued, for example, that at the end of the Austro-Turkish wars in the late seventeenth and early eighteenth century, and then again after the emancipation of Serbia and Greece, the Ottoman authorities sought to reduce the disproportion between Christians and Muslims in the parts of the Balkans that remained under their control, by moving Turks (as well as Albanians) from Anatolia to the eastern part of Macedonia, generally along the left bank of the Vardar.⁷

THE DHIMMIS, “PROTECTED UNBELIEVERS”

The second major reason for the (at least relative) permanence of religious identities in eastern Europe after the Ottoman conquest had to do with the regime’s policy concerning religion. Contrary to the discourses of the time—an emblematic example is provided by the previously cited letter from John Hunyadi to the pope—and contrary to an idea that is still very widespread, the conquerors did not conduct a systematic Islamization policy or, more generally, “cultural assimilation,” to use a contemporary expression.

The Orthodox Church was therefore retained with its institutions, clergy, and hierarchy. The determining act of Mehmed II on January 6, 1454, that is, a few months after the conquest of the city, was to reinvest the patriarch of Constantinople in the person of George Scholarios, called Gennadius, a monk who had become known for his ardent opposition to the union with Rome. Through that appointment, the sultan was affirming the multifaith character of his empire. In addition, not only was the Greek church preserved, its authority was also broadened in a sense, through the abolition of the old autocephalous Serbian and Bulgarian churches that had emerged in the Middle Ages. Bayezid I had already suppressed the Bulgarian patriarchate in 1393, and the Serbian patriarchate of Pec, created at the instigation of Stephen Dušan, was suppressed in 1459. Only two institutions mitigated that Greek ascendancy over the Orthodox Church in the Balkans. As the last remnant of independence for the Bulgarian church, the archdiocese of Ohrid maintained a relative autonomy; and Sokollu Mehmed Pasha, even before becoming the last grand vizier of Süleyman the Magnificent, reestablished the Serbian patriarchate of Pec in 1557, granting it to a very close relative (perhaps even a brother). The measures taken by the Ottomans would only gradually produce their full effect. It was not until the second half of the eighteenth century that the elimination or subordination of the former autocephalous sees allowed the patriarchate of Istanbul, under the control of the Ottoman sultan, to become the head of the Orthodox Church as a whole.

The other large non-Muslim religious communities of the empire had more or less similar evolutions, though it is necessary to correct the chronology presented in certain historical myths, which tend to attribute everything to Mehmed II’s conquest of Constantinople.⁸

It has been established, for example, that the head of the Armenians of Constantinople did not become patriarch under that sultan but in reality obtained that title of patrik, as well as a set of particular rights, only in the first half of the sixteenth century. In the case of the Jews, it was Mehmed II who recognized Mosheh Capsali, leader of the Jews of Constantinople, as the haham bashi. But what exactly does that title mean? Some have argued that it applied to the leader of the Jews of Constantinople but did not correspond to chief rabbi, whose authority extended to the entire empire. In effect, according to an autograph opinion of the haham bashi who succeeded him, Elijah Mizrahi (1498–1526), Capsali’s authority was confined to Istanbul and the surrounding area. But the question is largely theoretical, since during Capsali’s time, Mehmed II had assembled most of the Jews within his empire in the capital. The dispersal of the Jews across Ottoman Europe would, on the contrary, be a consequence of the great immigration of the Sephardis after the 1492 expulsion. Nevertheless, the post of haham bashi disappeared in 1526, reappearing only in 1835 in a completely different context. In addition, the Jews’ aversion to any centralization did not prevent the Ottoman power from recognizing their community authorities, religious or not, as they did for all the other religions in the empire.

In that matter, the Ottomans based themselves primarily on Islamic canon law, the Sharia, especially its Hanefite version, which they embraced. They inherited the principle of dhimma, which had been applied by most Muslim regimes before them (as we have seen in part I), with the exception of the most radical. In doing so, like many of their predecessors they displayed pragmatism, even as they placed themselves within Islamic legality: they had to take into account the fact that, especially in Europe, they ruled over regions where Islam was by far the minority religion.

By virtue of the dhimma pact that bound the sultan to the dhimmis, non-Muslim subjects who had pledged their obedience, the dhimmis enjoyed religious freedom, while at the same time—the less positive side—certain obligations and specific forms of discrimination were imposed on them. They were compelled to pay a specific tax symbolic of their submission: a capitation called jizya or kharj (or bashkarj, to distinguish it from the tribute of vassal countries). The Ottomans added a few other royalties specific to the dhimmis, just as they imposed special rates for certain taxes that the dhimmis paid along with Muslim subjects. Under these circumstances, the conversion of a dhimmi to Islam—however laudable in principle in the eyes of the authorities—represented a tax deficit to be filled, and was as such regrettable. The dhimmis were the object of a number of prohibitions: they did not have the right to bear arms, to own slaves, or to ride horseback in town. Some garments and some colors were forbidden them, as well as all marks of ostentatious luxury. The list of prohibitions, in fact, had to be adapted to reflect changes in mores and fashions. Only the more austere—even unattractive and humiliating—clothing was allowed, appropriate to a fundamentally debased condition. There was to be no confusion possible between true believers and infidels. All the same, the insistent repetition of proscriptions on that subject is clear evidence of the difficulty in applying them, especially given that an elite of these dhimmis was quite wealthy. Initiatives calling the dhimmis to order generally came from the local Muslim rank and file, whose virtuous indignation was probably not exempt from jealousy and resentment. In response to the complaints reaching it, the central authority could not fail to act as the guarantor of canon law.

Discrimination and harassment did not prevent the status of dhimmi from having one essential advantage: the right to claim to be the follower of any religion or sect whatsoever (so long as one remained outside Islam), to worship and perform rites. That was the difference between Christian Europe and Ottoman Europe, which therefore became a potential refuge for all those proscribed by European Christendom. That possibility had its limits, however. Public order was not to be disrupted, and the dhimmis’ ceremonies had to be performed discreetly. All signs of ostentation were prohibited: the Christians, for example, could have no bells or processions. Their houses of worship could in no case surpass the height of those of the Muslims. Reparations had to be officially authorized and could be made only to return a restored or rebuilt structure to its initial configuration and dimensions. The construction of new places of worship was in principle not allowed.

It is clear, however, that on this point as well, the dhimmis discovered means for circumventing the law, especially by striking bargains with local judges. The writings of Machiel Kiel in particular demonstrate that not only were a great number of churches and monasteries, even relatively modest ones, restored, but new churches were founded and built during the Ottoman period, in Bulgaria and in continental Greece and the Greek isles.⁹

It is true that, apart from maintaining an Orthodox religious life in Ottoman Europe, the centers of high ecclesiastical culture—the Patriarchal Academy of Constantinople and the monasteries of Mount Athos—amounted to very little. The cultural centers that remained most vibrant were therefore located outside the empire: in Crete before the Ottoman conquest of the island and in Italy (Venice in particular), where the printing of religious and especially liturgical texts in Greek, as well as selections from profane Greek literature, proliferated.

The status of dhimmi also entailed a certain community autonomy, especially in judicial matters, since questions of personal law at least (marriage, divorce, inheritance, custody) belonged to the realm of the respective religious laws. Similarly, representatives of the different clergies quite naturally oversaw their respective communities of faithful and served as intermediaries with the Ottoman authorities, especially in matters of taxation, though the secular elites gradually came to compete with them for these functions. Moreover, these aspects of autonomy and self-government do not justify tracing the so-called millet system so far back in time, as is often done. The millet mode of administering the communities of the empire, which established special courts of “personal law” for religious minorities, was much more centralized and structured than what had existed in the previous periods. It would become a reality only in the nineteenth century, in the age of reform.

CONVERSIONS TO ISLAM

The argument that the Ottoman conquest did not entail a systematic policy of forced conversion of the subject populations, but rather their reduction to the status of dhimmis, does not rule out the possibility that such conversions sometimes occurred, during particularly violent and troubled episodes of Ottoman history. In one of the accounts that attest to them, it is claimed that, at the conquest of Tirnovo in 1394, the only ones who escaped the massacre were notables who agreed to convert. Such accounts cannot necessarily be dismissed as anti-Muslim clichés in every case. Moreover, there is nothing implausible about such deeds within the context of invasions of Ottoman territory, during the wars of the late seventeenth and early eighteenth centuries. The Christian subject, always vulnerable to the suspicion that he was collaborating with the invader, found himself in an extremely critical situation after the invader withdrew. It is therefore understandable why, in 1689, Arsenije Crnojevi, patriarch of Pe, who had previously preached insurrection against the Turkish masters, resolved to flee north of the Danube after the Austrians’ departure and was followed by part of his flock—reportedly some thirty thousand families, though the matter remains controversial. In such a situation, the only way to prove one’s loyalty to the sultan was truly to adopt Islam, since the religious and the political were inextricably mixed.

There was one case, moreover, when forced conversion was not “accidental” but rather at the foundation of an institution essential to the state: the devshirme. Through that procedure, which consisted of rounding up young Christian boys from the villages of Anatolia and especially Rumelia, then forcibly converting and circumcising them, the sultan recruited a large portion of his permanent army, especially Janissaries, and constituted a significant part of his ruling elite, political and military, at least until the late seventeenth century. But that procedure was a clear violation of Muslim law. The cavils made by certain jurists in an attempt to justify it scarcely conceal the fact that, in this matter, reason of state had prevailed. In addition, whatever the emotional trauma caused by the practice of tearing children away from their parents (the popular literature in that part of Europe provides ample depictions of it), and regardless of the tendency of Balkan historiographers to focus on that factor of de-Christianization and “denationalization,” it is necessary to assess more accurately the real demographic impact of devshirme during the period it was in force. If we are to believe a late sixteenth-century Ottoman chronicler, Sa’adü-d-dn, that policy was responsible for two hundred thousand conversions.¹⁰

In any event, to minimize the role of conversions imposed by force is not to claim that there were no conversions of any other kind.¹¹ Apart from violence and blackmail, other motivations, complex in some cases, could lead some people to opt, willingly or not, for Islam. These behaviors fall into the general category of social opportunism. Someone might have wanted to escape the taxes on the dhimmis or even a legal punishment (as in the emblematic case of the “Jewish messiah” Shabbetai Tzevi, who escaped execution by adopting Islam in 1666); or to advance in the social world and especially, gain access to public sector positions; or to survive banishment by one’s community of origin and cut short one’s prosecution; or finally, to obtain a reward, a job, or a pension. Moreover, for vulnerable persons—slaves, or even wives or orphans—pressure alone could make conversion inevitable. It seems that, in spite of everything, these individual acts were few in number. It has been calculated, on the basis of the jizya registries, that in sixteenth-century Rumelia, conversions of that kind did not number more than a few hundred a year. It is true, however, that not only the wartime situations noted earlier but also the evolution in the nature of Ottoman power itself may have had some influence on the volume of conversions. As that power became more closely identified with Islam, proselytism in ruling circles became more forward and insistent. That was especially the case in the second half of the seventeenth century under the reign of Sultan Mehmed IV and the grand vizierate of Köprülü Fazil Ahmed Pasha, both followers of an Islamic radicalism of the “Salafist” type.

In certain parts of Ottoman Europe, conversions were much more massive in scope: in Albania, Bosnia, and Crete, and in the regions of Bulgaria and Macedonia inhabited by Pomaks. What causes are we to attribute to these phenomena? Without becoming involved in the often very sharp polemics, I shall confine myself to two remarks. First, the causes were certainly not the same everywhere or at all times. It is therefore necessary to seek out the dynamics at work in the contexts particular to each case. Second, Islamization seems to have occurred more slowly than people came to believe in retrospect. In the case of Bosnia, the Ottoman census records for 1489, that is, twenty-six years after the conquest, list twenty-five thousand Christian families and only forty-five hundred Muslim families. The conversion movement began among the elite in the Bosnian feudal system. The picture was completely different in the late eighteenth century: at the time, 265,000 Muslims, 253,000 Orthodox, and 80,000 Catholics were counted.¹²

UNDER THE DOMINATION OF THE CRESCENT

All in all, the Islamization of Ottoman Europe remained limited, but that centuries-old occupation produced a “Balkan Islam” whose legacy survives today, though the map was appreciably modified by the wars of national liberation and the Balkan Wars of the nineteenth and twentieth centuries.

Although Muslims were numerically in the minority in Ottoman Europe, Islam was the dominant religion there, in that it was the religion of the masters: the sultan and the civil and military representatives of his power. It was in terms of that one true religion that they judged the two earlier “religions of the Book”: partly true but incomplete religions and, since the teachings of the ancient prophets had been forgotten, erroneous on several points. Followers of these religions, persisting in error, could not help but inspire a certain disdain. Conversion was the best thing that could happen to them, and though they could not be forced to take that step, they could only be praised for doing so. It was with Islam alone that the state identified itself; the resources of the treasury went to its buildings and charitable works through the pious foundations of the sultan, his family, and his dignitaries. Although in a sense the officiants of all faiths (the patriarch, his metropolitans, and his bishops, for example) could be considered cogs in the state machine, the ulemas took precedence and were the only ones who could benefit from the state’s largesse and speak in its name.

THE PLACE OF NON-MUSLIMS

The traditionally accepted schema, according to which careers in the administration and in the military were for Muslims, while Christians and Jews were confined to economic occupations (farming, artisanship, and commerce) is not baseless in Ottoman Europe, but a few nuances need to be added. Apart from the fact that there were not only Muslim peasants and artisans but also merchants and even large-scale traders—even more than has been said—the state, in the interest of pragmatism, did not always systematically forgo the dhimmis’ military service. After the conquests of the fourteenth and fifteenth centuries, local Christians of distinction could be granted timars without having to change religion. In Albania in 1431, that is, twenty years after the first Ottoman conquests in that country, 60 of the 335 timariots were identified as Christians (the region even had a Jewish timariot at the time). In 1455, in the district of Tirhala in Thessaly, 36 of the 182 timariots were Christian. At the same date, that is, sixty years after the conquest, the proportion was 27 out of 170 around Priština, Kosovo. Similarly, in the Serbian district of Branievo, in the Timok Valley, 62 Christians were identified in 1468 among the 125 timariots.¹³ Although there were still attributions of timars to Christians under Bayezid II, the phenomenon did become more unusual in the sixteenth century, and, in any case, the descendants of Christian timariots converted to Islam.

More long-lasting, however, was the persistence of old Balkan military organizations under the Ottoman regime. When necessary, the new masters modified their original character and function, reducing them to paramilitary organizations or auxiliary corps. But the Ottomans nevertheless retained them, along with their old hierarchies and their Christian composition, total or partial. Such was the case for the voynuk, which, under the name voynici, constituted a petty nobility in Stephen Dušan’s empire and became, in the Ottoman armies, a corps specializing in the breeding and keeping of horses. Similarly, the role of border guards fell to the “Vlaches” of Serbia, and the martolos served as auxiliaries in the garrisons of fortresses or of local police forces.

To conclude, it is clear that the Ottomans radically altered the fates of the Balkan peoples, even while refraining, on principle or by necessity, from any true assimilation policy. They left in place a situation upon which, when the time had come, national rebirth would base itself in that part of Europe. In addition, the existence of these Christians, placed under the “Turkish yoke” in a lasting manner, would constitute an important factor in relations between the other Europe, the Christian one, and its infidel adversaries. On one hand, “captive” coreligionists would naturally be seen as potential allies and protected persons, whom fate would provide with pretexts for intervention; on the other, the dhimmi would come to be regarded as a potential traitor and, if need be, as a hostage.