CHAPTER 6 The Fall of Norman Africa and the Legacy of Zirid-Norman Interactions

Roger II, Norman Africa, and the Mediterranean

The campaigns that Roger II had undertaken in Italy, Ifriqiya, and the eastern Mediterranean during the last twenty years of his reign expanded the frontiers of the Kingdom of Sicily but also made many enemies for the Normans. When William I inherited the throne from his father in 1154, these animosities boiled over in the form of internal unrest and external invasions. Although William was largely successful in suppressing this discord, his preoccupation with threats in Italy and Europe left his Ifriqiyan possessions neglected.

After the conquest of Mahdia in 543H (1148–49), Roger II engaged in efforts to expand the boundaries of the Kingdom of Sicily. He set his gaze east to Constantinople in an attempt to deter a looming alliance between the Byzantines and the Germans.¹ George of Antioch launched an ambitious raid that penetrated the port of Constantinople, and then defeated a Byzantine navy in the eastern Mediterranean in 1149. Around the same time (and possibly on the same expedition), he rescued Louis VII of France from a hostile Byzantine fleet that had captured the French king while he was returning from the disastrous Second Crusade.² The Norman king may have even had a hand in persuading certain groups in the Balkans to revolt against Byzantine rule in the autumn of 1149.³ Ibn al-Athir wrote that Roger “would have ruled all the lands of Ifriqiya” had he not devoted his navy to these expeditions in the eastern Mediterranean.⁴ While these campaigns provided additional plunder and prestige for Roger II’s navy, they also contributed to continued animosity between the Normans and Byzantines that would lead to future conflict.

Roger’s ambitions in the eastern Mediterranean were enhanced by his second and third marriages, which connected him to elite families in western Europe that had a storied history in the crusading movement. His first wife, Elvira of Castille, had died in 1135 from a disease that had also nearly killed him. Roger was inconsolable after her death and withdrew from the life of his court to such an extent that for a time rumors circulated of his own death.⁵ He did not remarry until 1149, when he wed Sibylla of Burgundy. The same political overtones that had characterized the marriage of Roger and Elvira were present in this second marriage.⁶ Sibylla’s brother governed an important duchy in Burgundy that straddled French and German territories, and this family also had a long-running commitment to crusading.⁷ Roger II, always looking for opportunities to expand his power in Antioch and Jerusalem, potentially saw this marriage as a way to further his interests in the Latin East.

The union between Roger and Sibylla did not last long. Sibylla died from a miscarriage in 1150, severing the Kingdom of Sicily’s tie to the House of Burgundy. Roger quickly sought another bride, and his eventual choice of Beatrice of Rethel shows his continued interest in the Crusader states.⁸ Beatrice came from an esteemed family that could trace its lineage to Charlemagne and had substantial power in the Latin East: her cousin was Melisende, the queen of the Kingdom of Jerusalem. The marriage of Roger and Beatrice in 1151 came during a time of steep divisions in Jerusalem, as Melisende and her son Baldwin wrestled for the loyalty of elites in the Crusader states who were shaken by the cataclysmic failure of the Second Crusade. In view of Roger’s earlier attempts to insert himself into the affairs of the Crusader states, it is likely that he saw this marriage as a way of furthering his power in the Kingdom of Jerusalem given the right circumstances.

An opportunity for Roger II to insert himself in the Crusader states did not materialize in the early 1150s. On the opposite edge of the Mediterranean, however, seismic political changes occurred that would have a lasting impact on Norman Africa and the deposed Zirid emir al-Hasan ibn ʿAli. During the 1140s and early 1150s, ʿAbd al-Muʾmin, the caliph of the Berber dynasty known as the Almohads, established his authority in the Maghreb and much of al-Andalus by overthrowing the ruling Almoravids. The Almohads saw themselves as unique among the dynasties of the Maghreb. They rejected much contemporary Maliki jurisprudence and instead sought to return to two of the oldest sources of Islamic law, the Qurʾan and hadith.⁹ They also discouraged pilgrimage to Mecca, instituted a new call to prayer, minted square coinage, and abolished the protected status of Jews and Christians (the dhimma), which had been a hallmark of Muslim governments for centuries.¹⁰ Without this protected status, Christians and Jews under Almohad rule lived a precarious existence, and many chose to emigrate even as the Almohads simultaneously negotiated treaties with Italian city-states (especially Pisa and Genoa) to allow Christian merchants to trade in their ports.¹¹ ʿAbd al-Muʾmin thus utilized “all the material, ideological and symbolic resources at his disposal, regardless of mutual contradictions” to legitimize his rule, a strategy that would have significant repercussions in Ifriqiya.¹²

During the 1150s, the Almohads rapidly expanded east from the Maghreb into western Ifriqiya.¹³ In 547H (1152–53), ʿAbd al-Muʾmin deposed the last Hammadid emir, Yahya ibn al-ʿAziz, bringing an end to the dynasty.¹⁴ These conquests proved a boon for the Hammadid’s prisoner, the Zirid emir al-Hasan ibn ʿAli. Following the Almohad capture of Bougie, ʿAbd al-Muʾmin made al-Hasan a “close companion” and bestowed on him a “high position” in the Almohad ranks.¹⁵ Although his exact title and role in the Almohad administration is unclear, al-Hasan’s presence at the conquest of Mahdia and his later role as the city’s co-governor gives some sense of his ability to ingratiate himself with the court of ʿAbd al-Muʾmin. Given the Almohads’ strict adherence to Sunni Islam, it is possible (even likely) that al-Hasan abandoned his Shiʿa heritage in favor of his rescuers’ Sunnism.

The rapid conquests of the Almohads did not escape the attention of other rulers in Ifriqiya, who were wary of the prospect of governance from another outside power. Following the defeat of a tribal coalition at the hands of the Almohads in a battle near Bougie in late 547H (1152–53), a number of Arab tribes “from Tripoli to the furthest Maghreb” met to resist these newcomers from the Maghreb.¹⁶ Ibn al-Athir reports that Roger II encouraged this resistance and even offered five thousand of his own soldiers to fight with this coalition on the condition that they provide hostages to him. Tribal leaders refused this offer, however, saying, “We do not ask for help from anyone other than Muslims.”¹⁷ The Almohads met this coalition at the Battle of Sétif in the spring of 548H (1153) and emerged victorious, taking many of the resisters’ wives and children back to Marrakesh.¹⁸

Despite the resistance that ʿAbd al-Muʾmin encountered on this campaign in the eastern Maghreb and western Ifriqiya, he showed leniency to those who had opposed him. Almohad administrators sent letters to the defeated Arab leaders, telling them that they should come to the Maghreb to be with their families. Upon receiving these messages, many of them indeed headed west to the Almohad court in Marrakesh. ʿAbd al-Muʾmin treated these tribal leaders well on their arrival and returned their families to them, which caused them to remain loyal to him. This was a strategy common among the Almohads during their early conquests. This nascent dynasty relied on “the integration into leadership positions in the Empire of all defeated or converted enemies from previous overturned regimes” for its survival.¹⁹ Through this strategy, the Almohads were able to ensure the integration of tribes in the Maghreb and Ifriqiya that had once been united against them.²⁰

Soon after the Battle of Sétif, Roger II looked to create a buffer zone against the ascendant Almohads by expanding his own lands westward. He sent a navy under the command of his eunuch Philip of Mahdia to the city of Bône. With the aid of unspecified Arabs, Philip conquered the city in the autumn of 1153, took many of the inhabitants captive, and appointed a member of the Banu Hammad to govern the city.²¹ In the same year, likely on the same expedition, the Normans once again seized the Kerkennah Islands and attacked Djerba, whose population had presumably revolted against Norman rule.²² The timing of Philip’s attacks on the Kerkennah Islands and Djerba may have been linked to the Almohad campaigns that immediately preceded them. Perhaps the Muslim inhabitants of these islands thought the time was right to revolt against the Normans and pledge their loyalty to the Almohads (as the people of Tripoli would do several years later)? In any case, the Normans managed to suppress these revolts. Philip of Mahdia, though, met an unfortunate end shortly after returning from this successful campaign; he was tried and executed in late 1153 on charges of being a crypto-Muslim and showing leniency to certain families of Bône.²³

Although Norman territories in Ifriqiya had remained intact during the Almohads’ first excursion into the region, trouble was brewing for the Normans in the eastern Mediterranean. In the last few years of Roger II’s reign, the peaceful relationship between the Normans and the Fatimid caliphate soured. The deaths of the Fatimid imam-caliph al-Hafiz, his ex-vizier Bahram, and George of Antioch in the late 1140s/early 1150s destabilized the productive relationship that had defined the two dynasties in decades past. The next generation of Fatimid officials had a less generous view of the Norman administration, due in no small part to the ascendency of a Zirid prince in the Fatimid regime.²⁴ In chapter 3, I recounted the death of the emir Yahya ibn Tamim in 1116 and the steps his successor ʿAli took to ensure his own safety. ʿAli sent possible conspirators Abu al-Futuh, his wife Ballara, and their son al-ʿAbbas ibn Abi al-Futuh into exile at the Fatimid court at Cairo. The account of the Fatimid court provided by the twelfth-century writer and politician Usama ibn Munqidh relates their fortunes in Egypt.²⁵

Following the death of Abu al-Futuh in an unspecified year, his widow Ballara married Ibn al-Sallar, the Fatimid vizier. By virtue of this marriage, al-ʿAbbas ibn Abi al-Futuh became the stepson of one of the most powerful individuals in the Fatimid administration. Tensions between Ibn al-Sallar and other high-ranking Fatimid officials, however, led to his assassination in 1153 in a plot that was masterminded partly by his own stepson (and Usama ibn Munqidh himself). Through these machinations, al-ʿAbbas ibn Abi al-Futuh, the son of Ballara and the Zirid prince Abu al-Futuh, became the Fatimid vizier.²⁶ The presence in the Fatimid court of this Zirid noble, who must have known about Norman conquests in Ifriqiya, facilitated a break between the courts of Palermo and Cairo. Although al-ʿAbbas was vizier for only a year (he was killed by Frankish soldiers while fleeing Cairo in the summer of 1154), his presence in the Fatimid government nonetheless contributed to its deteriorating relationship with the Kingdom of Sicily.

The Troubled Reign of William I

Roger II appointed his son William as coruler and heir to the throne on Easter 1151 in a ceremony conducted by the archbishop of Palermo without the approval of the papacy.²⁷ When Roger died in February 1154, William I became the sole ruler of the Kingdom of Sicily—a polity that was powerful, wealthy, and coveted by a host of powerful enemies. Pervasive conflict during William’s reign has earned him the endearing epithet “the Bad,” which is perhaps unfair given the circumstances of his rule.²⁸ From the beginning of his reign, William was caught between a rock and several hard places. The papacy refused to recognize the validity of his rule, the Byzantines were planning to invade southern Italy, the friendship between the courts of the Normans and Fatimids had soured, and there was unrest in Djerba and the Kerkennah Islands. The troubled relationships that Roger II had with his various rivals came crashing down on William in the early years of his reign, and, when combined with the ascendency of the Almohads, would lead to the destruction of the Norman kingdom of Africa.

Mounting animosity between the Normans and Fatimids culminated in a Norman raid against the Fatimid city of Tinnis around 1154.²⁹ This raid shows that the policy of nonintervention between the Fatimids and Normans during the reign of Roger II had given way to animosity, and that Norman ships were no longer safe along the southeastern shores of the Mediterranean. Although this falling-out began in the early 1150s due to the deaths of prominent officials in both administrations and the ascendency of a Zirid prince as Fatimid vizier, it first led to military conflict in this raid against Tinnis during the first year of the reign of William I. The policy of neutrality in Ifriqiyan affairs that the Fatimids had adopted during the 1130s and 1140s was now no longer guaranteed. Fortunately for the Normans, however, political unrest in the Fatimid administration meant that the imam-caliph and his armies were in no position to directly threaten lands in the Kingdom of Sicily.³⁰

In southern Italy, the threat of violence was very real. During the first two years of William I’s reign, the Byzantines launched a fleet against southern Italy in retribution for years of raiding and conflict across the eastern Mediterranean. Around the same time, Norman nobles revolted against William with the encouragement of Pope Adrian IV, who did not recognize the legitimacy of the new Norman king and, thus, the Kingdom of Sicily.³¹ William defended his lands in largely successful campaigns across southern Italy in the mid-1150s and even conducted retaliatory raids against Byzantine territories in the eastern Mediterranean. Following these victories, he was able to reach peace settlements with his enemies. William and Pope Adrian IV signed the Treaty of Benevento in 1156, which recognized the legitimacy of the Norman king.³² Several years later, William reached made peace with the Byzantines as well.

William’s focus on affairs in Italy led to the apparent neglect of his African possessions. At the beginning of 1156, while William was campaigning in southern Italy, a series of popular uprisings in Ifriqiya resulted in the Normans losing many of their coastal bases. The impetus for these revolts is complex and requires some explanation. Chapter 5 discussed how the Normans brought about societal changes in Ifriqiya that favored Christians at the expense of Muslims. Although this sea change contributed to uprisings against Norman rule, it was not the only motivating factor. As the economy of Ifriqiya rebounded from the devastating drought of the 1140s, Muslim jurists presented legal arguments to justify revolt against their non-Muslim overlords. Local governors, too, had been raised in a tradition of opportunistic holy war against Christians, which easily justified violence against the Normans. This movement to revolt was further buttressed by Almohad conquests across northwest Africa, which provided some Ifriqiyan lords with a Muslim dynasty to which they could pledge their loyalty.

Long-held Islamic legal perspectives were a driving factor in motivating the revolts against Norman rule in the mid-1150s. In both Ifriqiya and the Maghreb, the dominant school of Islamic jurisprudence, or fiqh, was the Maliki school—a vast body of legal rulings and customs derived from the work of the eponymous Malik ibn Anas.³³ Of relevance to Norman Africa is the body of Maliki literature concerning how Muslims ought to live when their lords are not Muslims. The legal rulings of the twelfth-century jurist al-Mazari provide the most descriptive analysis of this topic, written around the time of the Norman kingdom of Africa.

Al-Mazari was a Sicilian jurist who immigrated to Ifriqiya during the late eleventh century.³⁴ After studying under prominent Maliki jurists at Sfax and Sousse, he became head of a judicial school in Mahdia and wrote a number of fatāwā on issues pertaining to Muslims living under and interacting with Christians, which survive today in a later compilation by al-Wansharisi (d. 1508).³⁵ As is customary in this genre of legal text, the fatāwā of al-Mazari take the form of questions posed to and answered by him. In one legal opinion, for example, al-Mazari was asked to adjudicate a dispute between three men who had formed a partnership to travel from Ifriqiya to Sicily to sell their goods.³⁶ After strong winds forced the ship back to Mahdia, one member of this partnership declared that this voyage had ended, a proposition that the other two men in the partnership did not support. Al-Mazari was asked to resolve this dispute, and he ruled that the business contract between the partners was finished and that the initial investments made by the three men should be returned to them.³⁷

Other opinions from al-Mazari concern questions about the legality of trade conducted in Christian lands (namely Sicily) and the legitimacy of judges appointed by non-Muslim rulers. In one case from the late 1080s or early 1090s that revolved around the metallic worth of coinage in Sicily and Ifriqiya (where standards for minting differed), al-Mazari condemned merchants traveling and conducting business in Norman Sicily because it was “illegal for Muslims willingly to travel to a country under infidel rule.”³⁸ This opinion was controversial, however, and scholars in Mahdia debated the legality and necessity of trading with Sicily—especially with regard to wheat, which was a cornerstone of commercial relations for merchants operating between Ifriqiya and Sicily.

In another legal opinion, al-Mazari was asked about the legitimacy of a Muslim judge handing down rulings in Christian Sicily.³⁹ At the heart of this legal opinion was the question of how ʿadala (justice as defined by Islamic law) ought to be administered when living under the rule of someone who exhibits ʿadawa (hostility to Islamic law). This was an important question in the context of Norman Sicily, for Muslims living on the island were subject to Christian administrators whose policies did not necessarily align with the standards of Islamic law. Al-Mazari ultimately decided that the rulings of such a judge should be accepted, but he also said that Muslims should not live in the dar al-ḥarb (house of war, i.e., a place governed by a non-Muslims) unless absolutely necessary.⁴⁰ If Muslims do live in this state, though, al-Mazari thought it was better that they accept the judgment of the local Muslim judge than have no legal rulings whatsoever. In other words, “even bad government was better than no government.”⁴¹

If we accept the rulings of al-Mazari to be broadly indicative of Maliki legal scholars living in Ifriqiya during the time of Norman dominion, we learn several important things. The first is that the rulings of Muslim judges that the Normans appointed in their cities (like Abu al-Hajaj in Tripoli) were valid even though Roger II and William I were not Muslim. Even though Abu al-Hajaj was living in the dar al-ḥarb, he could justly provide legal advice to Muslims in his community. The second is that the circumstances of Muslims living under Norman control in Ifriqiya were by no means ideal. While al-Mazari acknowledges that sometimes it is necessary to live in this poor environment, he asserts that Muslims should endeavor to either leave or revolt if they are able. This legal mindset thus legitimized accepting Norman rule when the situation required it, and then revolting against Norman rule when the opportunity presented itself.⁴²

If Maliki legal judgments were the primary form of judicial thought through which qadis and governors in Ifriqiya interpreted their submission to the Normans, why did they not resist or revolt against Norman rule sooner? The answer to this question is likely tied to drought, famine, emigration, and military conflict in Ifriqiya during the 1140s. Although the presence of Christian Normans was a bitter pill to swallow for many Muslims living in Ifriqiya, it stimulated the economy and allowed the region to modestly recover from the events of the previous decade.⁴³ Once cities had recovered economically, though, the advantages of living under Norman rule disappeared, thus paving the way for revolt.⁴⁴

This legal framework from the Maliki school was further buttressed by a tradition of opportunistic holy war in Ifriqiya. Across the twelfth century, local lords in Ifriqiya marshaled support for military campaigns against Christians from Muslim allies, sometimes using religious rhetoric to bolster their cause. The Zirid emirs ʿAli ibn Yahya and al-Hasan ibn ʿAli forged an alliance with the Almoravids against the Normans that took the form of raids against Norman lands in the 1120s. In 517H (1123–24), al-Hasan boasted of the Arab tribes waging “jihad” when they defeated the Normans at the Battle of al-Dimas.⁴⁵ In 548H (1153–54), a coalition of Arab tribal leaders rejected Roger II’s offer of five thousand Norman knights to fight alongside them against the Almohads because they wanted aid only from other Muslims.⁴⁶ After ʿAbd al-Muʾmin conquered Mahdia from the Normans in 554H (1159–60), he tried to sway local tribes to campaign with him in al-Andalus, promising jihad and the reconquest of lands that had been conquered “at the beginning of Islam.”⁴⁷

Contemporary sources do not make explicit all of the rhetoric that surrounded these alliances of Muslim lords in Ifriqiya against Christians. Nonetheless, there is explicit evidence that al-Hasan in 517H (1123–24) and ʿAbd al-Muʾmin in 554H (1159–60) used jihad to motivate Ifriqiyan tribes to fight on their behalf. This mentality of holy war was opportunistic, though, and it applied only when it suited the interest of its rulers. We cannot ignore the multiple occasions in which Muslim rulers in Ifriqiya worked alongside Christians to achieve their political ends. Rafiʿ ibn Makkan al-Dahmani of Gabès, for example, sought the protection of Roger II from Zirid aggressors in 511H (1117–18).⁴⁸ Al-Hasan ibn ʿAli likewise requested help from Roger when the Hammadids attacked Mahdia in 529H (1134–35).⁴⁹ And Yusuf of Gabès asked for the allegiance of Roger in 542H (1147–48) so that he could rule the city.⁵⁰

Rhetoric of jihad is nowhere to be found in these encounters. Instead, local governors threw in their lot with whatever power was best positioned to help their cause. In some cases, this involved making alliances with the Christian Normans. But when the situation called for fighting the Normans, the rhetoric of war changed to include a religious dimension not found in intra-Muslim conflict. This is not dissimilar to how legal scholars could use the writings of al-Maqrizi to justify both acceptance of and resistance to Norman rule when it suited them. During the 1150s, when the economy of Ifriqiya was stabilizing, these ideologies found common ground and the result was widespread revolts against Norman rule.

The above examples indicate opposition to the Christian Normans among Muslim governors and scholars. There is further evidence, however, of grassroots opposition to Norman rule among the Muslims of coastal Ifriqiya. When Yusuf usurped the governorship of Gabès in 542H (1147–48) and pledged his loyalty to Roger II, the agitated Muslim populace of the city worked with the army of al-Hasan to overthrew Roger’s newest vassal.⁵¹ A year later, Ibn ʿIdhari reports that the people of Tunis overthrew their governor for simply trying to trade with the Normans.⁵² These incidents, the details of which are unfortunately lost, speak to some level of grassroots animosity toward Norman rule among Muslims in Ifriqiyan cities. When revolts against the Normans occurred in the 1150s, they were led by Muslim governors but supported by these masses who opposed Norman domination.

Widespread distaste for Norman rule in Ifriqiya was likely reciprocated by members of the Norman elite who had contempt for the Muslims of Ifriqiya. Evidence for this mindset comes from the Tristia ex Melitogaudo, a poem written by an unknown member of the Norman nobility who sought to win back the good graces of George of Antioch and escape exile from the island of Gozo.⁵³ In this poem, the author demeaned the Muslims that surrounded him to gain sympathy from the court of Palermo:

To what end have I been flung in the midst of the trackless seas

where the children of godless Hagar [i.e., Muslims] live,

[I,] not enjoying even little comforts,

I, woe betide me, not drinking any wine,

[not] even for the good of my stomach.…⁵⁴

The author, surrounded by godless Hagarenes, is unable to enjoy the benefits of alcohol or other comforts of life. In another section of the poem, he laments that he cannot even speak to the Hagarenes without the use of interpreters, which only increases his sense of isolation:

What mind may be able to comprehend

the whole throng of evil happenings from which I, wretched me, suffered harm,

even if someone were to be well-versed in countless languages?

For who would endure to utter, in uncouth speech,

the unjust arrogances of the intemperate,

the insolence, [the] lawlessness of foreigners,

the hunger and thirst and seizure of things possessed,

sleeping-mats both of an unfortunate and of a well-adjusted [person],

and the change and the turned-back posture,

the manner of sitting and of rising up,

the unholy bridlings of words,

alas, the suppression of unallowed hymns?⁵⁵

The author of the Tristia ex Melitogaudo wrote this poem with a clear goal in mind: to escape exile from Gozo using all of the rhetorical tools at his disposal.⁵⁶ Although much of the poem is filled with laudatory allusions to the Greco-Roman world that bolstered the standing of the Kingdom of Sicily, the author also thought that outlining indigenous Muslim practices (diet, language, and worship) would win the sympathy of the Norman court by tapping into their anti-Muslim biases. He presumably reasoned that George of Antioch and his colleagues would read or listen to this poem and have real sympathy for anyone who had to spend so much time surrounded by Muslims. By tapping into some faction of high-ranking administrators who held unfavorable opinions about Muslims, the author thought he could escape his present situation. This disdain of some Norman elites for Muslims may have translated into quotidian interactions otherwise lost in contemporary sources that further contributed to widespread tension in Ifriqiya.

Revolts against Norman Rule in Ifriqiya

Momentum for an uprising against Norman rule was building in Sfax at the beginning of 551H (1156–57). Ibn Khaldun vaguely writes that Christians of Sfax had brought harm to Muslims, which spurred local unrest.⁵⁷ The governor of the city, ʿUmar ibn Abi al-Hasan al-Furriyani, capitalized on this anger by leading the people of Sfax in revolt. According to Ibn al-Athir, this revolt was almost inevitable for ʿUmar. When George of Antioch had initially conquered Sfax in 543H (1148–49), he wanted to install the father of ʿUmar, Abu al-Hasan, as governor. Abu al-Hasan, however, requested that his son be made governor, to which the Normans agreed only on the condition that they could take Abu al-Hasan as hostage. Before leaving for Sicily, Abu al-Hasan told his son to revolt against the Normans when “the right opportunity enables you.” Thus, in 551H (1156–57), ʿUmar ibn Abi al-Hasan al-Furriyani channeled the ire of his people and (if this apocryphal anecdote is believed) acted on the parting words of his father to revolt against Norman rule. Ibn al-Athir records his call to action: “ ‘Let a group from among you climb the walls and a group proceed to the housing of the Franks and the Christians, all of them, and kill them all.’ [The people of Sfax] said to him, ‘But our lord the shaykh, your father, we are frightened for him.’ He said, ‘He commanded me to do this. If thousands from among the enemies are killed for the shaykh, then he has not died.’ The sun had not risen before they killed the Franks to the last.”⁵⁸

This uprising in Sfax was successful in overthrowing Norman rule, presumably in a battle that involved the defeat and massacre of the city’s garrison and administrative officials. No contemporary sources describe these grisly details, although they do relate the fate of ʿUmar’s father in Sicily. When word reached William I of the Sfaxian revolt, he demanded that Abu al-Hasan write to his son and demand his return to Norman authority. Abu al-Hasan refused, saying that “one who has the audacity to do this will not return because of a letter.”⁵⁹ Undeterred, William sent an envoy to Sfax. On reaching the city, the envoy was not allowed inside. Instead, the following day, he watched as the people of Sfax carried a coffin outside the city in anticipatory memory of the doomed Abu al-Hasan. When William heard this, he had Abu al-Hasan crucified.⁶⁰

The revolt of ʿUmar ibn Abi al-Hasan al-Furriyani was the first in a succession of uprisings that occurred in Norman Africa in the mid-late 1150s.⁶¹ Abu Yahya ibn Matruh led a revolt in Tripoli, Muhammad ibn Rushayd instigated one in Gabès, and the islands of Djerba and Kerkennah followed suit. Of these revolts, only the one at Tripoli is described in any detail.⁶² Al-Tijani relates how the Normans demanded that Muslim religious officials in Tripoli denounce the Almohads from their pulpits—presumably due to some level of popular support for the Almohad cause. The officials, led by the qadi Abu al-Hajaj, refused to do so because it was not part of the contract between the city and the Normans. Although the Normans withdrew their demand, this interaction prompted Abu Yahya ibn Matruh and other nobles in the city to plan a revolt. On an unspecified day in 553H (1158–59), the people barricaded the streets and defeated a Norman garrison that rode out to suppress the uprising. Christians in Tripoli were arrested and Abu Yahya ibn Matruh became the independent governor of the city.⁶³

The only rebellion that the Normans were able to suppress during the late 1150s was in Mahdia’s suburb of Zawila. Incited by the revolt of ʿUmar at Sfax, the people of Zawila rose against Norman control and attacked Mahdia in late 551H (1157) with the aid of “Arabs from the region,” the people of Sfax, and unspecified others.⁶⁴ While these armies besieged Mahdia by land, King William I offered the first semblance of his resistance to these uprisings by supplying the defenders of Mahdia with reinforcements and a fleet of twenty galleys. His agents also managed to bribe several Arab tribes to defect when the Franks sortied out of the city.⁶⁵ The remaining attackers initially continued to fight, but when the Normans looked to surround them, the fighters from Sfax fled. The people of Zawila then attempted to retreat but found the gates of Zawila locked (potentially by their former Arab allies). The Normans won a decisive victory, pillaged the suburb, and killed those in it. The leniency that Roger II and George of Antioch had shown when they first conquered cities in Ifriqiya had disappeared under William I. Following the Norman victory at Mahdia, William settled Christians, including an archbishop, in Zawila in an apparent attempt to take more direct control of the city.⁶⁶ When the dust had settled at the end of 553H (1158–59), William had control over only Mahdia, Zawila, and Sousse, the last of which was the only city controlled by the Normans that did not revolt against their rule.⁶⁷

The Arrival of the Almohads

Amid widespread revolts against Norman rule, the Almohads had amassed an army and sought to complete the conquest of Ifriqiya they had started a decade earlier. Their arrival into the region, however, was met with a mixed response. One particular alliance, between ʿAbd Allah ibn ʿAbd al-ʿAziz of Tunis (a member of the Khurasanid dynasty) and Muhriz ibn Ziyad, who ruled over the nearby ruins of Carthage, stands out in showing the unpopularity of Almohad rule among some governors.⁶⁸ In 543H (1148–49) a council of elders in Tunis had appointed Muhriz ibn Ziyad to be the next governor of the city, but his appointment was rejected by the people of the city and he was forced to flee.⁶⁹ Nearly a decade later, in 552H (1157–58), an Almohad army led by one of the sons of ʿAbd al-Muʾmin besieged Tunis.⁷⁰ The Almohads were eventually defeated by reinforcements led by none other than Muhriz ibn Ziyad. A letter sent from the governor of Tunis to Pisa corroborates this victory (and discusses commercial advantages to be given to Pisan merchants by the city of Tunis).⁷¹ After the siege was broken, Muhriz even allowed the Khurasanids to continue governing Tunis. The prospect of rule by an outside group like the Almohads was so unwelcome that to stop their advance, Muhriz was willing to assist the very city that had previously shunned him.

This animosity toward the Almohads was not shared uniformly across Ifriqiya. In Tripoli, for example, the qadi Abu al-Hajaj refused to denounce the Almohads in 553H (1158–59), despite an order to do so from William I.⁷² That William felt the need to make this demand indicates anxieties in Palermo about the rising sympathy for the Almohad cause.⁷³ Others in Ifriqiya directly appealed to the Almohads for aid. In the aftermath of the failed revolt at Zawila, survivors fled to ʿAbd al-Muʾmin in Marrakesh and implored him to defeat the Normans.⁷⁴ The Almohad caliph distributed money to these refugees and promised that he would undertake this expedition on their behalf. The pleas of these Zawilan refugees were reinforced by those of their emir, al-Hasan ibn ʿAli, who remained in Marrakesh as a loyal adviser to the Almohad caliph and pressed him to capture Ifriqiya from the Normans. These factors were enough to persuade ʿAbd al-Muʾmin to set out with an army and navy in 553H (1158–59), secure in the knowledge that some of Ifriqiya’s governors and people would welcome his arrival.⁷⁵

The Almohad threat was apparent to William I even in the early years of his reign. William came to the throne shortly after the Battle of Sétif in 548H (1153–54), in which the Almohads had destroyed a coalition of Ifriqiyan forces and were well positioned to move further east into Norman-controlled territories. When William succeeded his father, he ordered the striking of a triumphal dinar in Mahdia that was similar to the one minted in honor of his father’s victory in Mahdia in 543H (1148–49), but with a key difference in honorific titles.⁷⁶ Whereas the text on Roger II’s coin refers to the dinar being struck “by order of the sublime king,” the text on William’s dinar reads that it was struck “by order of the Guide [al-Hādī] according to the command of God.”⁷⁷ This is a conspicuous difference that is likely due to the threat posed by the Almohads, whose imperial ideology centered on the self-proclaimed Mahdi (Guide) Ibn Tumart.⁷⁸ By appropriating an Almohad title for his own coin, William sought to present himself as the equal to this ascendant dynasty and, thus, to discourage local Ifriqiyan governors from defecting to this other “Guide” from the Maghreb. The widespread revolts of the mid-1150s against Norman rule proved this strategy ineffective, however, and it was not long after these insurrections that the Almohads came to end the Norman presence in Ifriqiya and establish their own control over the region.

When ʿAbd al-Muʾmin and his forces arrived on the western frontier of Ifriqiya in 554H (1159–60), they were not welcomed with open arms.⁷⁹ Tunis, which had already driven off the Almohads in 552H (1157–58), resisted once again.⁸⁰ The plans of the city’s defenders were foiled, though, when a group of nobles from the city covertly descended to ʿAbd al-Muʾmin under the cover of night and offered their surrender, which he accepted. The Almohad army seized half the goods of the people of Tunis (except for those of the surrendering nobles, who were granted immunity) and ordered that the Christians and Jews in the city either convert to Islam or be executed. The governor of Tunis, who had not been involved in these secret negotiations, was then exiled to Bougie. ʿAbd al-Muʾmin installed one of his viziers as governor before departing for Mahdia, where he arrived in late July/early August 1159.⁸¹

Other Ifriqiyan lords were more welcoming of the Almohads. Al-Tijani, quoting directly the Zirid prince Ibn Shaddad, glowingly describes how Arab tribal leaders and their people flocked to the support of the Almohads as they marched across western Ifriqiya.⁸² Ibn al-Athir likewise notes that advancing soldiers did not damage “an ear of corn” as they marched along the fields of Ifriqiya, and that all of the army prayed as one behind their imam. The swelling ranks of the Almohad army was an unwelcome sight to the Norman garrison at Mahdia, which retreated to the city’s walls at the arrival of ʿAbd al-Muʾmin. At Zawila, the Almohads marched freely into the suburb, which soon “became a flourishing town” as auxiliary forces of Berbers, Arabs, and other unspecified locals joined the army.⁸³

Abd al-Muʾmin, realizing it would take a prolonged siege to take the city, collected provisions of such quantity that they looked like two mountains of wheat and barley: a tangible reminder to both the besieged and the besiegers of the power of the caliph (and somewhat ironically, an indicator of the economic recovery that the region had experienced under Norman rule). But the Norman defenders of Mahdia did not give way to the Almohads. They conducted sorties from the city and made successful attacks on the front ranks of the siege camp.⁸⁴ ʿAbd al-Muʾmin responded by erecting a wall separating his camp from Mahdia to keep the Normans at bay. He also sailed around the peninsula with al-Hasan ibn ʿAli to assess the city’s fortifications and even asked (with apparent incredulity) how the Zirid emir had managed to lose such a well-fortified city. While the siege dragged on, the Almohads secured the loyalty of Tripoli, Sfax, Sousse, and inland Gafsa without conflict. The city of Gabès and a number of other unspecified territories, though, briefly resisted the Almohads before capitulating to their armies.

Around a month after the arrival of the Almohads, William I sent reinforcements to the city. In early September 1159, a fleet that had been fighting Almoravid pirates in Ibiza arrived outside Mahdia on the orders of the Norman king. Arabic chronicles recount that the Franks were horrified at the number of soldiers assembled outside the city and were subsequently defeated in a battle in the waters outside the city by the Almohad navy.⁸⁵ The outcome of this naval battle is corroborated by Romuald of Salerno, who mentions that the Norman commander Qaʾid Peter lost a naval battle to the Almohads at the cost of many galleys. Another Latin chronicler, Hugo Falcandus, even accuses Qaʾid Peter of being a crypto-Muslim and of treacherously retreating from combat because he was in in league with his coreligionists.⁸⁶

ʿAbd al-Muʾmin commemorated his victory by distributing some twelve thousand dinars and equipment to his men. The Norman fleet responded by raiding nearby Sousse and capturing the governor of the city in an apparent act of retribution for the city’s capitulation to the Almohads.⁸⁷ The defenders of Mahdia, meanwhile, resisted the Almohads for another six months before finally capitulating in January 1160.⁸⁸ The most detailed account of the city’s fall in the Latin tradition comes from Hugo Falcandus, who emphasizes the treachery in the court of Palermo that permitted the city to fall.⁸⁹ He accuses certain palace eunuchs in Palermo of sending letters to the Almohads that revealed that William I did not plan to send further assistance to the city. Falcandus also asserts that Maio of Bari, the successor to Philip of Mahdia, used Mahdia as a tool for destabilizing the rule of William I. Maio had evidently convinced the king that his popularity would be unaffected by the loss of Mahdia. Apparently, Maio believed that the people of Palermo would think the king had become mad for giving up such a prosperous city, revolt against him, and install Maio on the throne. Instead, Hugo Falcandus reports that many people in Sicily thought Maio “permitted [Mahdia] to be captured,” which caused the popularity of the king’s closest adviser to plunge.⁹⁰

Whereas Falcandus’s description of the fall of Mahdia is grounded in the internal politics of the Norman court, Arabic chroniclers instead emphasize the might and generosity of ʿAbd al-Muʾmin. The Almohad caliph made multiple peace offers to the Norman defenders in Mahdia that initially required them to convert to Islam. Although the defenders refused these offers, ʿAbd al-Muʾmin still allowed them to correspond with the court of Palermo. Eventually, though, he won over the defenders, and the garrison surrendered the city without having to convert.⁹¹ ʿAbd al-Muʾmin provided the defenders with ships to leave the city, but many drowned in the choppy winter waters of the Mediterranean and only a few reached Sicily. It is possible that the leniency shown to the Norman knights was partially due to warnings from William I, who had threatened the Almohad caliph with the massacre of Sicilian Muslims should his soldiers be harmed. Following his entrance into the city, ʿAbd al-Muʾmin organized the affairs of Mahdia and repaired its fortifications. He appointed an (unspecified) Almohad to rule the city along with al-Hasan ibn ʿAli, the Zirid ruler whom the Normans had ejected twelve years prior. Both al-Hasan and his sons received land around the city before ʿAbd al-Muʾmin departed for the Maghreb at the beginning of February 1160.

The Legacy of Norman Africa

The Almohad conquest of Ifriqiya resulted in a substantial realignment of the political landscape, whereby the entirety of north-central Africa came under the control of one dynasty. This upheaval did not sit well with a number of Ifriqiyan tribal leaders, who did not want to swear fealty to yet another outside power. Their reluctance to submit to Almohad rule led to conflict when ʿAbd al-Muʾmin requested aid from them in his campaigns against Christians in al-Andalus in early 1160. Although these Ifriqiyan lords initially agreed to follow him on this campaign, they later reneged. ʿAbd al-Muʾmin then sent a contingent of his army back to corral the deserters. The two sides met at the Battle of Horn Mountain in 555H (1160) near Kairouan, which ended in a decisive Almohad victory. Muhriz ibn Ziyad was killed in the battle, the goods and families of the deserters were taken to the Maghreb, and the surviving Arab tribesmen were forced to pledge their allegiance to ʿAbd al-Muʾmin. Following this victory, Ibn al-Athir reports that Ifriqiya was overwhelmingly “safe and peaceful” under Almohad rule.⁹²

Al-Hasan ibn ʿAli, the last Zirid emir, served as the co-governor of Mahdia and Zawila alongside an Almohad vizier for about three years.⁹³ After ʿAbd al-Muʾmin died in 1163, his son and successor Yusuf ordered al-Hasan to relocate to Marrakesh—presumably to ensure that al-Hasan did not seek to subvert his authority. En route to the Almohad capital, al-Hasan fell ill and died near Rabat in 563H (1167–68).⁹⁴ How al-Hasan had governed Mahdia during his final years as co-governor and the extent of his actual power in the city are not known. I suspect that his presence in the city was largely ceremonial—a physical manifestation of Almohad goodwill to the people of Mahdia, demonstrating that the new dynasty would honor the memory of the Zirids. This continuity of power would have helped to legitimize Almohad rule, particularly in comparison to the Christian Normans, and to minimize the risk of unrest. When al-Hasan was recalled to Marrakesh and later died, it brought an end to the Zirid dynasty.

Almohad rule across Ifriqiya realigned the balance of power in the region and eliminated any changes to its society that had been brought about by Norman rule. The Almohads’ policy of forced conversion to Islam reduced the indigenous Christian presence in the region.⁹⁵ Furthermore, the Almohads crafted a policy of forced resettlement that saw the relocation of tribal leaders and their families from Ifriqiya to the Maghreb, where they could live in comfort under the watchful eye of the Almohad court. This happened twice in a decade, once after the Battle of Sétif in 548H (1153–54) and again after the Battle of Horn Mountain in 555H (1160). The lenient treatment of these lords helped ʿAbd al-Muʾmin cultivate a loyal base of power that he could leverage to his advantage—as was the case in 551H (1156–57) when he persuaded Ifriqiyan chieftains to support his right to choose his son as heir to the Almohad throne. These tribesmen also provided military service. The skill of these Ifriqiyan soldiers was such that they were paid more than their Maghrebi counterparts: twenty-five dinars for an equipped Ifriqiyan cavalry, compared to only ten dinars for a Maghrebi cavalry.⁹⁶

The Almohad policy of forced migrations continued after the expulsion of the Normans from Mahdia and into the 1190s.⁹⁷ When the Almohad administration began to fragment in the late twelfth and early thirteenth centuries, these imported tribal leaders (and their descendants) took sides and drove infighting within the caliphate.⁹⁸ In Ifriqiya proper, local ambivalence with regard to the Almohads became apparent in the mid-1180s when a displaced Almoravid tribe from Majorca called the Banu Ghaniya invaded Ifriqiya and captured a number of prominent cities with ease: Bougie, Algiers, Constantine, Tozeur, Gafsa, and Gabès.⁹⁹ The Banu Ghaniya made alliances with Ifriqiyan tribes who were dissatisfied with Almohad rule, and together they managed to inflict multiple defeats on the Almohads, who sent several armies far afield to deal with this new threat. The Almohads eventually triumphed over the Banu Ghaniya in a number of battles in the early 1200s that eliminated the threat posed from this dynasty.¹⁰⁰

When the Banu Ghaniya waged their campaigns in Ifriqiya, they most often clashed with the Banu Hafs, a dynasty that the Almohads had appointed as governors of Ifriqiya during the 1190s.¹⁰¹ As Almohad authority in al-Andalus and the Maghreb faltered, however, the Hafsid governors of Ifriqiya, too, began to see themselves as separate from their Almohad overlords. Drawing on their hereditary connection to Ibn Tumart, the Mahdi and spiritual founder of the Almohad movement, the Hafsids removed the caliph’s name from their call to prayers at the beginning of 627H (1229) and subsequently introduced their own names into the khutbah in 634H (1236–37). The Almohad caliphate—stretched thin by repeated (and increasingly unsuccessful) campaigns in al-Andalus, politically fractured by repeated succession crises, and torn between various ideological, tribal, and legal divisions—was incapable of quelling this revolt. The Hafsids thus ruled Ifriqiya as one of several successor dynasties to the Almohads until the end of the sixteenth century.

The religious overtones that had characterized Almohad campaigns in Ifriqiya against the Normans did not prevent other Christian powers in the Mediterranean from seeking to trade in Almohad ports. Genoese merchants made treaties across the 1160s with the Almohads, the Banu Ghaniya in Majorca, and Ibn Mardanish in al-Andalus.¹⁰² The Pisans followed suit, not wanting a rival city-state like Genoa to have exclusive trading rights with Muslim powers across the western Mediterranean.¹⁰³ These contracts allowed foreign powers to establish funādiq (commercial spaces that provided essential services for visiting merchants) for the exchange of textiles, spices, foodstuffs, gold, and industrial goods.¹⁰⁴ Ifriqiya was just one front where Christian powers on the Italian Peninsula sought to extend their commercial reach in the mid-late twelfth century; Pisa concluded a favorable treaty with the Fatimids in 1154 and, despite the objections of religious officials, Fatimids and Pisans served as crew members on the same ships.¹⁰⁵ The lucrativeness of these trade routes was such that even the Normans were willing to set aside their prior conflicts with the Almohads.¹⁰⁶ In 1180, King William II of Sicily signed a treaty with the Almohad governor of Tunis, Abu Yaʿqub, to allow Norman traders to conduct commerce in their Ifriqiyan ports.¹⁰⁷ The terms of this treaty even favored William, who received tribute from Abu Yaʿqub—potentially to guarantee certain grain prices for shipment to Tunis.¹⁰⁸

In the Kingdom of Sicily, the loss of Mahdia fueled unrest and conspiracy during the reign of William I. The grand admiral of Sicily, Maio of Bari, was widely blamed for the fall of Norman Africa and became immensely unpopular. As related by the decidedly anti-Maio and anti-William chronicler Hugo Falcandus, rumors spread throughout Palermo that Maio sought to usurp the Sicilian crown for himself.¹⁰⁹ Falcandus reports that Maio had seized the royal regalia by conspiring with the queen consort and that he plotted with the help of the papacy to assassinate William himself. These rumors—though likely untrue—were enough to rouse a group of Norman elites to conspire to assassinate the grand admiral, which they did in November 1160 on the streets of Palermo. William’s troubles did not end with the death of Maio, however. Several months later, in March 1161, discontented rebels led by a handful of Sicilian nobles marched on the royal palace and captured William. While they were considering how to best deal with the captive monarch, the Norman army and people of Palermo rallied to William, demanding the release of their king. With this popular support, William suppressed the attempted coup and continued to rule the Kingdom of Sicily until his death from natural causes in 1166.

Although William I’s reign after this coup was largely peaceful, there remained troubling undercurrents of Christian-Muslim tension during it.¹¹⁰ After the execution of Philip of Mahdia near the end of the reign of Roger II, crypto-Muslim eunuchs of the royal court became the targets of increased scrutiny and suspicion. A series of grassroots anti-Muslim riots in 1161, led by Italian immigrants and intensified by the loss of Norman possessions in Africa, also shows “widespread popular resentment” among many Latin Christians toward Muslims in Sicily.¹¹¹ Perhaps counterintuitively, these pogroms initially led to a closer connection between the Norman monarch and his Muslim subjects, as persecuted Muslims on the island flocked to their ultimate overlord for protection. William I even appointed one of his crypto-Muslim eunuchs, Qaʾid Peter, the military commander who was partially responsible for the loss of Mahdia, as part of his intimate three-person advisory council, the familiares regis (familiars of the king).¹¹²

The relationship between William I and Muslims in Sicily was such that Hugo Falcandus notes that Saracen women lamented the most over the king’s death in 1166.¹¹³ Even after William’s passing, crypto-Muslim eunuchs continued to play a substantial role in the kingdom. Qaʾid Peter leveraged his proximity to the queen regent to elevate his position over other administrators as “chamberlain of the royal palace.”¹¹⁴ Peter’s position in the Norman court was nevertheless precarious. In the summer of 1166, whether due to his desire to return to his coreligionists or because of his fear of an internal plot, he defected to the Almohads. Although the Norman kings continued to appoint crypto-Muslim eunuchs to government posts after the defection of Qaʾid Peter, the safety of non-elite Muslims in Sicily became increasingly precarious at the end of the twelfth and beginning of the thirteenth century as dynastic crises fueled popular violence and anti-Islamic papal doctrines justified the removal of Muslims from Christian lands.¹¹⁵ By the end of the thirteenth century, when Sicily and southern Italy were controlled by Hohenstaufens, the indigenous Muslim community of Sicily was all but destroyed.¹¹⁶

A Deeper Mediterranean Legacy?

Norman conquests in Ifriqiya did not have enduring resonance in Latin Christendom. Latin chroniclers writing both within and outside the Kingdom of Sicily have precious little to say about these campaigns, especially compared to Arabic source materials. Only the work of Hugo Falcandus tackles the fall of Norman Africa, and this description is mostly concerned with dysfunction and intrigue at the court of Palermo. This paucity indicates that contemporaries (or near contemporaries) of Norman Africa viewed it as peripheral to other Norman military campaigns and political machinations of the time, or that so little information about Norman Africa reached the shores of Europe that these authors had little to write about it. Leaders who undertook campaigns in Ifriqiya after the fall of the Norman dynasty show a similar disinterest and do not recall the earlier campaigns of Roger II and George of Antioch.¹¹⁷ Louis IX of France did not mention the Normans when he invaded Tunis in 1270 as part of his second Crusade.¹¹⁸ Charles V did not evoke the Normans when he attacked Tunis in 1535.¹¹⁹ Zirid-Norman conflict likewise does not find a place in the narratives about crusading, holy war, or Christian-Muslim relations that were written in the late medieval and early modern periods.

The same cannot be said of medieval Arabic sources. Despite the ephemeral nature of the Norman conquests in Ifriqiya, Muslim authors interpreted them as one element of an enduring and present threat that Christians posed to Muslims on the southern shores of the Mediterranean.¹²⁰ Al-Kamil fi al-Tarikh (The complete history) by Ibn al-Athir and Kitab al-ʿIbar (The book of examples) by Ibn Khaldun present Norman involvement in Ifriqiya as part of a trans-Mediterranean assault on the lands of Islam. Both authors, drawing on source material ultimately deriving from the Zirid prince Ibn Shaddad, interweave narratives of Zirid-Norman conflict in Ifriqiya and contemporary campaigns of Christians fighting Muslims in other parts of the Mediterranean. Exchanges between the Zirids and Normans were thus mobilized by both authors as one component of a larger interfaith battle between Muslims and Christians.

Ibn al-Athir wrote Al-Kamil fi al-Tarikh, his longest and most famous work, while living in his family’s estate near Mosul.¹²¹ The chronicle presents a year-by-year summary of events in the Muslim world up to the year 628H (1230–31).¹²² This text is part of the genre called tārīkh: a “history,” but with the connotation of using events of the past as a way to teach by example—specifically, to teach Muslim elites who could learn from historical individuals and dynasties.¹²³ Ibn al-Athir outlines his approach to the genre in the introduction of his chronicle. He argues that previous histories had provided disparate narratives that disconnected the histories of the West (the Mediterranean and Middle East) and East (greater Iran and India).¹²⁴ By carefully combining these narratives into one work about the entire umma and scrutinizing previous histories, Ibn al-Athir thought he could create a chronicle that was closest to an objective truth.¹²⁵

One such truth made apparent in the sections that discuss the Norman conquest of Sicily in 482H (1089–90) to the end of the chronicle in 628H (1230–31) is the idea of a trans-Mediterranean Frankish assault on Islamic lands. Ibn al-Athir argues that the Franks sought to establish a dawla (state or dynasty) in al-Andalus, Ifriqiya, Sicily, and Syria through a series of interrelated campaigns that posed a substantial threat to the survival of the Muslim community.¹²⁶ It is within this larger framework that exchanges between the Zirids and Normans occur, and Ibn al-Athir makes explicit the interrelatedness of Norman aggression in Ifriqiya and other instances of Frankish aggression in the Mediterranean.¹²⁷

I recounted the first of these interrelated encounters in chapter 3. In 490H (1096–97), an envoy of crusader Baldwin I sent a message to Roger I that outlined his plans to invade Ifriqiya.¹²⁸ One of Roger’s advisers voiced his opinion that this was a great idea because then “the lands would become Christian lands.” On hearing this, Roger farted loudly and proclaimed, “My religion be true! That was more useful than your words!” He explained that should Baldwin take Ifriqiya, he would have to commit numerous resources to facilitate the conquests, he would lose money from the grain trade, and he would have to break his treaties with the Zirids. Although Roger refused Baldwin’s proposal, he ominously foreshadowed the future conquest of Ifriqiya at the end of the episode when he said, “When we find the strength, we will take it!” This apocryphal episode illustrates the extent to which Ibn al-Athir considered the Frankish campaigns in greater Syria and Ifriqiya to be related. Even though an alliance between Baldwin and Roger I never came to fruition, Ibn al-Athir thought his audience should know about this supposed instance of Frankish collusion.

The interconnectivity of Muslim-Frankish conflicts is seen some fifty years later, in 539H (1144–45), during an encounter that an unnamed scholar related to Ibn al-Athir.¹²⁹ When news arrived in Sicily that a fleet of Roger II had plundered the lands around Tripoli, the Norman king asked a respected Muslim scholar, “Where was Muhammad for those lands and their people?” The scholar replied, “He was conquering for them. He witnessed the conquest of Edessa, which the Muslims have now conquered.” Although members of Roger’s court laughed at this prophecy, news reached Sicily several days later of the Muslim seizure of Edessa. Like the story of Roger I’s interaction with Baldwin, this episode is not found in any other Arabic sources. Instead, it was a story that Ibn al-Athir heard from a Muslim scholar with some connections to Norman Sicily and thought prudent to record in his chronicle. Its presence in Al-Kamil fi al-Tarikh once again highlights the perceived interconnectivity of conflicts between Muslims and Christians through the eyes of Ibn al-Athir: a loss for the Muslims of Ifriqiya corresponded to a gain for the Muslims of Syria.

The language that Ibn al-Athir uses in Al-Kamil fi al-Tarikh further solidifies the idea of a monolithic Christian enemy—embodied through the use of the term “Frank,” which broadly means Latin Christians.¹³⁰ Although Ibn al-Athir names individual Frankish leaders, there is little variance in how he describes the people over whom they ruled or the Muslims against whom they fought.¹³¹ A leader of the Third Crusade fighting in the Latin East is a Frank, an Aragonese commander fighting in al-Andalus is a Frank, and a Norman commander fighting in Ifriqiya is a Frank.¹³² Through this rhetorical strategy, Ibn al-Athir reduces interfaith conflicts throughout the Mediterranean to this simple Muslim-Frankish dynamic. Ibn al-Athir’s use of the same term across all these exchanges shows that he perceived these Franks and their campaigns against Muslims to be interrelated, and he hoped to convey this to his thirteenth-century readers.

A similar strand of thought is found in Ibn Khaldun’s Kitab al-ʿIbar, which (as discussed in the introduction) has been the subject of more scholastic inquiry than any other text written in North Africa during the Middle Ages. Ibn Khaldun utilized a host of earlier sources, including the chronicle of Ibn al-Athir, to reinforce his idea of the cyclical nature of history.¹³³ This methodological framework leads Ibn Khaldun to treat Zirid-Norman conflict in two instances: the first as part of larger Frankish aggression in the Mediterranean and the second in interspersed narratives related to the histories of individual Ifriqiyan dynasties. In the latter section, the Franks (i.e., the Normans) have a fleeting presence that is ultimately undone by the Almohads.¹³⁴ In the former, however, Ibn Khaldun echoes some of the narrative strategies of Ibn al-Athir by presenting Zirid-Norman conflict alongside other instances of Franks fighting Muslims in the Levant.

Ibn Khaldun first mentions Zirid-Norman interactions in the chapter titled, “An account of the Franks, who captured the coasts of Syria and its ports, and how they subdued them, and the beginning of their authority there and their fate.”¹³⁵ Sections in this chapter, beginning with the conquest of Jerusalem in 1099, consider the regions of the Islamic world threatened by Frankish aggression in the eastern Mediterranean, Sicily, and Ifriqiya. Al-Andalus, which Ibn Khaldun considers in different sections, is absent from this narrative; this represents a departure from Al-Kamil fi al-Tarikh of Ibn al-Athir. Nonetheless, Ibn Khaldun advocates for a broad Frankish assault on Islamic lands in the central and eastern Mediterranean, which include Ifriqiya.¹³⁶ This is seen in both the narrative he tells and the vocabulary he uses to tell it. The section headings of the Kitab al-ʿIbar show how he weaves the Normans’ campaigns in Ifriqiya (italicized) with roughly chronological events in the Levant:

The lord of Damascus conquers Banias

Shamas al-Malik seizes Beaufort

The Franks seize the island of Djerba in Ifriqiya

The lord of Damascus conquers some of the forts of the Franks

The Franks seize Tripoli of the West

The Franks seize Mahdia

The Franks seize Bône and the death of Roger, lord of Sicily, and reign of his son William

The Franks seize Ascalon

The uprising of the Muslims on the coast of Ifriqiya against the Franks, who had been triumphing there

The reconquest by ʿAbd al-Muʾmin of Mahdia from the hand of the Franks

The Franks siege Asad al-Din Shirkuh at Bilbeis

The Franks siege Cairo¹³⁷

Ibn Khaldun’s narrative alternates between events in Ifriqiya and the Mashriq. The core of the Norman conquests is told in the sections on the Franks conquering Tripoli, Mahdia, and Bône, though these are separated from other events that involve their presence in Ifriqiya: the conquest of Djerba and the Almohad reconquest. The intertwining of these events with those in the Middle East, combined with the use of the term “Franks” to refer to these people as a collective, shows the perceived interrelatedness of these military campaigns.

Ibn Khaldun’s rhetoric when describing the Almohad conquest of Mahdia is also evocative. He writes of the “reconquest by ʿAbd al-Muʾmin of Mahdia from the hand of the Franks.”¹³⁸ This word choice is curious, for neither ʿAbd al-Muʾmin nor the Almohads ever held Mahdia before this conquest. Instead, such language implies that the Almohads retook Mahdia on behalf of all Muslims from the Franks, whose conquests merely interrupted Muslim rule.¹³⁹ This rhetoric delegitimizes Norman rule as little more than an interlude between two more legitimate dynasties: the Zirids and the Almohads. Writing in the fourteenth century, Ibn Khaldun thus frames this centuries-old encounter between the Normans and Zirids/Almohads as one of many examples of Frankish-Muslim conflict in the Mediterranean.

Norman rule in Ifriqiya was short lived. When William I came to power in 1154, rebellious lords and foreign powers like the Byzantines sought to conquer his lands. With his military resources stretched across southern Italy, William was unable to combat widespread revolts in Ifriqiya that occurred during the mid-1150s. These uprisings, which were driven by opportunistic governors backed by Maliki jurists and a disaffected Muslim majority in Ifriqiya, resulted in the Normans losing the vast majority of their Ifriqiyan possessions by 553H (1158–59). It took an outside force in the form of the Almohads under ʿAbd al-Muʾmin to finally displace the Normans from Mahdia, Zawila, and Sousse a year later.

The Almohads dramatically altered the political landscape of Ifriqiya. They deported tribal leaders to areas closer to their capital of Marrakesh, which ensured that Almohad supremacy in Ifriqiya was relatively uncontested. Nonetheless, external invasions from the Banu Ghaniya during the late twelfth and early thirteenth centuries strained regional infrastructure and emboldened the Hafsid governors of Ifriqiya, who soon after declared their independence. Although indigenous Christians in Ifriqiya suffered during the Almohad conquests of the 1150s, Italian (and later, Norman) merchants prospered as they flocked to Almohad ports across the southern Mediterranean. In Sicily, William I managed to hold onto his throne despite the loss of his African territories. Nonetheless, tensions between Christians and Muslims that had manifested in conflicts during the 1140s and 1150s in Ifriqiya likewise cropped up in Sicily. Grassroots pogroms against Sicilian Muslims in the 1160s were the first in a series of popular persecutions that contributed to the decimation of these communities, a process made complete with the help of royal and papal doctrine in the thirteenth century.

While the legacy of Norman involvement in Ifriqiya was virtually nonexistent in Christian Europe, it became one component of a larger narrative in Arabic chronicles. Authors like Ibn al-Athir and Ibn Khaldun, who wrote sweeping histories of the Muslim world, used Zirid-Norman conflict to reinforce the idea of a Frankish threat to Muslim lands across the Mediterranean. Although the geographical and chronological scope of Norman involvement in Ifriqiya was minor compared to that of larger Frankish campaigns in al-Andalus and the Mashriq, these authors nonetheless saw it as a relevant third theater of conflict. By folding Zirid-Norman interactions into this larger narrative, Ibn al-Athir and Ibn Khaldun ensured both dynasties’ continued relevance to Muslim history into the late medieval period.

________________

1. 1. Roger also intervened in German affairs by hosting Welf VI, a rival to Emperor Conrad III, in 1148/49 in the hopes that Welf would be sympathetic to the idea of causing trouble for Conrad. Houben, Ruler between East and West, 90.

2. 2. The details and chronology of this incident are unclear. I follow the narrative presented by Stanton, which rejects the story given in the Byzantine chronicle of John Kinnamos. Stanton argues that a Sicilian fleet gave assistance to a fleet of Louis VII and Eleanor of Aquitaine that was being attacked by a Byzantine navy. Then, Roger brought Louis and Eleanor to safety in Calabria and met with the French king for three days at Potenza in August 1149. Houben, Ruler between East and West, 93; Stanton, Norman Naval Operations, 98–102.

3. 3. Chalandon, Histoire de la domination normande, 1:146–47.

4. 4. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 544H, 11:73.

5. 5. Birk, Norman Kings of Sicily, 103–5.

6. 6. Hayes, Family, Faith, and Empire, 61–63.

7. 7. Jonathan Riley-Smith, “Family Traditions and Participation in the Second Crusade,” in The Second Crusade and the Cistercians, ed. Michael Gervers (New York: Palgrave Macmillan, 1992), 101–8.

8. 8. Hayes, Family, Faith, and Empire, 64–72.

9. 9. Maribel Fierro, “The Legal Policies of the Almohad Caliphs and Ibn Rushd’s Bidāyat al-Mujtahid,” Journal of Islamic Studies 10, no. 3 (1999): 227–28.

10. 10. Jean-Pierre Molénat, “Sur le role des almohades dans la fin du christiansme local au Maghreb et en al-Andalus,” Al-Qanṭara 18 (1997): 393–401.

11. 11. Mohamed Ouerfelli, “Personnel diplomatique et modalités des négociations entre la commune de Pise et les états du Maghreb (1133–1397),” in Les relations diplomatiques au moyen âge: Formes et enjeux (Paris: Éditions de la Sorbonne, 2011), 119–32.

12. 12. Pascal Buresi, “D’une péninsule à l’autre: Cordoue, ’Uṯmān (644–656) et les arabes à l’époque almohade (XIIe–XIIIe siècle),” Al-Qanṭara 31, no. 1 (June 2010): 7–8. For a discussion of the capacity for integration in the Almohad caliphate, see Pascal Buresi, “L’apogée almohade: La bataille d’Alarcos et son contexte historique,” in Averroès et l’averroïsme (XIIe–XVe siècle): Un itinéraire historique du Haut Atlas à Paris et à Padoue, ed. André Bazzana, Nicole Bériou, and Pierre Guichard (Lyon: Presses Universitaires de Lyon, 2005), 111–13.

13. 13. ʿAbd al-Muʾmin’s motivations for expanding east into Ifriqiya are explored in Baadj, Saladin, 53–54.

14. 14. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 547H, 11:79–80; al-Tijani, Al-Rihla, 247; Ibn Khaldun, Kitab al-ʿIbar, 6:223–24. Yahya ibn al-ʿAziz made plans to escape from North Africa via Sicily so that he could then travel on to Baghdad, but these plans were thwarted when ʿAbd al-Muʾmin sent him to the Maghreb. He died in the city of Sale in 557H (1161–62). Some of Yahya’s family may have succeeding in fleeing to Sicily; others attempted to defect but were executed. Idris, Berbérie orientale, 1:363–70; Simone, “Mezzogiorno normanno-svevo,” 286–87; Baadj, Saladin, 54–56.

15. 15. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 547H, 11:79.

16. 16. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 548H, 11:92–93; Ibn Khaldun, Kitab al-ʿIbar, 6:235–36.

17. 17. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 548H, 11:92.

18. 18. Lévi-Provençal, Trente-sept lettres, 26–34. On the Almohads’ campaigns in Ifriqiya, see Ibn Abi Zar, Al-Anis al-Mutrib bi Rawd al-Qirtas (Rabat: Dar al-Mansur, 1972), 198–200; Baadj, Saladin, 53–56.

19. 19. Buresi and El Aallaoui, Governing the Empire, 42.

20. 20. The importance of Ifriqiya to the larger Almohad plan of control in the western Mediterranean is further seen in the intimate involvement of their sovereigns, who participated in 75 percent of expeditions to the Maghreb and Ifriqiya but only 50 percent of those to al-Andalus. Pascal Buresi and Hicham El Aallaoui, “La chancellerie almohade,” in Los almohades: Problemas y perspectivas, ed. Patrice Cressier, Maribel Fierro, and Luis Molina (Madrid: Consejo Superior de Investigaciones Científicas, 2006), 498.

21. 21. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 548H, 11:93; al-Idrisi, Nuzhat al-Mushtaq, 1:291–92.

22. 22. Al-Idrisi, 1:305; Idris, Berbérie orientale, 1:376; Metcalfe, Muslims and Christians, 47, 102.

23. 23. The life and death of Philip of Mahdia has received substantial scholarly attention. Philip was a court eunuch in the Norman administration, one of many employed in Palermo in a conscious emulation of the Fatimid (and possibly Zirid) court. Through this system, the Normans employed Muslim converts to Christianity and placed them in high-ranking government positions. According to Ibn Jubayr, who traveled to Palermo in the 1180s, many of these “palace Saracens” secretly practiced Islam in defiance of their Christian masters. Ibn Jubayr, Rihla Ibn Jubayr (Beirut: Dar Sadir, 1964), 299–300. Philip of Mahdia was one such official. The legacy of Philip’s trial and execution is complex. While many scholars have seen this episode as laying bare anti-Muslim sentiment in the court of Palermo late in the reign of Roger II, recent work on the manuscripts documenting the fate of Philip (which were written in the thirteenth century) indicates that descriptions of the trial revolved around contemporary (thirteenth-century) issues rather than those of the 1150s. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 548H, 11:93; Romuald of Salerno, Chronicon, 7.1:234–36. See also Brian Catlos, “Who Was Philip of Mahdia and Why Did He Have to Die? Confessional Identity and Political Power in the Twelfth-Century Mediterranean,” Mediterranean Chronicle 1 (2011): 89; Johns, Arabic Administration in Norman Sicily, 212–34, 289–97; Takayama, Administration of the Norman Kingdom of Sicily, 73–114; Nef, Conquérir etgouverner, 588–91; Birk, Norman Kings of Sicily, 139–72; Theresa Jäckh, “Verbrechen und Strafe im normannisch-staufischen Königreich Sizilien: Der Prozess gegen Philip von al-Mahdiyya,” Deutsches Archiv für Erforschung des Mittelalters 76, no. 1 (2020): 23–60.

24. 24. Romuald of Salerno briefly mentions that “for his honor and benefit, [Roger] made peace with the King of Babylon.” This king of Babylon is probably al-Hafiz, the Fatimid caliph. The timeline and circumstances for this reconciliation, though, are vague. Romuald does not mention Roger being at war with the king of Babylon previously, and the nearest chronological event to this peacemaking is the death of John II Komnenos, the Byzantine emperor, which happened in 1143. Romuald of Salerno, Chronicon, 7.1:227.

25. 25. Usama ibn Munqidh, The Book of Contemplation: Islam and the Crusades, trans. Paul M. Cobb (New York: Penguin Classics, 2008), 26–34. See also Cobb, Usama Ibn Munqidh, 38–41.

26. 26. Carl H. Becker and Samuel M. Stern, “ʿAbbās b. Abi ‘l-Futūh,” EI2; Johns, “Malik Ifrīqiya,” 98–99.

27. 27. Houben, Ruler between East and West, 96.

28. 28. Graham Loud, “William the Bad or William the Unlucky?,” Haskins Society Journal 8 (1999): 99–113.

29. 29. The timeline of this raid is unclear. Ibn al-Athir records that a Sicilian fleet raided Tinnis in 548H (1153–54). The Tarikh Mansuri dates a Sicilian attack on Tinnis to 549H (1154–55). The Egyptian historian al-Maqrizi mentions a Sicilian raid against Tinnis, Rosetta, Alexandria, and Damietta in Jumada II 550H (August 1155) launched by “Roger son of Roger.” The continuation of the chronicle of Sigebert of Gembloux notes that William I launched a fleet to Tinnis in 1154. An 1156 letter from the Fatimid vizier Talaʾiʿ ibn Ruzzik to Pisa refers to a Sicilian attack on Tinnis in an unspecified year. These sources are considered in Amari, Storia dei musulmani di Sicilia, 3:433; Johns, “Malik Ifrīqiya,” 98.

30. 30. Brett, “Abbasids, Fatimids and Seljuqs,” 715–20.

31. 31. Falcandus, Historia o liber de regno Sicilie, 14, 20–21.

32. 32. Enzensberger, Guillelmi I diplomata, document 12. Pope Adrian IV had earlier addressed him as “lord.” Houben, Ruler between East and West, 166–67.

33. 33. Cottart, “Mālikiyya,” EI2.

34. 34. Summaries of the life of al-Mazari can be found in Charles Pellat, “al-Māzarī,” EI2; H. R. Idris, “L’école mālikite de Mahdia: L’imām al-Māzarī (m. 536 H/1141),” Études d’orientalisme dédiées à la mémoire de Lévi-Provençal (Paris: G. P. Maisonneuve & Larose, 1962), 1:153–63.

35. 35. Al-Mazari’s legal opinions are not restricted to issues relating to Christian-Muslim interactions. See, for example, his ruling on a commercial dispute between a Muslim and Jew in Gafsa. Camilla Adang, “A Fatwā by Al-Māzarī (d. 536/1141) on a Jewish Silk Merchant in Gafsa,” in Jewish-Muslim Relations in Past and Present: A Kaleidoscopic View, ed. Joseph Meri (Leiden: Brill, 2017), 162–71.

36. 36. Al-Wansharisi, Al-Miʿyar al-Muʿrib, 8:181–82. See also Davis-Secord, “Muslims in Norman Sicily,” 60–61.

37. 37. Unfortunately, the date of this ruling is unknown, though there are clues within the text as to its time of composition. This particular case likely dates to sometime in the late eleventh or early twelfth century.

38. 38. Davis-Secord, “Muslims in Norman Sicily,” 64.

39. 39. Al-Wansharisi, Al-Miʿyar al-Muʿrib, 10:107–8.

40. 40. Brett, “Muslim Justice under Infidel Rule,” 326.

41. 41. Metcalfe, “Muslims of Sicily,” 296.

42. 42. Al-Hasan’s decision to flee Mahdia rather than aid the Normans in their assault on Gabès was partially informed by this legal tradition, as the Maliki school held that it was not permissible in most situations for Muslims to aid non-Muslims in combat. Brett, “Muslim Justice under Infidel Rule,” 340; Alan Verskin, Oppressed in the Land? Fatwās on Muslims Living under Non-Muslim Rule from the Middle Ages to the Present (Princeton, NJ: Markus Wiener, 2013), 19–34; Michael Lower, “Christian Mercenaries in Muslim Lands: Their Status in Medieval Islamic and Canon Law,” in The Crusader World, ed. Adrian Boas (New York: Routledge, 2015), 420–25.

43. 43. Incoming grain from Sicily combined with a return to typical levels of rainfall helped the port cities of Ifriqiya bounce back from the events of the 1140s. OWDA 1149–60.

44. 44. Brett, “Muslim Justice under Infidel Rule,” 359.

45. 45. Al-Tijani, Al-Rihla, 338.

46. 46. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 548H, 11:92–93.

47. 47. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 554H, 11:121–22.

48. 48. Al-Tijani, Al-Rihla, 398–99; Ibn al-Athir, Al-Kamil fi al-Tarikh, year 511H, 10:247.

49. 49. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 529H, 11:17–18.

50. 50. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 542H, 11:61; Ibn Khaldun, Kitab al-ʿIbar, 5:232–33.

51. 51. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 542H, 11:61; Ibn Khaldun, Kitab al-ʿIbar, 5:232–33.

52. 52. Ibn ʿIdhari, Al-Bayan al-Mughrib, 1:349.

53. 53. The author of the Tristia ex Melitogaudo did not hold a high opinion of Roger II, which could have been a cause of his exile in the first place. In the poem, Roger is compared unfavorably to Dionysius, the Tyrant of Syracuse, and Herod. Busuttil et al., Tristia ex Melitogaudo, xxix–xxii, 69, 113, 171.

54. 54. Busuttil et al., Tristia ex Melitogaudo, 171.

55. 55. Busuttil et al., Tristia ex Melitogaudo, 69.

56. 56. On the possible identity of the author of the Tristia ex Melitogaudo, see Lauxtermann, “Tomi, Mljet, Malta,” 156–59; Christina Rognoni, “Leggendo l’anonimo Maltese: Alcune considerazioni su Giorgio di Antiochia,” Νέα ῾Ρώµη: Rivista di Ricerche Bizantinistiche 14 (2017): 315–31.

57. 57. Ibn Khaldun provides two slightly different accounts of the revolt at Sfax. One echoes that of Ibn al-Athir, almost to the word. The other, though, includes the statement that the Christians of Sfax brought harm to Muslims. These two descriptions are not necessarily incompatible. Ibn Khaldun, Kitab al-ʿIbar, 5:205, 6:224. Brett finds the second of these entries a more convincing explanation for the Sfaxian revolt. Brett, “Muslim Justice under Infidel Rule,” 356.

58. 58. Quotations from this encounter are from Ibn al-Athir, Al-Kamil fi al-Tarikh, year 551H, 11:100–101.

59. 59. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 551H, 11:100. See also al-Tijani, Al-Rihla, 75; Ibn Khaldun, Kitab al-ʿIbar, 5:237–38.

60. 60. To my knowledge, this is the only account of someone being crucified in Norman Sicily. Ibn al-Athir reports that Abu al-Hasan “did not cease to call out to God the Almighty until he died.” The overtly religious tones in this narrative could indicate that Ibn al-Athir was misrepresenting the fate of Abu al-Hasan so that he could be portrayed as a martyr. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 551H, 11:100.

61. 61. Ibn Khaldun, Kitab al-ʿIbar, 5:237–38.

62. 62. Al-Tijani, Al-Rihla, 241–42.

63. 63. Al-Tijani places this revolt as happening in 553H (1158–59), which runs counter to the narratives of Ibn al-Athir and Ibn Khaldun, who write of Tripoli’s uprising in 551H (1156–57). Circumstantial evidence makes me trust al-Tijani’s dating of Tripoli’s revolt. The presence of an Almohad army near Tunis in 552H (1157–58) may have caused William I to fear unrest in the city and demand the denouncement of the Almohads. Al-Tijani, Al-Rihla, 247–48; Idris, Berbérie orientale, 1:384.

64. 64. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 551H, 11:101.

65. 65. Ibn Khaldun, Kitab al-ʿIbar, 5:237–38.

66. 66. Robert of Torigni, Stephen, Henry II., and Richard I, 4:191.

67. 67. The fate of Sousse during these revolts is unrecorded in contemporary sources. Idris speculates that the Normans managed to retain control of the city until the arrival of the Almohads in 554H (1159–60). Perhaps its proximity to Mahdia allowed the Normans to retain greater control over Sousse than in other urban areas that revolted? Idris, Berbérie orientale, 1:394.

68. 68. Al-Tijani, Al-Rihla, 247–48. See also Ibn Khaldun, Kitab al-ʿIbar, 6:218.

69. 69. See footnote 105 in chapter 4.

70. 70. Ibn Khaldun, Kitab al-ʿIbar, 6:218–19; Ibn ʿIdhari, Al-Bayan al-Mughrib, 1:349–50.

71. 71. Mas Latrie, Traités de paix et de commerce, 23–26; Louis de Mas Latrie, “Documents sur l’histoire de l’Algérie et de l’Afrique septentrionale pendant le moyen âge: Relations avec Pise,” Bibliothèque de l’École des Chartes 5 (1848–49): 137–39. See also Silvia Orvietani Busch, Medieval Mediterranean Ports: The Catalan and Tuscan Coasts, 1100–1235 (Leiden: Brill, 2001), 223–25.

72. 72. Al-Tijani, Al-Rihla, 241–42.

73. 73. Two letters from the Almohad court written in 551H (1156–57) further indicate support for their campaigns, as they include a petition from Ifriqiyan leaders that the son of ʿAbd al-Muʾmin be made the governor of the region. Although these letters were written after the Almohad victory at Sétif, the leniency with which the Almohad caliph treated Ifriqiyan leaders spurred cooperation between them. The first of these letters does not contain a date, though Lévi Provençal argues that it can be dated to the beginning of 551H (1156). Lévi-Provençal, Trente-sept lettres, 55–66; Évariste Lévi-Provençal, Un recueil de lettres officielles almohades: Étude diplomatique, analyse et commentaire historique (Paris: Librarie Larose, 1942), 35–38.

74. 74. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 554H, 11:118; al-Tijani, Al-Rihla, 344–45.

75. 75. Chronicles are ambiguous about the time when these conquests were prepared and the strength of the force of ʿAbd al-Muʾmin. Idris, Berbérie orientale, 1:384–89. On the importance of Ifriqiyan ports to the Almohads, see Christophe Picard, The Sea of the Caliphs: The Mediterranean in the Medieval Islamic World, trans. Nicholas Elliott (Cambridge, MA: Harvard University Press, 2018), 170–74.

76. 76. Abdul-Wahab, “Deux dinars normands de Mahdia,” 215–17.

77. 77. Johns, “Malik Ifrīqiya,” 92–93.

78. 78. On Ibn Tumart’s troubled visit to Ifriqiya, see Idris, Berbérie orientale, 1:326–33.

79. 79. The details of these campaigns are considered in Lévi-Provençal, Trente-sept lettres, 99–121. See also Idris, Berbérie orientale, 1962, 1:378–79, 384–400.

80. 80. Ibn Khaldun, Kitab al-ʿIbar, 6:218–19; Ibn ʿIdhari, Al-Bayan al-Mughrib, 1:349–50.

81. 81. Ibn al-Athir dates the arrival of the Almohads at Mahdia to 18 Rajab 554H (August 5, 1159) while al-Tijani places it at 12 Rajab 554H (July 30, 1159).

82. 82. Al-Tijani, Al-Rihla, 346–47.

83. 83. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 554H, 11:119.

84. 84. Al-Tijani, Al-Rihla, 347–48.

85. 85. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 554H, 11:120; al-Tijani, Al-Rihla, 348–49.

86. 86. Romuald of Salerno, Chronicon, 7.1:242; Falcandus, Historia o liber de regno Sicilie, 25–28. Qaʾid Peter, also known as Peter Ahmad Barrun, was a leading figure in the court of William I during the early years of his reign. His life story is discussed in Johns, Arabic Administration in Norman Sicily, 222–28.

87. 87. The timeline for this raid is unclear and is mentioned only in al-Tijani. The governor of Sousse, Jabbara ibn Kamil al-Fadighi, was taken to Sicily and eventually ransomed. Al-Tijani, Al-Rihla, 30.

88. 88. Évariste Lévi-Provençal, Documents inédits d’histoire almohade: Fragments manuscrits du “legajo” 1919 du fonds arabe de l’escurial (Paris: Paul Geuthner, 1928), 200–202. See also Idris, Berbérie orientale, 1:394–400.

89. 89. Falcandus, Historia o liber de regno Sicilie, 24–28. Nef argues that the backlash against Maio for losing Mahdia is indicative of the personal investments in Africa held by many in Norman Sicily. Nef, Conquérir et gouverner, 601–7.

90. 90. Falcandus, Historia o liber de regno Sicilie, 25. Loud argues that there was an “atmosphere of suspicion, paranoia and factional dispute” in the court of William I during the time of the Almohad conquest of Mahdia. Falcandus, History of the Tyrants of Sicily, 23.

91. 91. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 554H, 11:120; al-Tijani, Al-Rihla, 349.

92. 92. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 554H, 11:122.

93. 93. Ibn al-Athir writes that al-Hasan was a governor of Mahdia, while al-Tijani specifies that he governed Zawila. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 554H, 11:120; al-Tijani, Al-Rihla, 349. Ibn Khallikan writes that al-Hasan exercised power in Mahdia with the aid of an Almohad nāʾib (“representative” or “lieutenant”). He also specifies that the Almohads provided al-Hasan with two farms in Mahdia and a house for his sons and entourage. Ibn Khallikan, Wafayat al-Aʿyan wa-Anbaʾ Abnaʾ al-Zaman, 6:218–19.

94. 94. Al-Tijani, Al-Rihla, 349–50; Tibi, “Zīrids,” EI2.

95. 95. Evidence exists for the limited survival of indigenous Christianity, especially near Gafsa. Brett, “Muslim Justice under Infidel Rule,” 360; Valérian, “Permanence du christianisme au Maghreb,” 149.

96. 96. Baadj, Saladin, 60.

97. 97. Amira K. Bennison, “Tribal Identities and the Formation of the Almohad Élite: The Salutory Tale of Ibn ‘Aṭiyya,” in Biografías magrebíes: Identidades y grupos religiosos, sociales y políticos en el Magreb medieval, ed. Mohamed Meouak (Madrid: Consejo Superior de Investigaciones Científicas, 2012), 245–72.

98. 98. Buresi and El Aallaoui, Governing the Empire, 78–80.

99. 99. Abun-Nasr, History of the Maghrib, 93–100.

100. 100. Conflict between the Banu Ghaniya and Almohads in Ifriqiya has led some later scholars to compare the former’s campaigns to those of the Banu Hilal during the 1050s. See, for example, Georges Marçais, “3ẖāniya,” EI2.

101. 101. Baadj, Saladin, 62–85.

102. 102. Epstein, Genoa and the Genoese, 38–53; Dominique Valérian, “Gênes, l’Afrique et l’Orient: le Maghreb almohade dans la politique génoise en Méditerranée,” in Chemins d’outre-mer: Études d’histoire sur la Méditerranée médiévale offertes à Michel Balard, ed. Damien Coulon, Catherine Otten-Froux, and Dominique Valérian (Paris: Éditions de la Sorbonne, 2004), 827–37; Hilmar Krueger, “The Routine of Commerce between Genoa and North-West Africa during the Late Twelfth Century,” in Latin Expansion in the Medieval Western Mediterranean, ed. Eleanor Congdon (New York: Routledge, 2013), 47–68. On this reorientation in the context of the southern Mediterranean, see Dominique Valérian, “Les relations commerciales entre Alexandrie et le Maghreb, XIe-XIIe siècle: de l’unité à la rupture?,” in Alexandrie médiévale 4, ed. Christian Décobert, Jean-Yves Empereur, and Christophe Picard (Alexandria: Centre d’Études Alexandrines, 2011), 229–38.

103. 103. Letters between the Pisans and Almohads reveal how both sides used a combination of personal trust, public institutes, and legal precedent to facilitate economic exchange—especially in the case of the crew of a Pisan ship looting and killing individuals in an Almohad port. Travis Bruce, “Commercial Conflict Resolution across the Religious Divide in the Thirteenth-Century Mediterranean,” Mediterranean Historical Review 30 (2015): 19–38.

104. 104. Constable, Trade and Traders, 68–106.

105. 105. For Muslim rulers, the commercial utility of Italian city-states clashed with the aid these polities provided to the Crusader states. Saladin even wrote a letter to Baghdad complaining of “contradictory behavior of the Pisans, Genoese, and Venetians.” Salvatori, “Corsairs’ Crews,” 39–40. See also David Jacoby, “Les italiens en Egypte aux XIIe et XIIIe siècle: Du comptoir à la colonie?,” in Coloniser au moyen âge: Méthodes d’expansion et techniques de domination en Méditerranée du 11e au 16e siècle, ed. Michel Balard and Alain Ducellier (Paris: A. Colin, 1995), 76–88.

106. 106. Commercial contracts between the Almohads and Italian city-states are more numerous in the thirteenth century than the twelfth century. Pascal Buresi, “Traduttore traditore: À propos d’une correspondance entre l’empire almohade et la cité de Pise (début XIIIe siécle),” Oriente Moderno 88, no. 2 (2008): 297–309; Pascal Buresi, “Les documents arabes et latins échangés entre Pise et l’empire almohade en 596–598/1200–1202: La chancellerie au cœur des relations diplomatiques,” in Documents et histoire: Islam, VIIe–XVIe siècle, ed. Anne Regourd (Geneva: Libraire Droz S. A., 2013), 13–88.

107. 107. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 576H, 11:225. See also Mas Latrie, “Documents sur l’histoire de l’Algérie,” 137–40; Louis de Mas Latrie, Relations et commerce de l’Afrique septentrionale avec les nations chrétiennes au moyen âge (Paris: Imprimeurs de l’Institut de France, 1886), 82–97; Abulafia, “Norman Kingdom of Africa,” 43–45.

108. 108. Giovanna Palombo, “The Normans of Sicily from ‘the Other Side’: The Medieval Arabic Sources,” in Sicily and the Mediterranean: Migration, Exchange, Reinvention, ed. Giovanna Summerfield and Claudia Karagoz (New York: Palgrave Macmillan, 2015), 40.

109. 109. Graham Loud, “The Image of the Tyrant in the Work of ‘Hugo Falcandus,’ ” Nottingham Medieval Studies 57 (2013): 1–20.

110. 110. In 1163, Frederick Barbarossa of the Holy Roman Empire sought to negotiate an alliance with Italian city-states to attack the Normans but was ultimately unsuccessful. Loud, “Norman Sicily in the Twelfth Century,” 457–58.

111. 111. Birk, Norman Kings of Sicily, 207. See also Falcandus, Historia o liber de regno Sicilie , 70; Romuald of Salerno, Chronicon, 7.1:246–47.

112. 112. Hiroshi Takayama, Sicily and the Mediterranean in the Middle Ages (New York: Routledge, 2019), 33–51.

113. 113. Falcandus, Historia o liber de regno Sicilie, 87–91.

114. 114. Tolan, Saracens, 194–96; Birk, Norman Kings of Sicily, 224.

115. 115. Nef, “Déportation des musulmans siciliens,” 457–65.

116. 116. See Taylor, Muslims in Medieval Italy, 1–32.

117. 117. Ibn ʿIdhari intriguingly mentions two attacks of the Franks in Ifriqiya in the decades after the Almohad conquest of Mahdia: one in 558H (1162–63) that hit Mahdia and Sousse and another that raided Mahdia in 573H (1177–78). Neither of these attacks are corroborated in other sources, and it is unclear where Ibn ʿIdhari heard of these attacks. Perhaps William I and his successor, William II, did seek brief retribution for the fall of Mahdia in 1160? Regardless, whatever attacks occurred during these years, they did not lead to lasting changes in the relationship between the Normans and the Almohads, as the two powers brokered a commercial treaty in 1180. Ibn ʿIdhari, Al-Bayan al-Mughrib, 1:350.

118. 118. Michael Lower, “Tunis in 1270: A Case Study of Interfaith Relations in the Late Thirteenth Century,” International History Review 28, no. 3 (September 2006): 504–14; Michael Lower, The Tunis Crusade of 1270: A Mediterranean History (New York: Oxford University Press, 2018), 144–73.

119. 119. Otto von Habsburg, Charles V, trans. Michael Ross (London: Weidenfeld and Nicolson, 1970), 133–48; James D. Tracy, Emperor Charles V, Impresario of War: Campaign Strategy, International Finance, and Domestic Politics (New York: Cambridge University Press, 2002), 7, 145–49.

120. 120. In considering the Zirids in this larger Mediterranean context, I echo the call of Hillenbrand and Cobb for scholars to consider these interfaith episodes from the perspective of non–Latin-Christian communities. This methodology also requires moving away from the Biblioteca arabo-sicula of Michele Amari, which, while convenient, isolates texts from the context in which they were written. Hillenbrand, Crusades, 613; Cobb, Race for Paradise, 279.

121. 121. The fifteenth-century scholar al-Sakhawi called Ibn al-Athir a “hadith expert and great scholar,” and The Complete History “the best of all histories in recording the happenings clearly and distinctly.” Franz Rosenthal, A History of Muslim Historiography (Leiden: Brill, 1952), 413.

122. 122. The annalistic format of The Complete History breaks in several places. For example, the account of the rise of Ibn Tumart, which is listed under the year 514H (1120–21), extends from 505H (1111–12) until 524H (1129–30). This break in form is likely due to Ibn al-Athir’s reliance on earlier sources that did not adhere to annalistic styles. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 514H, 10:266–70.

123. 123. Historians today are dubious of many of these genealogies. In his discussion of the genre, Brett states that tārīkh “is not in fact history; it is a peculiarly Islamic genre of miscellaneous information about humanity and the world, arranged with a passion for chronology which gives it simply the appearance of history in the Western sense.” Brett, “Way of the Nomad,” 252. Many Muslim scholars during the Middle Ages were disdainful of this methodology because they saw it as conflicting with the idea that human progress was preordained by God. François de Blois et al., “Tārīkh,” EI2. See also Fozia Bora, Writing History in the Medieval Islamic World: The Value of Chronicles as Archives (New York: I. B. Tauris, 2019), 11–23.

124. 124. Azmul Kamaruzaman, Norsaeidah Jamaludin, and Ahmad Fadzil, “Ibn Al-Athir’s Philosophy of History in Al-Kamil fi al-Tarikh,” Asian Social Science 11, no. 23 (2015): 30.

125. 125. Françoise Micheau, “Le Kitāb al-kāmil fī l-tā’rīkh d’Ibn al-Athīr: Entre chronique et histoire,” Studia Islamica 104/105 (2007): 84–85, 100–101. Ibn al-Athir primarily used Arabic sources for his narrative, though he also likely drew on Latin texts in his narration of the Crusades. Konrad Hirschler, “The Jerusalem Conquest of 492/1099 in the Medieval Arabic Historiography of the Crusades: From Regional Plurality to Islamic Narrative,” Crusades 13 (2014): 71–73.

126. 126. Cobb, Race for Paradise, 40.

127. 127. Ibn al-Athir was not the first to do this. As early as the first decade of the twelfth century, ʿAli ibn Tahir al-Sulami wrote about how the Norman conquest of Sicily was the first strike in a multipronged Frankish assault on the lands of Islam. Chevedden, “Crusade from the First,” 201–6. Chevedden’s argument is contested in Nef, Conquérir et gouverner, 47–49, 61–62.

128. 128. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 491H, 10:126–27.

129. 129. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 539H, 11:51–52.

130. 130. Variation in the usage of the term “Frank” means that it “never reached the status of an uncontested generic term for all Christian peoples of Western Europe.” König, Arabic-Islamic Views, 230.

131. 131. On rare occasions, Ibn al-Athir praises named Frankish leaders. This is true in the case of Roger II, whom Ibn al-Athir extolls for treating Muslims well in Sicily. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 484H, 10:92. See also Metcalfe, Muslims and Christians, 100; Nef, “Dire la conquête et la souveraineté,” 4–10.

132. 132. See, for example, Ibn al-Athir, Al-Kamil fi al-Tarikh, year 496H, 10:170–71, year 497H, 10:173–74, year 499H, 10:192–93, year 552H, 11:110–11.

133. 133. Circumstantial information in the Kitab al-ʿIbar shows that Ibn Khaldun used the works of Ibn al-Athir and al-Tijani as the basis for his own. In his first description of the Normans’ attack on Mahdia, Ibn Khaldun writes that Roger II assembled a fleet of 250 ships. In a different section of the book, though, he writes of the same attack but mentions that Roger launched a fleet of 300 ships. This discrepancy can be attributed to his use of multiple sources, for Ibn al-Athir wrote that Roger had a fleet of 250 ships and al-Tijani wrote that Roger had a fleet of 300 ships. Ibn Khaldun’s naming of George of Antioch similarly indicates that he used al-Tijani as a source. On several occasions, Ibn Khaldun introduces George as “George ibn Mikhael” and “George ibn Mikhael the Antiochene.” These titles are found in the writings of al-Tijani but not those of Ibn al-Athir. Ibn al-Athir, Al-Kamil fi al-Tarikh, year 543H, 11:63–64, year 544H, 73; al-Tijani, Al-Rihla, 241, 333, 341–42; Ibn Khaldun, Kitab al-ʿIbar, 5:233, 6:214. See also Charles Issawi, “Ibn Khaldun on Ancient History: A Study in Sources,” in Cross-Cultural Encounters and Conflicts (New York: Oxford University Press, 1998), 51–78.

134. 134. See, for example, Ibn Khaldun, Kitab al-ʿIbar, 6:223–24.

135. 135. Ibn Khaldun, Kitab al-ʿIbar, 5:209.

136. 136. This section on the Franks is flanked by narratives of the Seljuq and Artuqid dynasties.

137. 137. Ibn Khaldun, Kitab al-ʿIbar, 5:230–43.

138. 138. Ibn Khaldun, Kitab al-ʿIbar, 5:238. The crucial word in question, irtijāʿ, has the connotation of returning something to its previous form. Thus, the conquest of ʿAbd al-Muʾmin returns Mahdia (and Ifriqiya) to its previous form of governance by a Muslim dynasty.

139. 139. Al-Tijani similarly writes that Islam “returned” (ʿāda) to Mahdia. Al-Tijani, Al-Rihla, 349.