CHAPTER 3

THE SOCIAL INFERIORITY OF RELIGIOUS MINORITIES

Dhimmis and Mudejars

CHRONICLES OF HOLY WAR celebrated the feats of arms against the infidels and minimized those against coreligionists, except in cases where the latter were portrayed as heterodox. But once the conquest was achieved, the new subjects had to be integrated into the political and social order. These religious “minorities,” who in actuality were often in the numerical majority immediately after the conquest, were usually granted a protected but subordinate place in society. Theologians and jurists justified their subordination, defining their role with reference to the founding texts (Qur’an, Hadith, Bible, or Roman law). From Barcelona to Baghdad, large minorities lived in Muslim and Christian societies. They were sometimes the victims of persecutions, acts of violence, and expulsions, but in general they enjoyed a status where their theoretical inferiority (religious and legal) did not prevent some of them from achieving clear economic and social success.¹

PROTECTED AND INFERIOR: THE DHIMMIS IN THE MUSLIM SOCIETIES OF EUROPE (AL-ANDALUS AND SICILY)

Let us first consider how Muslim law defined the status of the dhimmi, or protected person.² Although the Qur’an does not clearly establish the legal framework for non-Muslims within the dr al-islm, it declares that the Muslim must not force the “peoples of the book” (ahl al-kitb, the Jews and Christians) to convert. By contrast, he may oblige them to recognize the superiority and suzerainty of Muslim authority and to pay “humbly” the jizya, the capitulation tax (Qur’an 9:29). During the great conquests, the victorious Muslims gave guarantees to the conquered peoples, granting them far-reaching legal autonomy and freedom of worship. According to certain chroniclers, restrictions were sometimes among the conditions of surrender applied to the defeated Christians. This is apparent in the Pact of ‘Umar, which, according to Muslim tradition, the second caliph, ‘Umar ibn al-Khattab (634–644), imposed on the Christians of Syria. In fact, these restrictions were imposed gradually, throughout the first century A.H. (beginning in 622 C.E.), and expanded under ‘Umar II (717–720).³ The first author to give us a full version of the Pact of ‘Umar is the Andalusian traditionist al-Turtsh (d. 1126) in his Siraj al-mulk. In that text, the Christians of Syria send a missive to Caliph ‘Umar to remind him of the pledge they made at the time of their surrender. They present a long list of prohibitions that they agreed to respect: on building new churches and monasteries, teaching the Qur’an, wearing “Muslim” clothing or turbans, bearing arms, and so on. A numbers of these measures were aimed at limiting or proscribing the public expression of Christianity. Hence the Christians pledged not to put crosses on their churches, not to display their scriptures in public, not to participate in certain public processions, not to pray in a noisy or ostentatious manner, not to ring their bells too loudly.⁴

The tradition attributed that pact to ‘Umar, a great general during the conquests and the second caliph, probably to grant authority to a status that took definite shape only slowly in the early Muslim centuries. It was during the eighth and ninth centuries C.E. that the Umayyad, then the Abbasid, caliphs and jurists defined and circumscribed the status of the dhimmi. By paying the jizya, the dhimmi marked his submission to Muslim authority and as a result enjoyed its protection. If he owned lands, he also paid the kharj, a property tax higher than the one the Muslim had to pay. At the same time, the dhimmi accepted his social inferiority. The theoretical restrictions were not uniformly respected, however. Far from it: many churches and synagogues were built in Muslim countries; the clothing prohibitions were applied very unevenly, and a number of Christians and Jews occupied positions of authority in the entourage of princes. There were times of tension, even persecution: the most notorious example was the reign of the Fatimid caliph al-Hkim (996–1021), who imposed on the Jews and Christians the wearing of distinctive clothing, prohibited them from drinking wine and from holding their processions and public feast days, and had many churches and synagogues razed.⁵ But that policy was an aberration, and Christians and Jews were soon allowed to rebuild their places of worship and to practice their religions as before. The taxes could nevertheless be burdensome, especially on the most destitute. In Fatimid Egypt, for example, a Cairo artisan had to pay a jizya of about 1 2/3 dinars, the equivalent of two weeks (twelve days) of his salary, an altogether acceptable amount; by contrast, for a worker (for example, a peddler), the same sum represented twenty-two weeks (132 days) of work.⁶

In Europe, it was primarily the Christians and Jews of Sicily and the Iberian Peninsula who were under Muslim domination, beginning in the eighth century. The sources describing the conquest of Spain are all of late date, but we possess a curious document, a pact of surrender, dating to 713, between Theodomir (Tudmir in Arabic), a Visigoth lord of large territories in the southeast of the peninsula (in the region of the present-day city of Murcia), and ‘Abd al-Aziz, governor of al-Andalus and son of the conqueror Musa ibn Nusayr. Tudmir surrendered the cities of the region to ‘Abd al-Aziz and promised to pay a tribute in kind. In return, the governor recognized the lordship of Tudmir and guaranteed his security and that of his subjects, the enjoyment of their personal property, and their freedom to practice Christianity.⁷

Historians call the Christians of al-Andalus “Mozarabs,” a word that may be derived from the Arabic must‘arib, meaning “Arabized.”⁸ Throughout the nineteenth and twentieth centuries, historians inquired at length about them: How many were there at different times in the history of al-Andalus? How many had converted to Islam (and when)? Where and until when had their communities survived? The debate has sometimes been bitter because it is ideologically charged. For some nineteenth- and early twentieth-century Spanish historians, the Mozarabs represented the “true” Spaniards, forcibly subjected by the Muslim “outsider.” The existence of the Mozarabs made it possible to justify the war as a reconquest (reconquista) waged by northern Christians to “liberate” their coreligionists from the yoke of Islam. For other historians, the near total disappearance of the Mozarabs before the thirteenth century demonstrated the widespread Arabization and Islamization of the peninsula; the invasion of men from the north was not a reconquest but quite simply a conquest. The lack of documentation has played a large role in the virulence of the debate, since historians are obliged to speculate.

What is certain is that the Mozarabs represented almost the totality of the population during the Muslim invasions of the eighth century, that they remained the majority for the rest of the eighth century, and that, by contrast, they were almost nonexistent by the mid-thirteenth century. If we are to believe Mikel de Epalza, their decline was rapid, less because of individual and voluntary conversions than for lack of ecclesiastical structures. In the absence of bishops and priests, the inhabitants of the rural zones of the peninsula were deprived of the essential sacraments of Christianity, especially baptism. Within the space of a few generations, they could no longer remain Christians and were considered Muslims.⁹ The situation was different in large cities such as Toledo, Merida, Seville, and especially Cordova: there, the Umayyad authority maintained privileged relationships with the bishops and other prelates, often important figures in the court of the emirs (and later of the caliphs). For Christians, the presence of these prelates at court symbolized the Muslim authority’s acceptance and reflected the universal power of the caliphs of Cordova (mirroring that of their predecessors in Damascus). A heavy tax burden fell on the dhimmis: it has been estimated that, in the mid-eighth century, a protected person had to pay the state about three and a half times what a Muslim owed.¹⁰ That burden helps explain the reactions of Christians, many of whom converted to Islam, emigrated to the Christian kingdoms to the north, or joined revolts against the Umayyad authority within Andalusian society.

Religious differences were only one factor in a society riven by ethnic and regional divisions—between southern Arabs, northern Arabs, Berbers, and muwalladun (autochthonous peoples who had converted to Islam). The emirs attempted to deal with the revolts produced by these divisions, even as they manipulated them to prevent united opposition to their power. They therefore cultivated personal relationships with each community, including the Christians. For a long time, Muslims and Christians were thrown together even in their main place of worship: they shared the Cathedral of Cordova until ‘Abd al-Rahmn I (756–788), judging the place too cramped, purchased the building from the Christians and allowed them to construct churches in the new neighborhoods of the capital.¹¹ That emir and his successors named a Christian “count” (comes in Latin; kumis in Arabic), an intermediary between the Christians and the sovereign, responsible for taxation and the justice system in the Christian community.

Christian notables had a presence at the caliphal court of ‘Abd al-Rahmn in the tenth century: the caliph confirmed the nomination of bishops; and Christians served in the Umayyad administration, where they played an important role as ambassadors and translators in negotiations between Cordova and the Christian princes on both sides of the Pyrenees. The best-known example is no doubt that of Reccemundus, or Rab b. Zayd: as indicated by his two different names, he, like a number of Mozarabs in the tenth century, lived between the Latin and Arab worlds. ‘Abd al-Rahmn III sent him as an ambassador to both the Byzantine and the Germanic emperors. For his trouble, the envoy received the bishopric of Elvira from the caliph. It was apparently he who compiled the Calendar of Cordova in a bilingual (Latin and Arabic) version, dedicating it in 961 to the new caliph, al-Hakam II. But the Mozarabs of the caliphal period generally left few traces in the documentation or among the chroniclers.

For the period of the taifas (1031–1090), the information about the Mozarabs is even rarer. The Christians who remained were increasingly Arabized: essential Christian texts were translated into Arabic for readers who no longer knew Latin. Those playing an important role in diplomacy or politics became rarer; with the disappearance of the caliphate, it seems, no emir felt the need to surround himself with representatives of the Christian community, whose political importance was minimal. By contrast, the presence of Christians within the taifas seems in general not to have provoked any anxiety. No one feared they might form an alliance with the harbs (non-Muslim residents of the dr al-harb) to the north, who were becoming increasingly aggressive. This is particularly surprising when we realize that the Jews in certain taifas were sometimes accused of destabilizing the power structure, as was the case in Granada, where they were the victims of a massacre in 1066.

Under the Almoravids (1090–1147), the situation of the dhimmis on the peninsula grew grimmer. For Ysuf b. Tshfn and his followers, one of the fatal flaws of the petty kings of the taifas was precisely their lack of steadfastness in their relations with the Christians, dhimmis and harbs. There could now be no question of making peace with the harbs, much less of their paying parias. As for the dhimmis, it was necessary to limit and scale back their role in Andalusian society and minimize their contacts with Muslims, while still respecting the rights that the Sharia granted them. Ibn ‘Abdun’s manual of hisba (urban law) reflected that new state of affairs: it specifies that no Muslim ought to do “lowly” tasks for a Jew or Christian—take care of his animals, dispose of his garbage, clean his latrines, and so on. It was the dhimmi who was to execute these tasks, which corresponded to his inferior status. Christians were morally inferior as well, according to Ibn ‘Abdun: he advises prohibiting Christian women from going into churches except on days when mass is held, since it is well known, he says, that they go there to fornicate with the priests.¹²

In 1099, Ysuf b. Tshfn, on the advice of his ulemas, had the main church of Granada razed. Later, the Christians of the city appealed to the Aragonese king Alfonso I, who conducted a campaign of raids in Andalusia in 1125–1126, bringing back a good number of Mozarabs with him to Aragon. The role played in that affair by the Christians of Granada led to the deportation of a fair number of Mozarabs to Morocco, where it would have been difficult for them to conspire with their northern coreligionists and where they could perform the function of collectors of non-Qur’anic taxes. Other Christians and Jews did not wait for these expulsions to leave al-Andalus: some departed for other, more tolerant Muslim countries. The Jewish philosopher Maimonides, for example, settled in Cairo. Others fled to Christian Spain, increasing the Jewish and Mozarab population in border cities such as Toledo. The repression of non-Muslims by the Almohads led to further mass departures of Christians and Jews. At the time of the taking of the chief Andalusian cities by Christian kings in the thirteenth century, there were almost no dhimmis left. The inhabitants of the Nasrid emirate of Granada were almost exclusively Muslim.

In Sicily, almost half the population remained Christian until the end of the Muslim period. Some Christian communities on the island remained independent until the tenth century, others simply paid a tribute to the Muslim rulers, while still others were under the authority of the Muslims as dhimmis. These dhimmis had nearly the same status as elsewhere in the Muslim world, including al-Andalus. But whereas the Visigoth kingdom had completely collapsed in Spain, the Christian Sicilians still maintained ties (religious, cultural, and sometimes political) with Constantinople. Those who rebelled against the Muslim authority placed their hopes in the Byzantine Empire, and various emperors tried in vain to reconquer the island.¹³

Many legal texts, especially fatwas (legal opinions) and manuals of hisba, deal with the everyday contacts between Muslims and dhimmis. Let us consider a few concrete examples of the problems raised by the coexistence of Muslims and dhimmis: sexuality and marriage, food (especially meat and wine), and social hierarchies.

From the early days of Islam, the laws governing marriage were relatively clear: a Muslim man could marry a Jewish or Christian woman, though certain jurists affirmed that marriage to a Muslim woman was far preferable. In any event, the children of a Muslim father were Muslims. A Muslim man could also have sexual relations with his female slaves, whether or not they were Muslim. By contrast, a Muslim woman could marry only a Muslim man, since a female believer was not to be placed in a position of inferiority in relation to a dhimmi.¹⁴

Food raised further problems. Islam imposed a whole series of rites on the slaughter of animals: it was necessary to cut the trachea, the esophagus, and the two veins in the neck while reciting the tasmiya, an invocation of God’s name. At the same time, the vast majority of the jurists believed that it was permissible for a Muslim to purchase and consume meat slaughtered by the dhimmis. The Jews had slaughtering methods similar to those of the Muslims, but that was not at all the case among the Christians. The meat of Christians sometimes caused concern: some jurists prohibited Muslims from consuming it if they knew an invocation to Jesus had been uttered during the slaughter.¹⁵ Ibn ‘Abd al-Ra’f, an Almoravid jurist, acknowledged that the purchase of meat prepared by dhimmis was legitimate, but he strongly discouraged it, going so far as to declare that someone who purchased it was a “bad Muslim.” Who knows whether the meat had been consecrated “for their churches, or in the name of the Messiah or the Cross, or for some other reason of the same kind?” He concluded that it was better to abstain.¹⁶

Wine could be a troublesome subject. Although at certain times and in certain parts of the dr al-islm, Muslims readily drank wine,¹⁷ this practice often provoked the wrath of jurists. A mufti from Cordova in the first half of the eighth century declared that the house of every wine merchant should be burned down.¹⁸ In the twelfth century, Ibn ‘Abdun complained that Cordovan Muslims were crossing the Guadalquivir in boats at night, to go to the Christian neighborhood and buy wine. Ibn ‘Abd al-Raf recommended harsh punishments for the Muslim who drank wine and for the Christian who sold it to him, but also for the overzealous Muslim who tried to prevent the Christian from consuming it.¹⁹

The jurists endeavored to impose respect for social hierarchies, which relegated the dhimmis to an inferior place, and to discourage certain types of relations between Muslims and dhimmis. For example, a tenth-century Cordovan mufti railed against Muslims who participated in Christmas festivities and exchanged gifts with Christians, or who joined Christians in celebrating New Year’s or the winter and summer solstices.²⁰ It is likely that his fulminations were wasted effort and that the practices he denounced were widespread. Another mufti of the same period prohibited the Muslims from teaching the Qur’an to Christian children.²¹ A Cordovan mufti from the late ninth century declared that a dhimmi man who raped a Muslim woman would receive the death penalty, but if he embraced Islam in extremis, he could be pardoned, provided he paid the rape victim a dowry proportionate to her social status. Conversely, if his was a false conversion, he would be crucified.²²

Large Christian populations could be found in Sicily until the end of the Muslim period. In Spain, by contrast, as in the nearby Maghreb, Christianity tended to disappear under the Almoravid and Almohad dynasties, as a result of conversions or emigration. At the same time, more and more European Muslims found themselves under the yoke of Christian princes.

MINORITY MUSLIMS IN CHRISTIAN STATES: LAW AND PRACTICE

In the eleventh and twelfth centuries, a large number of Muslims fell under Christian domination during the Norman conquest of Sicily (1072–1092), the First Crusade in the East (1098–1099), and the conquests by Christian princes of the Iberian Peninsula—the Castilian-Leonese conquest of Toledo (1085) and the Aragonese takeover of Huesca (1096) and Saragossa (1118). Later, large Muslim populations came under Christian authority following the major peninsular conquests of the thirteenth century, when James I of Aragon took Majorca (1229) and Valencia (1238), and Fernando III of Castile captured Cordova (1236) and Seville (1248). Although every conquest resulted in a different situation, most often the Christian princes granted the defeated a status similar to that of the dhimmis in Islamic territory, with the same legal and religious guarantees and the same fiscal and social constraints.²³

At the time of the Norman conquest, Sicily had nearly 250,000 Muslims, slightly more than half the population, the rest of it composed primarily of Greek-speaking Christians and a few Jews.²⁴ During the conquest and in the years that followed, many Muslims left the island to return to the dr al-islm. It was principally merchants and the wealthy who could emigrate: peasants, whose minimal wealth was primarily in real property, had a hard time leaving their lands. The Norman military aristocracy imposed a feudal system on the majority Muslim peasantry. Count Roger I divided up the island into fiefs, which were distributed (along with the villani, dependent Muslim peasants) to his Norman and Italian vassals. Other Muslim communities possessed greater rights over their lands and paid only an annual tribute to the royal authority, but their semi-independent status deteriorated throughout the twelfth century. The Norman kings had no interest in seeing the Muslims of the island convert to Christianity. They were content to reproduce the system of dhimmis while reversing the roles: it was now the Muslims and Jews who had to pay the jizya (the Arabic word was retained by the Norman administration). Muslim peasants worked the land for their new masters and paid a royalty twice a year. Other Muslims enjoyed a more advantageous status, especially the residents of Palermo and other cities that had bowed to Norman authority relatively early on and had thus been able to negotiate certain rights, such as exemption from the jizya. They gradually lost these rights, however.

Although the Norman lords had an interest in protecting their Muslim peasants, some immigrants were less favorable toward them. Lombard peasants were sent to settle the eastern part of the island, a sparsely populated zone that needed a labor force; and it was these immigrants who massacred the Muslim peasants in the surrounding villages in 1160–1161, obliging the survivors to take refuge in the western parts of the island. On the island as a whole, the Muslims’ traditional farming method, an intensive polyculture, gradually gave way to cereal monoculture, no doubt because wheat was easier to sell on the Italian and Ifriqiyan markets, but also because of the decline and exodus of the Muslim peasantry. In the cities, Muslim artisans and merchants, essential to economic activity, increasingly found themselves in competition with immigrants—Lombards, Genoese, Pisans, Catalans, and others—who had every interest in seeing their rivals lose their privileges. In that extremely divided society, where the alliances between different parties were multiple and shifting, religious differences did not account for everything, though they remained important. The Sicilian Muslims’ increasingly precarious status owed less to the rise of religious intolerance than to a game of shifting and complex interests in which Muslims more often than not lost ground.

One of the most fascinating accounts about the Muslims of Norman Sicily was penned by the Andalusian traveler Ibn Jubayr, who was shipwrecked off the coast of the island in December 1184 and who remained there until March 1185.²⁵ In his Rihla, or travel narrative, he painted a very mixed picture of the state of Sicilian Islam: he was amazed at the survival and piety of the Muslims he met on the island but also worried about the difficulties they faced. In theory, a Muslim, according to the Sharia, should not live under the jurisdiction of an infidel but should emigrate to a Muslim country if he can. Some jurists acknowledged extenuating circumstances, however: the mufti al-Mzar (d. 1141) set forth various reasons that might justify a Muslim living in Norman Sicily.²⁶ Ibn Jubayr agreed entirely, going so far as to say that the eunuchs in the royal palace of William I, obliged to conceal their religion, were in a state of continual jihad, since they were working for their faith and the well-being of their Muslim brothers. He was delighted to see the importance the monarch granted to his Muslim advisers. But he felt there was also a danger there: Could the kindness of the Christians of Palermo and the beauty of the Christian women be a lure? Charmed by the Christmas mass sung in the Church of the Martorana, dazzled by the beauty of the light coming through the stained-glass windows, Ibn Jubayr confesses that everything “provoked a conflict in my soul, from which I seek refuge in God.” Children threatened their parents with converting to Christianity, and far-sighted parents married their daughters to Muslim travelers, so that the women could live in Islamic countries.

For many reasons, Muslims were on their way to losing their status in 1184–1185: intrigues at court pitted Muslim q’ids against Christian counts, and Lombard peasants rioted against the Muslims on the eastern part of the island. Then there was the geopolitical climate. So long as the Normans were conquering the African coast (as they had done during the first half of the twelfth century), the Sicilian Muslims did not constitute a threat; but when the Almohads embarked on the reconquest, and especially after they took Mahdiyya in 1160, the loyalty of the Muslim subjects began to be called into question. In Palermo, their weapons were taken from them, exposing them to even more violence.

But the final blow to the Sicilian Muslim community would be dealt by a prince whose enemies often presented him as an Islamophile: Frederick II Hohenstaufen, king of Sicily and German emperor. In the struggles that marked his contested succession, many Muslims on the island allied themselves with German rebels. Frederick crushed these revolts, not without difficulty, between 1221 and 1224, then decided to put an end to Muslim Sicily. He had about sixteen thousand rebels transported to Lucera in Puglia, where he established an exclusively Muslim colony. Some Muslim communities persisted on the island, but they were small and few in number, and they have left almost no traces in the documentation. All that is mentioned is the embassy sent by al-Kmil, sultan of Egypt, asking Frederick to leave the Sicilian Muslims in peace or at least to let them go to Egypt. The founding of Lucera served a dual objective. Far from their lands and coreligionists, the Muslims could no longer revolt; they were dependent on the emperor’s goodwill. In addition, by installing them in Italy, Frederick could use them to assert his power against the Italian barons. In the years that followed, Frederick authorized Dominican missionaries to preach to the city’s Muslims, who, it was said, spoke good Italian. It was only when popes Gregory IX and Innocent IV were seeking polemical arguments against Frederick (and against his son Manfred) that they presented Lucera as evidence of the emperor’s tepid faith, his penchant for Saracen culture and women, and even for the Muslim religion. When Charles I of Anjou, at the pope’s instigation, defeated Manfred and took the crown of Sicily, he promised to liquidate Lucera, and his son Charles II did so in 1300. Muslims who did not agree to be baptized were sold as slaves.²⁷

In the East following the First Crusade, the Europeans imposed their suzerainty on their Eastern subjects: Melkite, Syriac, or Jacobite Christians as well as Jews and Muslims.²⁸ There are few data that permit us to estimate the proportions of these different groups. It seems, however, that Eastern Christians lived primarily around Jerusalem, while most of the Muslims lived in the other, rural regions of the kingdom; Europeans probably did not represent more than a quarter of the population. The villages retained their own structures of government: the notables of the village governed it under the leadership of the village chief, or ra’is, who meted out justice. Often the village was granted to a member of the Frankish nobility, or in some cases to an ecclesiastical institution. In the East and in Sicily, the feudal system was superimposed on autochthonous village structures. The Eastern peasant (whether Christian or Muslim) generally paid a percentage of his harvest, roughly the equivalent of the kharj that the dhimmi peasant had to pay under Muslim authority. His status was often equivalent to that of a serf in the West, though with a few differences: for example, he had almost no corvées, since demesne lands were almost nonexistent. Ibn Jubayr, who traveled in the region of Acre, believed that the Muslim peasants under Frankish jurisdiction were actually less exploited than their brothers in Islamic countries. That was likely not the result of “tolerance” on the part of their masters, but an indication that these masters needed to hold on to their labor force.²⁹

In Spain, every conquest led to large-scale emigration to territories still under Muslim control, especially Granada and the Maghreb, in accordance with the prescriptions of the Sharia, which discouraged the Muslim from living under the yoke of the infidel. But a good number of Muslims remained, and the kings did their best to encourage them, sometimes going so far as to establish new Muslim settlers in underpopulated regions (such as the island of Minorca). It would be difficult to draw general conclusions about the size of these Mudejar communities (as they were called in Christian Spain), which varied enormously from one region to another. In Castile and León in the eleventh and twelfth centuries, Muslim populations were often expelled from the conquered cities.³⁰ Toledo was an exception: Alfonso VI is said to have allowed the Muslims who wished to remain there after the conquest of 1085 to do so, but the vast majority emigrated.³¹ In the Ebro Valley, by contrast, a good number of Muslim peasants stayed after the conquest of the region (Alfonso I of Aragon captured Saragossa in 1118). These Muslims actively participated in the local economy, selling and buying lands and other property; their new Aragonese lords were usually content to live off their seigneurial revenues. Hence the Aragonese saying: “Quien no tiene Moro no tiene oro” (he who has no Moor has no gold).³² During the major Castilian conquests of the thirteenth century, the Muslims from the cities who offered strong resistance were generally expelled, whereas those who negotiated their surrender were granted guarantees allowing them to remain. Some Muslim princes who accepted the suzerainty of the king of Castile had their titles and powers confirmed.³³

In Catalonia, few Muslims remained after the Christian conquest, except in the city of Lérida, where there was a large community until the sixteenth century. The major expeditions of King James I of Aragon added a large number of Muslim subjects to the Crown. During the conquest of Majorca (1229), a good part of the Muslim population, especially the social and economic elite, left the island; only a “headless” Muslim peasantry remained. The Muslims on the neighboring island of Minorca mounted such fierce resistance that the king, when he finally took the island in 1235, reduced the entire population to slavery. In the kingdom of Valencia, a significant Muslim population remained; in many surrender treaties, James guaranteed legal autonomy and religious freedom to the aljamas who recognized his sovereignty.³⁴

There was always the risk that these Muslims would form alliances with potential invaders: Mudejars allied themselves with various Muslim princes from Granada and the Maghreb and revolted against Christian authority. That was especially the case during an uprising orchestrated by two Muslim vassals of Alfonso X of Castile and León, Ibn al-Ahmar, emir of Granada, and Ibn Hud, emir of Murcia, in 1264–1265. (Rebel Christian vassals participated as well.) The Muslim populations revolted in several Andalusian cities, proclaiming their allegiance to Ibn al-Ahmar. They were aided by a contingent of three thousand Moroccan warriors. With some difficulty, Alfonso X succeeded in reasserting his authority and then in expelling the Muslim populations from certain hotbeds of rebellion, especially Jerez and Cadiz. In the region of Denia, which was now part of the kingdom of Valencia, al-Azraq led a major uprising against James I in 1247–1248. James managed to regain control and drove out many Muslims. In 1276, some Mudejars of Valencia revolted to such an extent that the king resolved to expel a good number of them. But his son Peter III of Aragon did not implement that decision.

In the twelfth and especially the thirteenth century, a large number of legal texts came to define the status of Muslims in Christian territory: treaties of capitulation, municipal or royal fueros, ecclesiastical councils. These documents demonstrate that the Muslims from Christian kingdoms could be slaves, free peasants, artisans, or mercenaries in the royal armies. The Muslims’ right to practice their faith was generally guaranteed. Religious conversions had to be voluntary—and only to Christianity, of course. The laws tried to maintain a certain level of segregation: marriage and sexual relations between Muslims and Christians were forbidden, public baths were not to accept Muslims and Christians at the same time, and so on. In theory, the Mudejar was supposed to be socially inferior to the Christian, just as the dhimmi in Islamic countries was inferior to the Muslim. Since the laws concerning the Muslims in Christian lands have been the object of many studies,³⁵ I shall simply give a few examples, by way of comparison, of the legal provisions introduced to define and circumscribe the place of Muslims in Christian societies.

First of all, the Muslims, like the Jews, were granted the right to practice their religion and to have places of worship. Alfonso X, for example, declared that the Moors could live “observing their law without insulting our own.” Their mosques were royal property; the king could therefore do with them as he pleased. Implicitly, that provision entailed the possibility of turning them into churches, or conversely, of setting some aside to continue to serve as mosques.³⁶ This tolerance tended to erode over time. A good example is the right to perform the adhan, the call to prayer issued by the muezzins, which was often among the concessions granted. In 1311, the Council of Vienne barred the adhan in any Christian territory. But that prohibition was respected in the breach: in Valencia, various fourteenth- and fifteenth-century kings and lords granted exceptions to that ban or declined to enforce it, sometimes drawing the wrath of church authorities.³⁷

Muslim subjects were kept in check by specific legal institutions. In the Latin kingdom of Jerusalem, all Eastern subjects (Christians, Jews, and Muslims) fell under the jurisdiction of the Court of the Fronde, composed of two Frankish magistrates and four Eastern Christians—but no Muslims or Jews.³⁸ Matters of justice within the Muslim community were normally left to the Muslim procurator, the qadi (alcalde in Castilian, alcaide in Portuguese). Some Christian sovereigns guaranteed their Muslim subjects the right to elect their own qadi. Others preferred to name him themselves, but it is likely that, in such cases, the sovereign’s choice was made through negotiations with his Muslim subjects.³⁹

In legal matters involving both Muslims and Christians, it was the Christian justice system (municipal, royal, or other) that prevailed. A Muslim witness sometimes had to swear on the Qur’an, just as the Jew swore on the Torah and the Christian on the Gospels. The Siete partidas of Alfonso X of Castile established a precise and elaborate rite: the oath had to be taken at the door of the mosque; the Muslim witness had to swear in the name of Muhammad and his law and had to declare that, if his testimony was not truthful, he agreed to be deprived of all the goods belonging to Muhammad and the prophets and to suffer all the punishments that the Qur’an destined for the infidels.⁴⁰ The court placed more faith in Christian witnesses than in Muslims or Jews: a Muslim could not testify against a Christian, except in cases of treason.⁴¹ Penalties and fines often reflected the Muslim’s inferior status. The Leyes de estilo, compiled in Castile in the early fourteenth century, stipulated that the fine for the murder of a Moor should follow local custom but that it had to be lower than that levied for the murder of a Christian.⁴²

The legislation relating to Muslim minorities was derived from the traditional laws that limited the place of Jews in Christian society: in accordance with canon law, Jews could not have the slightest power over Christians. In particular, they were not allowed to own Christian slaves or to hold public office. Later laws extended these principles to the Muslims. The Third Lateran Council (1179) prohibited the Jews and Saracens from owning Christian slaves, a prohibition often repeated in royal legislation (for example, in the Siete partidas).⁴³ Various fueros in Iberian cities prohibited the Jews and Muslims from being judges in cases relating to Christians.⁴⁴

That legal inferiority did not always translate into true social inferiority. Muslims and Jews could in fact be found at every level of society. Mudejar contingents from Valencia played an important role in the army of the Aragonese kings: during the French invasion of Catalonia in 1285, for example, six hundred Valencian Mudejars, crossbowmen in particular, took part in the defense of Girona. Many Jewish and Muslim doctors were also in the service of princes and commoners. They sometimes provoked jealousy or distrust: William of Tyre complained that the wives of Frankish nobles preferred Jewish or Saracen doctors, and his translator added that these doctors were poisoning the grandees of the kingdom.⁴⁵

Many laws were aimed at prohibiting all sexual relationships between Christians and Muslims. Interfaith marriage was outlawed, except in cases where an already-married Muslim or Jew converted to Christianity. Such a person had the right to remain married to a non-Christian spouse, according to Gratian’s Decretum, and Pope Gregory IX confirmed that right in 1234.⁴⁶ In Christian Spain generally, the Christian woman and Muslim man who had sexual relations faced great risks, but that was not the case for the Muslim woman and her Christian lover. The Fuero de Sepúlveda decrees that a Muslim man who sleeps with a Christian woman will be thrown from a cliff and his lover burned; in the Fuero de Béjar, they are both to be burned. The Siete partidas is slightly more lenient toward the Christian woman: the Muslim or Jewish lover is to be stoned, while his accomplice loses half her possessions. If she is married, she faces the death penalty; if a prostitute, the two lovers are whipped together throughout the city. In all cases, the penalties for repeat offenders are harsher.⁴⁷ By contrast, no mention is made of sexual relations between a Christian man and a Muslim or Jewish woman, which seem to have been tacitly accepted. To judge by municipal jurisprudence, especially the fueros granted to Iberian cities by Christian sovereigns, it appears that sexual relations between Christian men and Muslim slave girls were commonplace. In the Crown of Aragon of the fourteenth century, a Christian prostitute ran the risk of being burned alive if she slept with a Jewish or Muslim man, and the Muslim authorities often demanded the death penalty for a Muslim woman who slept with a Jewish or Christian man. Nevertheless, the woman could escape punishment either by converting to Christianity or by becoming a slave—often of her Christian lover.⁴⁸ In law as in literature, sexual conquest became the metaphor for conquest generally. The pretty Muslim woman was not only a literary topos, she could also be found in the bed of many a Christian knight. A few authors did criticize or discourage these practices: The Castigos e documentos para bien vivir ordenados por el rey don Sancho IV tried to persuade the future king Fernando IV of Castile that having sexual relations with a Moorish woman was like sleeping with a dog, since she followed the irrational law of Muhammad.⁴⁹ Such views, one suspects, had little influence on the sexual practices of the prince and his entourage.

Contact with the religious adversary was often seen as an element of corruption or pollution to be avoided. Some fueros banned non-Christians from going to the public baths on the same days as Christians.⁵⁰ Christian wet nurses were prohibited from breastfeeding Jewish or Muslim children, Christians from using Muslim or Jewish wet nurses for their offspring.⁵¹ It was also in an attempt to better enforce the sexual prohibitions that the law imposed (or tried to impose) clothing restrictions. As early as 1120, the Council of Nablus followed its many sexual prohibitions with a ban on Muslims dressing as “Franks.”⁵² These clothing restrictions were supposed to help Christians identify Muslims and thus avoid any needless contact with them. In the same spirit, in 1215 the Fourth Lateran Council decreed that Saracens and Jews were to wear distinctive garb to prevent sexual relations with Christians, or rather, to prevent Christians from pleading ignorance to justify their affairs with non-Christians. These measures were in principle to be imposed on all Christendom, but they were unevenly applied. Sumptuary laws requiring distinctive signs for Muslims and prohibiting them from wearing “Christian” clothing can in fact be found in various eras: at the Cortes of Seville in 1252, at that of Valladolid in 1258, and again at that of Seville in 1261, proof that the measure taken by the council of 1215 was not respected to any great extent.⁵³

Not only sexual corruption but also spiritual corruption provoked fear. Innocent III and the Fourth Lateran Council strove to defend the Christians from the mockeries and blasphemies of the “infidels.” In order to protect the Holy Week rituals from such contamination, the council did not hesitate to ban Muslims and Jews from public places during that period, as legislation in Spain would later do.⁵⁴ Muslims and Jews were lumped together because of their supposed hostility toward the Christians. Both were “blasphemers,” according to the council, which claimed that members of these two groups paraded during Holy Week in gaudy clothes, making fun of the Christians who were ritually expressing their sorrow at the Passion of Christ. That hostility was specifically invoked to justify the ban on their holding public office: a “blasphemer” could not be granted the slightest power over a Christian. A polemical view of Islam fed these decisions of the Fourth Lateran Council: the council did not enumerate or distinguish the different “blasphemies” of the Muslims or Jews, but it declared that they were sufficient to justify their exclusion from any position of authority.

Alfonso X devoted one of the sections (títulos) of the seventh Partida to those who insulted “God, Mary, and the other saints.”⁵⁵ The last of the six chapters of the título has to do with the Jews and Moors who utter such insults. It notes that the Jews and Moors were allowed to live in “our country” even though they did not share “our faith,” but that this permission was granted them only if they did nothing to insult Christ, his mother, or the other saints. In addition to the ban on verbal insults, it was forbidden to spit on crosses, altars, and images of saints; to strike a holy object with the hand, foot, or another object; and to throw stones at churches.⁵⁶

The problem of conversion recurs often in legal documents. Alfonso the Wise made it the main subject of título 7.25 of the Siete partidas, called “De los moros”: seven of the ten laws are devoted to it. Five laws have to do with the punishments to be inflicted on the Christian who converts to Islam. The apostate shall lose all his possessions, which become the property of his heirs who have remained Christian; he can be accused of that crime until five years after his death. If an apostate returns to Christianity, he still loses the right to be a public official or a witness, and to enter into sales or purchasing contracts. The political and military context of thirteenth-century Castile instilled a fear of converts to Islam. It was a very real danger: conversions often occurred during captivity in Islamic territory or accompanied an act of treason.⁵⁷ By contrast, the conversion of a Muslim to Christianity was desirable, according to the Siete partidas, but always had to be voluntary: Christians had to try to persuade Muslims by reason and example, not by violence or constraint. No one had the right to prevent a Muslim from converting to Christianity, or to call the convert a tornadizo (renegade or traitor), or to insult him. According to Alfonso, it was the fear of being the object of such insults, as well as force of habit, that would keep Muslims from converting. Kings James I and Peter of Aragon decreed similar laws to protect converts from insults and the loss of their inheritance.

Conversion to the dominant religion was often motivated by social considerations. That was especially the case for slaves: baptism resulted in liberation, as specified, for example, in the Siete partidas or in the Furs of Valencia. But slave owners were naturally opposed to that principle and tried to prevent the evangelization of their slaves: Jacques de Vitry, bishop of Acre in the thirteenth century, complained that the Christians of the Holy Land did not allow him to preach to their slaves. Pope Gregory IX ultimately decreed that the conversion of a slave would not result in liberation, hoping that masters would thereby allow slaves wishing to convert to do so.⁵⁸

CAPTIVES AND SLAVES

In the Middle Ages, slavery was an integral part of Mediterranean societies, both Muslim and Christian. Slaves were often captives taken during a siege, sailors and passengers of a commandeered ship, or peasants from coastal regions rounded up by corsairs. The captive faced one of three fates. If he had wealthy or influential relatives, he was ransomed: from the point of view of his abductors, that was the most profitable operation. He could also be used as currency of exchange to secure the liberation of captives held by the adversary. Or the abductor could reduce him to slavery, either to profit directly from his services or to sell him on the international market in the extensive slave trade.

In the wars between the caliphal and the Byzantine armies, the many prisoners taken by each side had first and foremost a propaganda value: they were paraded about, displayed in public, because they embodied the victory over the infidel enemy. In about 900, for example, the emperor of Constantinople held a banquet for his Muslim prisoners on Easter: his “guests” were solemnly reassured that they would be served no pork. He also had a mosque built for the prisoners and for visiting Muslim diplomats.⁵⁹ These captives were later exchanged for Byzantines held by the Muslims. The Qur’an stipulates the fate of prisoners of war taken by the Muslims: “Grant them their freedom or take ransom (al-fid’) from them, until War shall lay down her burdens” (the term al-fid’ would later be used to designate both ransom and exchange).⁶⁰ In addition, the redemption of Muslim prisoners was considered an obligation and was one of the authorized uses for the zakat, alms that had become an obligatory tax for every Muslim. Many exchanges of captives between the Byzantine emperors and the Abbasid caliphs took place between 769 and 969.⁶¹

In the Holy Land in the age of the Crusades, many were taken captive on both sides. ‘Imd al-Dn, a chronicler in Saladin’s entourage, claimed that Saladin had redeemed twenty thousand Muslims from captivity among the Franks: an exaggerated figure no doubt, but other sources on both sides confirm that many were freed. In 1263, the Mamluk sultan Baybars is reported to have proposed a large exchange of captives to the Franks of Acre; the military orders refused, however, because their Muslim slaves were talented artisans whom it would have been impossible to replace.⁶² When Charles II of Anjou reduced the Muslims of Lucera to slavery in 1300, about ten thousand were sold on the markets of Naples, Bari, and other cities of the kingdom.

The owner of a perfume shop successfully purchased five hundred slaves in Spain after the Muslim conquest and brought them back to Syria, if we are to believe a chronicle written three centuries after the events. The same text speaks of a Visigoth slave purchased in Medina for a handful of pepper.⁶³ The summer raids by Umayyad troops into northern Spain in the ninth and tenth centuries often procured large quantities of slaves.⁶⁴ Sometimes the northern kings responded in kind, as in summer 913, when Ordoño II of León attacked Alentejo and took four thousand women and children captive.⁶⁵ Almohad chroniclers claim (exaggerating no doubt) that Caliph Ab Ysuf Ya’qb al-Mansr took as many as twenty-five thousand captives during his victory at Alarcos in 1195. Other sources speak of five thousand captives, reportedly exchanged for five thousand Muslim captives.⁶⁶ During the great conquests of the Spanish Christian kings (such as the taking of Majorca in 1229, of Minorca in 1235, or of Granada in 1492), captives were plentiful on the markets of Iberian and Italian cities.

More common were captures during actions on a smaller scale: small raids on border regions, where attackers quickly seized objects of value (including cattle, women, and children); and acts of piracy, with pirates commandeering a ship or making a rapid incursion into a coastal zone. When Muslims of Spain conquered Crete in the ninth century, they used it primarily as a base for conducting raids against the Byzantine Empire, enriching themselves with the booty and ransoming captives or selling them on the slave markets. Greek corsairs did the same on the coasts of Syria, Palestine, and Egypt.⁶⁷

Some of these actions were illegal, which in theory prevented the perpetrators from selling the booty, including slaves. In the case of those captured during a war, the purchaser was assured that captives were “fair game” (de bona guerra) and that, as a result, the purchase did not risk being voided. Otherwise, the legality of the sale was not assured. In fourteenth-century Catalonia, for example, if someone who was purchased as a slave turned out to be a Christian prince’s Mudejar subject, unjustly deprived of his freedom, or the subject of a Muslim ally of the king of Aragon taken during an illegal raid, the buyer could be obliged to free the slave, often without financial compensation.⁶⁸ That may be part of the reason that in Genoa young Tatar or Greek women were preferred to Maghrebis; the Genoese controlled the trade, without being overly moved that the Greek women were Christians. Over time, the enslavement of Christians raised fewer and fewer objections, provided they could be considered heretics, schismatics, or simply rebels. Two popes (in 1294 and then in 1305) threatened to reduce their Italian political adversaries to slavery.⁶⁹

Raids represented humiliation for those who were reduced to slavery and for their families, fellow citizens, and coreligionists. Many initiatives were therefore taken to liberate the captives, usually in exchange for ransom. In the early tenth century, the patriarch of Constantinople sent a pound of gold to the residents of Amalfi, who had commercial and political relations with the Aghlabids of Ifriqiya, in order to redeem Christian captives.⁷⁰ Bequests (often in wills) have been found dating to late tenth-century Catalonia for the redemption of captives taken by al-Mansr in 985.⁷¹ In the twelfth century, some fueros established that Christians had a right to purchase Muslim slaves at the market price, in order to use them as ransom for Christian captives in Muslim hands.⁷² In 1182, King Alfonso VIII of Castile captured the castle of Santafila and took seven hundred prisoners; the Muslims of Seville redeemed them for twenty-seven hundred gold dinars.⁷³ In the documents, the redemption of captives sometimes looks like one commercial practice among others, profitable for the merchant involved. In eleventh-century Egypt, the ransom for a captive varied between twenty-four and one hundred gold dinars, depending on the circumstances. One hundred dinars for three captives seems to have been more or less the “standard” price⁷⁴—it was, for example, what Greek corsairs selling Muslim captives in the Palestinian ports asked in the tenth century.⁷⁵ That commerce benefited the abductors but also the retailers, such as the Amalfitans who purchased Egyptian Jewish captives from their Rm (Byzantine or Italian) captors, brought them back to Egypt, and then offered the Jewish community of Alexandria the chance to redeem them.⁷⁶ Many examples exist, such as a contract concluded in 1454 in Genoa, by the terms of which two captives, Ahmet Mazus and Mohammed Zamai, acknowledged that they were in the power of a certain Giovanni Raibaldi, “who purchased us from the hands of our enemies.” They pledged to pay Raibaldi’s brother in Tunis 161.5 gold doubloons and a half-length of cloth within twenty days after their return to North Africa. Similarly, Italian, Catalan, and Marseillais merchants redeemed Christian captives in North Africa, in exchange for the promise of payment after their return to Europe.⁷⁷ The redemption of captives became a profession: the fakkk, Hispanized to alfaqueque, collected funds, crossed the Mediterranean, redeemed Muslim or Christian captives, and brought them home. By the thirteenth century, various Christian kings of Spain had named alfaqueques who acted in the state’s behalf.⁷⁸

Family initiatives, the pursuit of profit, state interests, and religious motivations all play their part in the accounts of the liberation of captives.⁷⁹ Usma ibn Munqidh tells of his missions to purchase Muslim captives from the hands of the Franks.⁸⁰ Ibn Jubayr was moved by the sight of chained Muslim prisoners in the streets of Acre. He writes that the princes and other Muslims of the region made abundant gifts to free these captives.⁸¹ And it was not only their coreligionists whom people took it upon themselves to ransom: Ibn Taymiyya says he paid a ransom to the Mongols for the liberation not only of the Muslim prisoners but also of the dhimmis, who were “under the protection” of the Muslims.⁸² In Spain, the gifts offered for the liberation of captive Christians in Muslim territory sometimes went through the intermediary of the alfaqueques, but increasingly they passed through religious institutions. These could be military orders (like that of Santiago) or new orders created specifically to work for the liberation of the captives: the Order of the Holy Trinity (founded in 1198) and the Order of Our Lady of Mercy (1218).⁸³ Monks collected funds from the faithful and traveled to meet with Muslim princes in Spain and North Africa to redeem Christian prisoners, whom they brought back to Europe.

Slaves remained numerous in most Mediterranean societies of the Middle Ages. Slavery was fostered not only by raids and roundups but also by international trade or trafficking extending from the Mongol Empire to western Africa.⁸⁴ Slaves were omnipresent in the Byzantine Empire: in the workshops of Constantinople, in the homes of owners of large rural demesnes, in the mines, in the imperial palace.⁸⁵ In Muslim territories, many slaves were imported from the dr al-harb: Slavs (pagan or Christian) from northern Europe, mamluks (often pagan Turks) from the Asian steppes, blacks from eastern and western Africa. These slaves occupied a not insignificant place in Muslim societies. Slave women and young girls were destined to perform household tasks and sometimes to become the concubines of their masters. The men formed large military contingents within the caliph’s or sultan’s armies: from Iraq to al-Andalus, the power of various princes rested on the military strength of their Slavic, black, or mamluk contingents. The princes used them as counterweights to Arab or Berber forces. For example, the Saqliba (Slavs) played an important role in the armies of the caliphate of Cordova in the tenth century and in the palatine administration.⁸⁶ The strategy was dangerous, however, and sometimes gave rise to armed slave revolts, assassinations (a number of Cordovan leaders were killed in the first third of the eleventh century), and even coups d’état, such as the one that overthrew the Ayyubids of Egypt in favor of the Mamluks in 1250.

Other men were subjected to castration and became eunuchs. They were assigned to guard the princely harems and were given tasks of an administrative nature. The owners of the extensive domains in certain parts of Europe used large numbers of Muslim slaves. In Cyprus and Sicily, for example, the slaves cultivated sugarcane. The introduction of sugar to the kingdom of Valencia, and then to Portugal, encouraged the growth of agricultural slavery, which was already well established. These practices spread to the Portuguese islands of the Atlantic in the fifteenth century and to the American colonies in the sixteenth.⁸⁷ Other intensive farming, such as that of mulberries (for silkworms) also relied on slavery, especially in the region of Genoa.⁸⁸ But most of the slaves in medieval Europe were domestics in aristocratic and bourgeois houses, responsible for the cooking, the housework, and the care of children. Some served as wet nurses: a young nursing woman had a much greater value on the market. In Genoa and Venice, girls were purchased from the Maghreb, or more often (beginning in the thirteenth century), from the lands bordering the Black Sea. Slaves could be found even in modest quarters, in the shops of artisans of Italian or Portuguese cities, for example; a young apprentice or a free servant girl would have been much more costly.⁸⁹ In the marriage contracts of aristocratic Italian families, a slave girl was often part of the dowry provided by the bride’s parents. It is difficult to estimate the slave population, but there were reportedly 13,750 “Slavs” (Saqliba), slaves and freedmen, in Cordova in the tenth century,⁹⁰ and 1,225 slaves were counted in Barcelona in 1431. In mid-fifteenth-century Genoa, there may have been 3,000 slaves (the vast majority of them women and girls), representing about 3 percent of the city’s population as a whole but 10 percent of the female population under thirty.⁹¹

Although narrative sources on these slaves are few, normative texts are abundant, ranging from the Bible and the Qur’an to treatises on municipal laws, sales contracts, and acts of manumission. In Roman and Byzantine law, the child of a slave mother or any captive taken during a war was a slave. In theory, punitive slavery was also recognized, but that practice became increasingly less common in the Middle Ages. A master who wanted to marry his slave had to free her beforehand. Muslim law prohibited reducing a Muslim or a dhimmi to slavery; these prohibitions were generally respected, except in rare cases of those considered heterodox, sometimes judged unworthy of their freedom.⁹² The child born of a slave mother was considered the slave of its mother’s master, unless the master himself was the father, in which case the child was free and legitimate and the slave mother became umm al-walad (mother of children): she retained her slave condition, but her master no longer had the right to sell her, and she became free upon his death. The son of such a union had the same rights to inheritance as the sons of free wives, and it was not uncommon for the son of a slave to reach the highest echelons of power.⁹³ The mother of the grand caliph of Cordova, ‘Abd al-Rahmn III, was a Frankish captive, his paternal grandmother a Gascon.⁹⁴

Jews, Christians, and Muslims practiced slavery in the Middle Ages. It was broadly accepted, including by men of faith and religious institutions such as monasteries and military orders, which owned many slaves. In Byzantium, the monks and nuns of aristocratic origin kept the personal slaves in their service at the convent.⁹⁵ Sometimes religious scruples impelled masters to treat their slaves with indulgence or even to free them. According to various Hadith, Muhammad reminded the faithful that God, who gave them authority over their slaves, could also have reversed the roles, and that they should therefore be gentle and fair to their slaves. According to several Christian authors, slavery is a punishment for human imperfection that must be endured with humility and patience. Hagiography, moreover, often portrays the captivity of Byzantine Christians raided by Muslim corsairs as an ordeal allowing the saint to display his virtues, and sometimes to perform miracles leading to his liberation or the conversion of his abductors.⁹⁶ For many Jews, Christians, and Muslims, liberating one’s slaves was an act of piety. The Qur’an encourages the Muslim to liberate his war captives, either for ransom, in an exchange of prisoners, or freely, without compensation. This last initiative was for many jurists a good way to receive forgiveness for one’s misdeeds and be reconciled with God.⁹⁷ One Hadith declares that anyone who frees a Muslim slave will be liberated from the torments of hell; another says that the Muslim man who educates a young slave girl, frees her, then marries her honorably, will obtain a dual reward in paradise.⁹⁸ Byzantine hagiography mentions various cases of saints who freed their slaves in order to then devote themselves to a pious life. There it was less the institution of slavery than worldly possessions (human or other) that were an obstacle to a holy life.⁹⁹ In Constantinople and in Italian cities such as Genoa and Venice, many acts of manumission are recorded, either to allow a slave woman who has served her master well to enter into marriage with a free man, or to free slaves upon the master’s death (though often the will containing the act stipulates that one or two slaves remain in the service of his widow). Acts of manumission readily cite reasons of a religious order: the love of God or of his saints, the desire to be forgiven of one’s sins or to obtain eternal salvation. Similar manumissions in extremis were carried out by Jewish merchants in Cairo.¹⁰⁰ In the large Italian and Iberian cities, religious institutions emerged specifically to attend to these liberated slaves.¹⁰¹

Manumission was sometimes granted as a reward for services, past or future, given to one’s master. A number of Muslim captives in Constantinople eventually converted to Christianity, married Christian women, and settled in the empire, where they often served in the army. If we are to believe the tenth-century geographer Mas‘di, the empire had a contingent of twelve thousand Christianized Arab cavalrymen. Byzantine and Frankish captives, only some of whom had converted to Islam, also served in the armies of Muslim princes.¹⁰²

The laws of Italian and Iberian cities tried to regulate the sex lives of slaves (especially women). In the large agricultural estates, the issue at hand was primarily unions between slaves, whose children were slaves like their parents, while in the cities it was more often a matter of liaisons between slave women and free men. In Genoa, any man who acknowledged impregnating someone else’s slave had to pay a fine to make amends to the master; the children born of such unions, usually free, were given to orphanages or adopted by their fathers. In Spain, by contrast, the child generally remained the property of the mother’s master. The Fuero of Teruel (1176 or 1177) imposed a fine of twenty gold coins on any man who raped a Moorish slave not his own; there was no punishment, however, for raping one’s own slave. If a slave gave birth and the master was not the father, the child would be the master’s slave until the father purchased it. Only children who were redeemed, and then freed, could inherit from their fathers. The provisions were similar in other fueros.¹⁰³

Slaves and captives sometimes tried to escape. This was not a major problem in Italian cities, where slaves were far from their native lands and could not count on the aid of the local population. In Byzantium, fugitive slaves could take refuge in monasteries and become monks. Hagiography provides various examples of saints, such as Saint Faith of Conques or Saint Dominic of Silos, who focused particularly on the miraculous liberation of Christian captives from the hands of the infidels.¹⁰⁴ But other saints, especially Theodore Tiron and John the Soldier, actually helped Christian masters recover their fugitive slaves.¹⁰⁵ In Spain, near the border, Malekite jurists tried to establish procedures for capturing the escaped slaves and returning them to their Muslim masters.¹⁰⁶ In the Christian kingdoms, fugitive Muslim slaves were sometimes aided or hidden by free Mudejars. It seems, in fact, that in certain areas of Spain there was a sort of underground railroad that helped escaped slaves flee to Muslim lands. In the mid-twelfth century, the Usatges de Barcelona established monetary rewards for returning fugitive slaves to their masters.¹⁰⁷ In the fifteenth century, the king of Aragon founded an obligatory insurance fund for slave owners, who paid an annual premium for each slave and received indemnities for each fugitive. The money collected was used not only to pay compensation but also to maintain a dedicated police force and to give rewards to those who helped recover the escaped slaves. Once captured, these slaves were auctioned off, and the money from the sale returned to the fund.¹⁰⁸

Religious coexistence in Europe is not the result of twentieth-century immigration: it existed in European Christian and Muslim societies throughout the Middle Ages. Religious minorities—Jewish, Christian, or Muslim—played a key role at that time, as they still do today, in the transmission of knowledge and culture and also in commerce.