Christian Martyrs in Muslim Spain

Kenneth Baxter Wolf


Christians in Muslim Córdoba

[5] In late summer of the year 711 a small Muslim contingent under the command of Mughîth ar-Rumî found its way to the bank of the Guadalquivir river opposite Córdoba.(1) There, in a stand of pine trees safely out of sight of the city's sentries, they set up camp and began to reconnoiter the area around the city. Some of the scouts came upon a shepherd, whom they apprehended and brought to Mughîth for questioning. From him the commander learned that the bulk of the Cordoban nobility had already fled north in anticipation of a Muslim assault, leaving the town with a depleted garrison of no more than four or five hundred men. The shepherd also informed him of a large fissure in the otherwise sound and formidable wall that girded the city.

When darkness came it brought with it an unseasonably cold rainstorm which not only helped to obscure the Muslim forces as they crossed the river, but compelled the Cordoban sentries to abandon their posts in search of temporary shelter. With the shepherd as his guide, Mughîth found the aperture and sent a handful of men through it with instructions to open the south gate adjacent to the old Roman bridge. The surprised soldiers fled as the locks on the gate were smashed, allowing the rest of Mughîth's men to ride into the city. Mughîth proceeded to occupy the abandoned palace, enlist the aid of the local Jews to help guard the city,(2) and compose a letter describing his success to his superior, Târiq ibn Ziyâd. The only threat to his control of the city was removed some three months later when, after capturing the commander of the Cordoban garrison as he attempted to flee north to Toledo,(3) Mughîth led a successful assault on the soldiers who still held out in the church of St. Acisclus.

The Akhbâr Majmûca, the anonymous source that describes this [6] first encounter between the Muslim conquerors and the peoples of Córdoba, says nothing about Mughîth's subsequent dealings with the vast Christian population that chose not to follow its leaders into exile. We can only surmise, based on the chronicler's silence as well as the general pattern of Muslim conquest, that the Christians who remained offered no resistance. Fortunately the independent account of the historian ar-Râzî allows us to fill in some of the gaps. In the course of his description of the founding of the famous Cordoban mosque, the chronicler explained that the conquerors appropriated half of the local church dedicated to St. Vincent for use as a mosque. This was not an uncommon stopgap measure for dealing with the religious needs of the victorious armies. Syrians in the town of Hims had experienced similar divisions in the wake of their conquest.(4) But with the steady influx into Córdoba of Arab immigrants over the next two generations, the Muslim worshipers found their quarters increasingly cramped. During the reign of the first cUmayyad emir, cAbd ar-Rabmân I (754-88), negotiations began between the emir and the leaders of the Christian community to resolve the problem. Finally, after a promise of a large cash payment as well as permission to rebuild one of the extramural churches that had been leveled at the time of the conquest, the Christians relinquished their half of St. Vincent's. The emir then ordered the demolition of the church to make way for the construction of the mosque that occupies the site to this day.(5)

For our purposes, the most significant fact about this episode is that the Christians specifically appealed to a capitulation agreement, worked out between their grandfathers and the original conquerors of Córdoba, as the legal basis for their claims to the church. Unfortunately the only stipulation of this lost treaty about which we have any direct knowledge is the one to which ar-Râzî referred: that the Christians would relinquish control over half of the church of St. Vincent's. But by reviewing the capitulation agreements that survive from other parts of the Islamic empire, we can reconstruct the basic components of the original pact offered to the citizens of Córdoba.

Every such agreement reached between Muslim leaders and the peoples they subjected was based in principle on two passages from the Qur'ân: [7]

Fight those who believe not
In God nor the last Day
Nor hold that forbidden
Which hath been forbidden
By God and His Apostle
Nor acknowledge the Religion
Of Truth (even if they are)
Of the People of the Book,
Until they pay the Jizya
With willing submission,
And feel themselves subdued.

Those who believe (in the Qur'ân),
And those who follow the Jewish (scriptures),
And the Christians and the Sabians, --
Any who believe in God
And the Last Day,
And work righteousness,
Shall have their reward
With their Lord: on them
Shall be no fear nor shall they grieve.(6)

The militant monotheism that infused the Islamic concept of jihâd could not tolerate the existence of religious alternatives that recognized neither God's unity nor the inevitability of his judgement. Thus polytheists were given no real choice by their Muslim conquerors but to embrace Islam. The Christians and the Jews were, however, a different matter. Having met the basic monotheistic criterion, their only failure in the eyes of the Muslims was their misinformed, but ultimately tolerable, reluctance to accept Muhammad as the most recent in the long line of prophetic successors to Abraham. This ambiguity in the religious status of the scriptuaries or "peoples of the book" led to a certain ambiguity in their legal position within Islamic society as well. Though tolerated and protected on the basis of their monotheism, they suffered political subjection and the stigma of a special tax, the jizya, as a penalty for their rejection of Muhammad's prophethood.

The Qur'ân was not the only, or even the most significant, factor influencing Muslim policy toward subject religious communities. The fact that the Muslims were vastly outnumbered by the Christians in the Mediterranean basin limited their options. They could not, even if their scriptures had demanded it, effect a mass [8] conversion of local populations. On the contrary, their early policies suggest that the Muslim rulers and jurists were more concerned about protecting their own people from the potentially polluting effects of close contact with large Christian populations. Limited by their numbers and wary of cultural absorption, the Muslims had every reason to allow a great deal of autonomy to any Christian community that recognized their authority and paid the jizya.(7)

The earliest examples of the "covenant of protection" (dhimma) which Mubammad and his generals used to encourage Christian cities to surrender illustrate the confluence of religious, economic, and strategic concerns. In the year 630 the Christians of Ayla received a promise of complete protection of person and property in exchange for the payment of a yearly tribute amounting to one dinar per adult male citizen.(8) In 632 Muhammad offered the same terms to the Christians of Najran in exchange for their promise to provide clothing and silver in amounts which varied according to the town's yearly production, horses and shields in the event of a local war, and hospitality to any Muslim envoy who might request it.(9)

The treaties offered to the scriptuaries of the cities and territories conquered after the prophet's death reflect his emphasis on tribute in exchange for the protection of traditional liberties. The only such agreement that has survived from the conquest of the Iberian peninsula is no exception. In it, Count Theodemirus of Murcia agreed to recognize the overlordship of cAbd al-cAziz and to pay tribute consisting of a yearly cash payment supplemented with specific agricultural products. In exchange, Theodemirus received cAbd al-cAzîz' promise to respect both his property and his jurisdiction in the province of Murcia.(10)

Even though we lack a copy of the Cordoban dhimma, then, we know, on the basis of the pattern established by Muhammad and followed by his successors, that it must have stipulated the payment of an annual tax, whether in kind or in cash, in exchange for a pledge on the part of the authorities to preserve the personal and proprietary liberties of the dhimmîs.(11) Here the certainty ends. Yet on the basis of the special circumstances of the Muslim settlement in Córdoba, we can speculate a bit further about the contents of the lost covenant.

[9] Ar-Râzî, as we have seen, explicitly recorded that the Muslims appropriated half of the principal Visigothic church for use as a mosque. This, along with Mughîth's occupation of the Cordoban palace, strongly suggests that from the very beginning the Muslims took up residence in the city itself. Though this arrangement may have become the norm in Spain, it contrasted with the settlement patterns in the east, where the Muslim armies preferred to establish garrisons apart from the subdued cities.(12) Such segregation made sense for both strategic and, as we have seen, religious reasons. Later, as Muslim political control grew more secure and the booty and tribute oriented economy gave way to one based more on commerce, the garrison and city began to merge. When this happened the original succinct covenants governing the relations between the dhimmî population and the Muslims required significant elaboration in order to salvage some of the social distance that the garrison system had achieved by simple physical segregation. In the generations that followed the early conquests, Muslim rulers promulgated new statutes in an effort to keep the line that separated Muslim and dhimmî from being erased.

The fact that Córdoba did not fall to the Muslims until the second decade of the eighth century, a time when Muslim jurists in the east were already implementing such regulations, may explain why its Muslim conquerors did not hesitate to settle within the city itself.(13) The same chronological connection allows us to hypothesize about the types of social limitations under which Christians in Spain lived on the basis of what we know about the better documented legal situation in the eastern Islamic world.

Much of this new legislation aimed at limiting those aspects of the Christian cult which seemed to compromise the dominant position of Islam. cUmar II (717-20), whose caliphate produced the first full series of laws designed to keep dhimmîs in their place, forbade the construction of any new churches.(14) The so-called pact of cUmar I, that came to be considered normative by the legal schools emerging in the eighth century, went even further, forbidding the repair of dilapidated, pre-existing churches.(15) The caliphs and jurists also prohibited rituals and activities that drew too much public attention to Christianity, such as bell ringing or excessively loud chanting in church.(16) Restrictions on processions and funerals were common for the same reasons. The most severe penalties, [10] however, were reserved for those Christians who showed disrespect for Islam: its prophet, its tenets, or its adherents. Any attempt to convert Muslims to Christianity or to prevent Christians who so desired from adopting Islam was considered an act of lèse majesté.

Aside from such cultic restrictions most of the laws were simply designed to underscore the position of the dimmîs as second-class citizens. cUmar II issued a series of sumptuary laws prohibiting dhimmîs from adopting certain characteristically Arab clothing and hair styles.(17) He also seems to have been the first caliph to order all non-Muslim subjects to wear what would later come to be known as the zunnâr, a distinctive type of belt or girdle, for identification purposes.(18) The pact of cUmar also included restrictions on Christian use of arms and saddles as well as characteristically Arabic seals and onomastic structures.(19)

Although from the beginning Christians were given virtual autonomy to govern the affairs of their own religious communities, their subordinate position theoretically restricted their access to government positions which entailed exercise of authority over Muslims. Slaves of Christians could, on the basis of this same principle, win their freedom simply by converting to Islam. Similarly the social position of the dhimmî precluded certain types of marital relationships. Sexual relations between a male dhimmî and a female Muslim were absolutely forbidden, though a male Muslim could legally marry a Christian or Jew. Muslim jurists may have borrowed this restriction from the Byzantine law books which frowned on the intermarriage of Christians and Jews.(20) Or it may have been a more autochthonous product of the tribal milieu from which Islam emerged. Traditional Bedouin society, which placed great weight on endogamous marriage, considered it a sign of weakness for a clan to lose its female members in marriage to the men of another.(21) If this tribal convention had any impact at all we might expect female members of the Islamic "tribe" (umma) to be considered off limits to Christian outsiders.

This proliferation of legal restrictions on Christian activity did not necessarily mean that the actual situation of Christians living under Islam deteriorated to any appreciable degree. For the same inexorable movement toward physical integration that occupied Muslim legal minds inevitably proved stronger than the laws they [11] created. Enforcement always depended on the inclination of the particular authorities involved, and it was typically not in their best interests to carry out the subordination of the dhimmî populations to the full extent of the law.

For one thing the Muslims in a given city were almost always outnumbered by the non-Muslim indigenous population, a balance which in many areas was not altered for centuries after the original conquest.(22) In those cities where Christians constituted either the majority or a sizeable minority, the bishops often served in an important ministerial capacity to the local Muslim governors, and not simply in matters pertaining to the Christian community. The late tenth-century bishop Severus of al-Ushmunain in Egypt composed a history of the leaders of the Alexandrian church, which is full of indications of a surprisingly close working relationship between the Egyptian rulers and the patriarchs, who were often used as ambassadors, consulted for political advice, or even solicited for prayer.(23)

The episcopacy was not, however, the only avenue to public service open to Christians. Muslim conquerors often found it to their advantage to leave local administrative structures intact, staffing them with native non-Muslims who were familiar with their workings. The fact that dhimmîs lived outside of the intricate network of Arab and Berber tribalism increased their attractiveness to Muslim rulers looking for loyal ministers and public servants who could, if necessary, be easily dislodged. For these reasons, caliphs like cUmar II never succeeded in effecting any complete or permanent purge of the bureaucracies of Islam, despite the perennial grumblings of Muslims who resented being placed under the authority of Christians.

While the Arabic sources that pertain specifically to ninth-century Córdoba are largely silent on matters of Christian-Muslim interaction, the Latin documentation provides a good deal of information about such contact and the social liabilities that the Christians of the time faced. On the basis of these sources not only can we reconstruct some of the restrictions on Christian activity that would presumably have been specified in the lost treaty, but we can assess the frequency with which the more universal proscriptions were being enforced.

The Christians of Córdoba were without a doubt subject to [12] regular taxation. Eulogius complained of a "monthly tribute" that constituted a financial hardship for the Christians.(24) Similarly Alvarus wrote of an "unbearable tax" that weighed heavily on Christian necks.(25) There is little doubt, given the tone of victimization that both men adopted when speaking of the levies and the regularity with which they were collected, that tributum, vectigal, and census were simply Latin synonyms for the universal dhimmî tax, the jizya.(26)

The Cordoban authorities also prosecuted Christians guilty of blasphemy. In the spring of 850, a priest named Perfectus was arrested and later executed for publicly expressing his opinions about the errors of Islam to a group of Muslims.(27) Months later a Christian merchant named Joannes suffered a severe lashing, public humiliation, and a long prison term for invoking the prophet's name as he sold his wares in the marketplace.(28) But though diligent in their enforcement of the laws against blasphemy, the emirs of the time were willing to let more venial violations of the ban on public displays of Christianity go unchecked. Bells rang in Córdoba to indicate the canonical hours for the benefit of the faithful.(29) Christian funeral processions passed through Muslim neighborhoods.(30) Interestingly enough, though these acts provoked no official censure, they did at times become a target for public derision. Eulogius complained that the "clang of the reverberating metal" evoked Arabic curses from the Muslims who were within earshot.(31) Alvarus expressed his shock that priests were often subjected to verbal abuse or even pelted with rocks and dung when taking the dead to the cemetery.(32) These popular outbursts, unconnected with any known official action, suggest that laws against bell-ringing and processions had once been promulgated in Córdoba, but had since fallen into desuetude.

The mockery and jeering that some priests suffered as they passed through Muslim quarters in their distinctive clerical garb may also have been a vestige of earlier limitations on the public expression of Christianity.(33) This type of abuse may have been the principal factor in the declining standards of clerical dress decried by Leovigildus, the ninth-century Andalusian author of the Liber de habitu clericorum. Leovigildus criticized what he regarded as a "lack of ardor" and a certain "fatuity among some of the clerics," prompted by "Ishmaelite oppression."(34) In the course of the [13] treatise he reviewed the symbolic significance of each aspect of the clerical regalia, apparently for the benefit of those who had reason to question the utility of a uniform that significantly increased their susceptibility to public ridicule. The fact that there are no examples of laymen encountering this type of unprovoked derision strongly suggests that the sumptuary laws which would have served to set every Christian apart were not enforced in ninth-century Córdoba. With the exception of the clergy, Cordoban Christians must have looked much like Cordoban Muslims.(35)

The evidence regarding church fabric in Córdoba also attests to the flexibility of the emirs. As we have seen, Ar-Râzî explicitly stated that all of the churches except St. Vincent's were torn down in the aftermath of the conquest. As part of the later agreement the authorities permitted the Christians to rebuild one of the extramural churches for their exclusive use. Yet in the course of his writings Eulogius makes mention of no less than thirteen different institutions located in the immediate vicinity of Córdoba: four basilicae and nine monasteries, at least two of which were established within the priest's own memory.(36)

Half of these monasteries lay in or at the foot of the mountains which border the Guadalquivir valley a few miles north of the city.(37) Most of the others were situated down river.(38) Unfortunately Eulogius was less precise about the exact locations of the churches.(39) But the Arab historian Ibn Haiyan refers specifically to churches both in and outside of Córdoba, suggesting that if indeed the original conquerors did level all the churches except St. Vincent's, subsequent rulers allowed many of them to be rebuilt.(40)

The many examples of Christians in positions of authority within the Cordoban government illustrate the remarkable extent to which the emirs had suspended laws controlling religious integration. cAbd ar-Rahmân II worked closely with the bishop Reccafredus to discourage the spontaneous martyrdoms.(41) But bishops were not the only Christians in Córdoba who found themselves in positions of power. The emir al-Hakam I (796-822), shaken by a rebellion in a Cordoban suburb in 805, created a special bodyguard headed by the leading secular member of the local Christian community, count Rabic, son of Theodulf.(42) According to other Arabic sources, the same count served as the emir's main tax collector until he was removed and executed for [14] alleged misappropriations.(43) The use of Christians as both soldiers and tax collectors was as commonplace in ninth-century Córdoba as it was throughout the Islamic world, owing, again, to the absence of the tribal loyalties that could potentially compromise the allegiance of an Arab or Berber.(44) Other positions within the Cordoban bureaucracy were equally open to Christians. Samson, who later became an abbot in one of the local monasteries, served on occasion as an official translator to the emir.(45) A Christian nobleman named Argimirus acted in some sort of judicial capacity as Muhammad I's censor.(46) Finally, one of Eulogius' own brothers, Joseph, was a member of the principatus in the late 840s, but our source failed to elaborate on the official duties involved.(47)

We also know of two Christians who, perhaps consecutively, occupied the office of exceptor reipublicae. Isaac, whom we shall later meet as the first of the spontaneous martyrs, was promoted to the position because of his fluency in Arabic.(48) After he retired from public life the office passed to another Christian, probably the count Ibn Antonian, whom Ibn al-Qûtîya and al-Khushanî both described as a kâtib, or secretary, to Muhammad I.(49) It appears that the office of exceptor is identical to what the Arabs referred to as the kâtib adh-dhimam (secretary of the covenant), whose purpose it was to "attend to the protection and security of the Christians and Jews."(50) The only view that Eulogius provided of the exceptor acting in his official capacity had him presiding over the episcopal council in 852 as the direct representative of the emir, a task which would seem to fit the responsibilities of the kâtib adh-dhimam quite well.(51) It made sense to fill this important position with Arabic-speaking Christians, who could most effectively communicate with the leaders of the largest Andalusian religious contingent.

When all of the evidence from both Latin and Arabic sources has been reviewed, it would appear that for the most part the laws designed to keep Christians and Muslims at what the jurists regarded as the proper social distance went unenforced in ninth-century Córdoba. Taxation and the proscriptions against blasphemy and apostasy are the only apparent exceptions. Of these only the jizya could have served as anything like a perennial reminder of the subordinate status of the Cordoban Christians. The occasions for enforcing the other two were, under normal circumstances, [15] simply too few and far between to underscore the religious divisions.

Such was the rather relaxed legal climate that prevailed in Córdoba prior to June 3, 851, when the monk Isaac descended from his mountain retreat and, without provocation, berated a Cordoban judge for failing to recognize the errors of Islam and its prophet. As we shall see in the next chapter, his example led to periodic waves of Christians flouting the proscriptions against blasphemy in their search for a martyr's death. But before we consider their actions in any detail, we must extend our overview of the relationship between the authorities and the Christian community a bit further into the 850s. For with the very first spontaneous martyrs came some significant modifications in the emir's treatment of the Christian population of Córdoba.

When the judge, at whom Isaac had chosen to vent his animosity toward Islam, reported the case to cAbd ar-Rahmân II, the emir ordered the decapitation of the monk and immediately promulgated an edict which reiterated the Islamic prohibition of blasphemy, threatening any subsequent violators with the same punishment.(52) When, over the next four days, seven more Christians followed Isaac's example, the authorities began to worry. Eulogius no doubt exaggerated their reaction when he wrote that "the Muslims were dumbfounded with fear as a result of these events and thought the doom of their republic and the ruin of their kingdom to be at hand."(53) But given the history of local and provincial uprisings against the emirs, it would be a mistake to underestimate their concern over this new form of dissent.

cAbd ar-Rabmán II had experienced more than his share of troubles during the reign of his father. As an adolescent he had witnessed the crucifixion of seventy-two men, from the rabad shaqunda on the bank opposite Córdoba proper, who, in 805, had conspired to displace al-Hakam with one of his cousins.(54) Thirteen years later, despite the emir's precautions, he came frighteningly close to another violent deposition at the hands of rebels from the same volatile quarter. cAbd ar-Rahmân probably took part in the three day sack and finally the razing of the rabad that followed the victory of al-Hakam's troops.(55) Though the depopulated right bank would host no more conspiracies, the memories no doubt left the new emir all the more wary of signs of local dissent. Not [16] surprisingly, then, when eight Christians in five days came forward to pronounce judgement on Islam, he reacted swiftly by ordering the arrest and detention of the clerical leadership of the local Christian community.(56)

The imprisonment of the clerics lasted anywhere up to four months, ending in November 851. At first it appeared to have achieved its purpose: the outbursts quickly subsided with only a pair of executions recorded over the next seven months. When, however, the Andalusian summer brought with it a new wave of martyrs, the emir turned again to the Christian leaders as the ones most capable of controlling the zealots. But instead of imprisoning them, he ordered them to convene a council in Córdoba to review the matter and develop some strategy for dealing with the dissidents internally.(57) The result was not exactly what the emir had in mind. The statement which the council drafted, though hardly a vote of confidence for the martyrs, still did not bring any significant ecclesiastical pressure to bear on Christians who might have been considering the same course. The week before cAbd ar-Rahmân's death on September 22, 852, saw the executions of four more Christians.

Muhammad I, who succeeded his father as emir, followed a different tack in his dealings with the martyrs. He immediately began to apply pressure to the Christian community as a whole by enforcing many of the dhimma restrictions that had either never been enforced in Córdoba or had long since been neglected. "On the very day that he received the scepter of the kingdom," wrote Eulogius, "he ordered all Christians to leave the palace, deprived them of their dignity, and removed their honor. . ."(58) This purge led to the expulsion of the incumbent exceptor reipublicae who, owing to his exceptional proficiency in Arabic, managed to hang on a few months longer than the rest of the expelled Christians. Ultimately, however, he was reinstated after having, in Eulogius' words, "spurned the holy Trinity and joined the perverse sect."(59)

One interpretation of the motives behind the purge has emphasized the influence of Malikite jurists in the emir's court, whose legal conservatism would not let them stand for breaches of the dhimma.(60)Eulogius himself noted that Muhammad I replaced the Christian officials with Muslims who "laboring with a similar zeal against God, afflicted, subverted and oppressed them everywhere. . ."(61)[17] And given the strong identification of ninth-century Andalusian jurisprudence with the teachings of Mâlik ibn Anas, we would expect Muhammad I to rely on Malikite ministers to advise him when he came to power.(62) But the same could accurately be said of his father cAbd ar-Rahmân II or his grandfather al-Hakam I, both of whom patronized disciples of Mâlik, yet apparently did not feel compelled to enforce the dhimma restrictions. Furthermore, if the driving force behind Muhammad's purge was a Malikite-influenced program to put the dhimmîs in their proper place, why were the Christians singled out for expulsion? Eulogius himself could not figure it out: "why . . . if the emir enjoyed such free exercise of power, did he not also force the Jews to be removed from his presence. . .?"(63) The answer is simply that the Christians, and not the Jews, were regarded as troublemakers.(64) The hitherto neglected proscriptions against dhimmîs were applied specifically to Christians for political reasons, to discourage any further Christian participation in Cordoban dissidence.(65)

Muhammad was undoubtedly taken aback when, after nine months without a single incident, the Christian Fandila ushered in a new series of martyrdoms with his denunciation of Islam in June 852. The emir responded by reactivating another dhimma restriction that, as we have noted, had long since gone unenforced: "He ordered that newly constructed churches be destroyed, as well as anything in the way of refinements that might adorn the old churches, added since the time of the Arab occupation."(66)

Interestingly enough Muhammad I was unable fully to implement this policy due to military distractions in other parts of al-Andalus. "On all sides the insurgent wars of rebellion brought him great anxiety, for all the cities of Spain, which his father had dominated and occupied with the strength of his forces and the excellence of his power, or had acquired with great remuneration, Muhammad now bore the responsibility of controlling. When, in various places, he saw his army defeated and put to flight, he lamented, as cities everywhere were diminished and given to ruin."(67) Eulogius, writing sometime in the summer of 853, was not specific about the nature of these bella dibellionum that distracted the emir. But fortunately the Arab historians, so attentive to the details of Andalusian military campaigns, recorded the events to which the priest was alluding. [18] The old Visigothic capital of Toledo, like so many of the other provincial Andalusian centers, perennially resisted the centralizing efforts of the emirs of Cordoba.(68) In Toledo's case much of its recalcitrance stemmed from the fact that neither the Arabs nor the Berbers had ever settled the city in any great numbers, leaving it in the hands of an indigenous population, which, by the mid-ninth century, was composed primarily of Christians and converts to Islam, or muwallads.(69) Upon the death of cAbd ar-Rahmân II in September of 852, the two Toledan factions joined in revolt against Córdoba and occupied the nearby town of Calatrava. During the following summer a Cordoban army under the direction of the new emir's brother managed to retake Calatrava and establish a permanent garrison there. But undaunted, the Toledan forces wreaked havoc in the Andújar area, where they ambushed a Cordoban battalion and set ablaze the farmlands of the Jandula valley. Meanwhile the Toledan leaders were cementing an alliance with the new Asturian king, Ordoño I, who, like many Spanish Christian kings before and after, was more than willing to take advantage of internal strife in al-Andalus to extend his own political power. In June of 854, at the Guadacelete creek just southeast of Toledo, the forces representing the Asturian-Toledan alliance met the Cordoban army, led by Muhammad himself, and were soundly defeated.(70) But in spite of the victory the emir was still unable to force Toledo into submission. Like Tudela, Zaragoza and Huesca which were at that time ruled independently by the powerful muwallad clan of the Banû Qasî, and Mérida then under the control of the Berber Banû Tajit, Toledo remained stubbornly outside of Muhammad's sphere of influence.(71)

It is interesting that Eulogius regarded these "wars of rebellion" that "brought great anxiety" to Muhammad I as a windfall which saved the bulk of the Cordoban churches from ruin. For in fact it would appear that the rebellion in Toledo was the main reason why the new emir applied such pressure on the Cordoban Christian community when the martyrdoms resumed in June 853.(72) The Toledan uprising, timed to correspond with an expected period of political weakness under a new emir, not only posed a dangerous challenge to Muhammad's authority, but, because it involved Toledan Christians attempting to forge an alliance with the Christian kingdom of Asturias, must have worn thin any restraint [19] that he might otherwise have exercised toward dissident Christians in his own capital. The purge of Christian officials that Muliammad had effected in September 852, had not, as it turned out, stifled the suicidal outbursts against Islam. Clearly the emir's plan to destroy churches was conceived with the same goal in mind: to pressure the Christian community as a whole into controlling its more radical exponents at a time of particular political stress.

That Muhammad regarded the Cordoban dissidents in much the same way as he regarded the rebels that threatened his control of the provinces is apparent from Eulogius' account of another of his abortive plans for dealing with the martyrs. In the wake of the new outbreak in the summer of 853, Muhammad toyed with the idea of "killing all Christian men and dispersing their women by selling them into slavery, except those who spurned their religion and converted to his cult."(73) His ministers, however, advised him against such an extreme measure. They pointed out that "no wise or urbane Christian, nor any of their leaders, had perpetrated these actions and on that basis asserted that Christians as a whole ought not to perish." The reasoning of the advisors suggests that they were contrasting the unorganized, apolitical disturbances in Córdoba, with the full-fledged rebellions of the north for the benefit of an emir who tended to conflate the two into a single challenge to his authority.

Combining the evidence from ninth-century Latin sources with the information we can glean from earlier Arabic sources from the eastern part of the Islamic empire, we have come to some important conclusions about the relationship between the Andalusian emirs and their Christian subjects. Ar-Râzî's account of the negotiations leading up to the construction of the Cordoban mosque attests to the existence of a dhimma regulating Christian activities and their relations with Muslims. Similar agreements between Muslim leaders and their dhimmî subjects in the east allowed us to speculate as to the contents of the lost Cordoban pact. Though we suspect that the Cordoban law books contained all the usual restrictions on Christian activity, there is little evidence to suggest that the cUmayyad emirs were ever diligent about enforcing them. In fact the Latin sources dating from the 850s and 860s lead to the exact opposite conclusion. Aside from the jizya, the liabilities of Christian life in al-Andalus were minimal. Of all the traditional legal restrictions [20] placed on the exercise of Christianity, the only ones for which we have any evidence of consistent enforcement are the proscriptions against blasphemy and apostasy.

But just because the laws were not being enforced does not mean that they could not be enforced at the discretion of the ruler. Just as it was in the emir's interest during times of peace and prosperity to elevate Christians to positions of public service, to permit them to build churches, and in general to allow them to assimilate into Cordoban society, so in times of stress did it make sense to restore the social barriers and the sense of Christian subordination by enforcing the traditional restrictions. The relations between the emirs cAbd ar-Rahmân II and Muhammad I and the Christian community of Córdoba in the early 850s should be seen in terms of one of these shifts in policy, one which was elicited by a coincidental series of provincial rebellions and local martyrdoms.

Notes for Chapter 1
1. Antonio Arjona Castro, Anales de Córdoba musulmana (Córdoba, 1983), pp. 13-14. For a variation of the account, see: Ahmed ibn Mohammed al-Makkari, The History of the Mohammedan Dynasties in Spain, tr. Pascual de Gayangos, 2 vols. (London, 1840), 1:277-80.

2. This was a common practice in Spain at the time of the conquest, as Eliyahu Ashtor, in his The Jews of Moslem Spain (Philadelphia, 1973), pp. 15-26, indicates. Arabic chroniclers report the establishment of similar Jewish garrisons in Granada, Toledo, and Seville (al-Makkari 1:280, 282, 284), and claim that the practice "became the fixed method of the conquerors" (Ashtor, p. 24). Most historians agree that the harsh treatment of the Jews under Visigothic rule, as evidenced by the Toledan councils of the seventh century, led the Jews to applaud the change of leadership. José Vives, ed., Concilios visigóticos e hispano-romanos (Barcelona, 1963), e.g., Toledo 4.57-66 (633), pp. 210-14, and Toledo 17.8 (694), pp. 534-6.

3. This Christian leader was, according to Arabic sources, the only "prince" captured during the conquest. All of the others either capitulated or fled with greater success. al-Makkari 2:15.

4. Some Arab sources indicate that the first mosque in Damascus was also part of a church, but other evidence suggests this was not the case. A. S. Tritton, The Caliphs and their Non-Muslim Subjects: A Critical Study of the Covenant of cUmar (London, 1930), pp. 9-10, 39, 40-2.

5. Al-Makkari 1:217.

6. The Holy Qur'an, trans., A. Yusuf Ali, 2nd ed., 1977, Sûra 9.29, p. 447; S.2.62, pp. 33-4. Cf., S.5.69, p. 265.

7. Thomas F. Glick, Islamic and Christian Spain in the Early Middle Ages (Princeton, 1979), p. 169. Many historians have marveled at this "tolerance." But, as Glick (p. 174) has observed, "The communal autonomy of the groups, often represented as the very symbol of tolerance, was in fact the institutional expression of ethnocentric norms which held such groups in abhorrence, as tolerated but alienated citizens who were not to share in social life on the same basis as members of the dominant religion." Robert I. Burns, S.J., noted the same phenomenon in thirteenth-century Valencia, where the political tables had been turned and the Christians were segregating themselves from the Muslim majority: "Tolerance at this extreme is not easily distinguishable from intolerance." Islam under the Crusaders (Princeton, 1973), pp. 186-7.

8. Majid Khadduri, War and Peace in the Law of Islam (Baltimore, 1955), p. 178. Antoine Fattal, Le statut légal des non-musulmans en pays dIslam (Beirut, 1958), p. 21.

9. Khadduri, p. 179.

10. Francisco X. Simonet, Historia de los mozárabes de España (Madrid, 1903), pp. 797-8. Although this is the only capitulation agreement that has survived from the peninsular conquest, others are mentioned by Arab historians. The heirs of Witiza and Roderick, the two Visigothic pretenders to the throne, both retained their rights to family property in exchange for their cooperation. Al-Makkari 2:14, 30. See also: Alfred M. Howell, "Some Notes on Early Treaties Between Muslims and the Visigothic Rulers of Al-Andalus," in Actas del I Congreso de historia de Andalucía, diciembre de 1976, Andalucía medieval, v. 1 (Córdoba, 1978), pp. 3-14.

11. A dhimmî is someone bound by the constraints of a dhimma, in the case of Spain, a Christian or a Jew living under Muslim rule. The conquerors did not typically extend the protection of property to those Christians who either resisted or fled. We know in the case of Mérida that the conquerors confiscated the property of all those who either died in combat or fled to Galicia. Simonet, p. 52.

12. The only exceptions during the initial phase of Islamic expansion were Ctesiphon, where morale problems prompted the change in policy in the first place, and cities in Syria where pre-existing treaties governed settlement patterns. Marshall G. S. Hodgson, The Venture of Islam, 3 vols. (New York, 1972), 1:208.

13. It is difficult to determine whether or not the lack of a garrison in Córdoba was typical of the Iberian conquest as a whole. Roger Collins apparently assumes that the use of separate garrisons was as characteristic of Spain as it was of other areas conquered, but he does not devote any detailed attention to the problem. Early Medieval Spain: Unity and Disversity, 400-1000 (New York, 1983), p. 160. In any case it is interesting to note that when regional resistance to Cordoban authority began to proliferate in the next century, the emirs typically ordered the erection of garrisons in the most strategic areas of the troubled provinces. Such was the case in Zaragoza in 802, Mérida in 834, and Calatrava in 853. Evariste Lévi-Provençal, Histoire de l'Espagne musulmane, 3 vols., 2nd ed. (Paris, 1950), 1:156, 209, 292.

14. Tritton, p. 43. cUmar II and al-Mutawakkil (847-61) were the two most significant reformers of this type. Both sought to restrict dhimmî activity during times of particular political stress. Bernard Lewis, The Jews of Islam (Princeton, 1984), pp. 46-9.

15. Norman A. Stillman, The Jews of Arab Lands (Philadelphia, 1979), p. 157.

16. Tritton, p. 104; Stillman, p. 158.

17. Tritton, p. 116.

18. Tritton, p. 117.

19. Stillman, pp. 157-8.

20. Lewis, p. 27.

21. Pierre Guichard, Al-Andalus: estructura antropológica de una sociedad islámica en occidente (Barcelona, 1970), p. 81.

22. Glick, pp. 33-5.

23. Severus, History of the Patriarchs of the Coptic Church of Alexandria, ed. R. Graffin and F. Nau, Patrologia Orientalis, vols. 1, 5, 6, 10 (Paris, 1907-15).

24. Memoriale sanctorum 1.21 (PL 115:754; CSM 2:385); Eulogius, Documentum martyriale 18 (PL 115:830; CSM 2:470).

25. Paulus Alvarus, Indiculus luminosus 3 (PL 121:518 CSM 1:275).

26. The exact amount of taxation seems to have always been left to the discretion of the individual emir. The Chronica muzarabica, a mid-eighth-century continuation of Isidore's Chronicon, is full of references to changes in the levels of dhimmî taxation. As early as the 720s the Christians were complaining of the vectigalia duplicata imposed by the governor cAnbasa. As soon as he was replaced by Yahya ibn Salama, however, the money was restored to the Christians. Chronica muzarabica 60-1 (CSM 1:39). The references in the writings of Eulogius and Alvarus could have reflected the general increase of taxes under the emir cAbd ar-Rahmân II that we know about from independent sources: the historian Ibn Sa'îd and Louis the Pious who, in 826, sent a letter to the people of Mérida hoping to secure an alliance against the emir. Al- Makkari 1:124. MGH Epistolarum 5.1, ed. Karl Hampe (Berlin, 1898), p. 115.

27. Memoriale sanctorum 2.1 (PL 115:766; CSM 2:398). Indiculus luminosus 3 (PL 121:518; CSM 1:276).

28. Memoriale sanctorum 2.1 (PL 115:746-7; CSM 2:377-8); Indiculus luminosus 5 (PL 121:520; CSM 1:277-8). It is significant that Sa'îd ibn Sulaiman, the qâdî or judge, at the time of Perfectus' death and John's punishment, was appointed in the latter part of cAbd ar-Rahmân II's reign (822-52), after the emir had removed his predecessor from the same office for excessive lenience in the sentencing of a Muslim blasphemer. This precedent no doubt influenced Sa'îd's treatment of the two arrested Christians. Historia de los jueces, pp. 127-9.

29. Indiculus luminosus 6 (PL 121:521; CSM 1:278-9).

30. Indiculus luminosis 6 (PL 121:521; CSM 1:278). Eulogius, Liber apologeticus martyrum 33 (PL 115:867; CSM 2:493).

31. Eulogius, Memoriale sanctorum 1.21 (PL 115:755: CSM 2:386).

32. Indiculus luminosus 6 (PL 121:521; CSM 1:278).

33. Memoriale sanctorum 1.21 (PL 115:754; CSM 2:385).

34. CSM 2:668.

35. The only mention by an Arab chronicler of any special clothing worn by the dhimmîs in al-Andalus is a reference to yellow as being a color reserved for the woolen caps worn by Jews. Al-Makkari 1:116. The only reference to a zunnâr being worn by Cordoban Christians is an eleventh-century description of clerical garb. Al-Makkari 1:246. Moreover, by the ninth century, the effects of intermarriage had begun contributing to the physical homogenization of the population.

36. The monasteries located in the nearby mountain villages of Tabanos and Pinna Mellaria were both recently established as family retreats. Memoriale sanctorum 2.2 (PL 115770; CSM 2:402), 3.11.2 (PL 115:812; CSM 2:453).

37. These include the monasteries at Tabanos and Pinna Mellaria, along with St. Martin's at Rojana and the monastery of St. Justus and St. Pastor at Fraga. Another, dedicated to St. Zoilus, was located still further north behind the Sierra Morena.

38. The monastery of St. Christopher lay just downriver within sight of the city, while those dedicated to the Virgin and to St. Felix were located further southwest in the village of Cuteclara.

39. The ones that he mentions are the basilicae of St. Acisclus, St. Zoilus, St. Cyprian, and St. Faustus, St. Januarius, and St. Martialis.

40. Al-Makkari 1:246. A Cordoban calendar, dating from 961, includes references to some of the same monasteries and churches, as well as many that Eulogius never mentioned. Reinhart P. A. Dozy, Le Calendrier de Cordoue, 2nd ed., trans. C. Pellat, Medieval Iberian Peninsula Texts and Studies, vol. 1 (Leyden, 1961).

41. Alvarus, Vita Eulogii 2.4 (PL 115:709; CSM 1:332). Reccafredus may have been the metropolitan of Seville. Though he attended a council in 839 as the bishop of Córdoba and Cabra (CSM: 1:141), Saul served as bishop of Córdoba during the martyrdoms. See: Colbert, p. 176.

42. Lévi-Provençal 1:164.

43. Ibid, 1:196.

44. Ibn Khaldun (d. 1406), in his description of the duties of the khattât ash-ashghal, or tax collector, indicated that it was customary under the cUmayyads to appoint liberated slaves or members of an ahl adh-dhimma (people of the covenant). Al-Makkari 1, app. B, p. xxxi. Samson, who became abbot of Pinna Mellaria in 858, composed his Apologeticus in response to what he regarded as reprehensible conduct on the parts of count Servandus of Córdoba and bishop Hostegesis of Málaga, who had both engaged, among other things, in tax collecting. Apologeticus 2 pref., 5 (CSM 2:551). Eulogius also railed vehemently against Christian tax farmers who "crucified Christ's members daily." Memoriale sanctorum 3.5 (PL 115:803; CSM 2:443). Sanctius, one of the first of the spontaneous martyrs, was a soldier in the emir's service, and the fact that one of Muhammad I's strategies for stemming the flow of martyrs was to reduce military pensions suggests that Sancho was not alone. Memoriale sanctorum 2.3 (PL 115:771; CSM 2:402), 3.1 (PL 115:800; CSM 2:440).

45. Apologeticus 2 pref., 9 (CSM 2:554).

46. Memoriale sanctorum 3.16 (PL 115:815; CSM 2:455-6).

47. Eulogius, Epistula 3.8 (PL 115:848; CSM 2:500).

48. Memoriale sanctorum 2.2 (PL 115:770; CSM 2:402).

49. Historia de los jueces de Córdoba por Aljoxani, trans.Julián Ribera (Madrid, 1914), pp. 159-64. Historia de la conquista de España por Abenalcotía, trans. Julián Ribera (Madrid, 1926), pp. 67-9. Memoriale sanctorum 2.15.2 (PL 115:796; CSM 2:435), 3.2 (PL 115:801; CSM 2:440).

50. Al-Makkari 1:103.

51. Memoriale sanctorum 2.15.2 (PL 115:796; CSM 2:435).

52. Memoriale sanctorum 1, pref., 3 (PL 115:738; CSM 2:368).

53. Memoriale sanctorum 2.1.6 (PL 115:770; CSM 2:401).

54. Lévi-Provençal 1:163.

55. Ibid., 1:166-8.

56. Eulogius, Epistula 3.10 (PL 115:849; CSM 2:501); Vita Eulogii 2.4 (PL 1 15:709; CSM 1:332). It is uncertain how many of the Cordoban clerics were victims of this detention, but we can safely assume, given the point of the arrests, that at least the most prominent ones were involved. Whether or not it was Reccafredus who came up with the idea in the first place, he was the one pinpointed by Alvarus as the culprit.

57. Memoriale sanctorum 2.15.3 (PL 115:796; CSM 2435).

58. Memoriale sanctorum 2.16.2 (PL 115:797; CSM 2:436), 3.1 (PL 115:800-1; CSM 2:439). Ecclesiasticus 10.2.

59. Memoriale sanctorum 3.2 (PL 15:801; CSM 2:440-1).

60. Most recently, see James Waltz, "The significance of the voluntary martyrs' movement of ninth-century Córdoba," Muslim World 60 (1970), pp. 152-3.

61. Memoriale sanctorum 3.1 (PL 115:801; CSM 2:440).

62. For background on the Malikite school in Spain, see, in particular, Hussain Mones, "Le role des hommes de religion dans lhistoire de lEspagne musulmane jusquà la fin du califat," Studia Islamica 20 (1964), pp. 47-88, and Abdel Magid Turki, "La vénération pour Mâlik et la physionomie de malikisme andalou," Studia Islamica 33 (1971), 41-65.

63. Memoriale sanctorum 3.4 (PL 115:802; CSM 2:441).

64. The fact that the Christian kingdoms of northern Spain were asserting themselves politically at the expense of the Cordoban emirate only served to make the Muslim rulers more suspicious of Christian dissent in their own backyard. The Andalusian Jews, on the other hand, had "no secret relationships with factors outside the government that might frighten the Moslem rulers of the peninsula." Ashtor, p. 56.

65. Lévi-Provençal was very much aware of the political nature of the crackdown on the Cordoban Christian community: "If the reigns of many Umayyad emirs were marked by persecutions of Christian communities, of that of Córdoba in particular, . . . one must recognize that these persecutions were dictated less by the fanaticism of the princes than by their political preoccupations." Lévi-Provençal, p. 226. This was an opinion inherited from Reinhart P. A. Dozy, Histoire des musulmans d'Espagne jusquà la conquéte de l'Andalousie par les almoravides, 3 vols., rev. Lévi-Provençal (Leiden, 1932), 1:338.

66. Memoriale sanctorum 3.3 (PL 115:801-2; CSM 2:441). Eulogius expressed his outrage, not so much at the emir's order, as at its execution. For in the process, "the prince of darkness also tore down temples that had been erected. . . 300 years ago." This reference to pre-conquest churches in Córdoba does not fit ar-Râzîs account very well. Perhaps the Arab historian was referring to a specific radius around Córdoba, outside of which the original conquerors ignored the churches.

67. Memoriale sanctoruni 3.4 (PL 115:802; CSM 2:441). The only known casualty was the monastery at Tabanos, which was destroyed sometime during the summer of 853 between the martyrdoms of Digna (June 14) and Columba (September 17). Memoriale sanctorum 3.8.2 (PL 115:805; CSM 2:445), 3.10.9 (PL 115:809-10; CSM 2:450). The tenth-century calendar of Córdoba lists some of the familiar churches and monasteries of the previous century as sites of saint day celebrations. Dozy, Le Calendrier.

68. Al-Hakam I had resorted to the establishment of a garrison at Calatrava to deal with the Toledans. Al-Makkari 2:428, note 25.

69. Guichard, p. 278.

70. Al-Makkari 2:127. Lévi-Provençal 1:293-4.

71. Guichard, pp. 274-5, 278.

72. Eulogius regarded the new series of anti-Christian measures in 857 as a direct result of peace in the provinces. Liber apologeticus martyrum 22 (PL 115:863; CSM 2:488-9).

73. Memoriale sanctorum 3.7.4 (PL I 15:805; CSM 2:445). According to Eulogius, Muhammad's predecessor had at one point considered the same course of action. Ibid., 2.12 (PL 115:793; CSM 2:432). These were not necessarily exaggerations on Eulogius' part, nor idle threats on the part of the emirs. Al-Hakam I, as we have seen, dealt with the revolt of 818 by razing the rebellious quarter and deporting its inhabitants.