15
While Roger de Hauteville was completing the conquest of Sicily, messengers came from the maritime city-states of Genoa and Pisa, inviting him to join them in a strike against Mahdia. The populous city of Mahdia on the coast of modern Tunisia was the capital of the Zirid rulers of North Africa. And the attack was framed in religious terms, as a strike against the forces of Islam – a proto-crusade of sorts. Religious zeal was not the only motive, however. Pisa and Genoa were looking to muscle in on the lucrative trade between the western and eastern Mediterranean. Given Roger’s growing prominence and developing maritime expertise, he was the obvious choice of ally.
Roger’s advisers were keen to cash in on this opportunity. Or so the later Arab chronicler Ibn al-Athīr would have us believe. In his account, the canny Hauteville was unimpressed. He lifted a thigh and released a thunderous fart, before proceeding to explain why such flatulence was better counsel than that proffered by the wiser heads at court. For as Roger observed, if the Normans were to join the venture, they would soon find themselves in the worst of all possible worlds. If they succeeded, the profits would go to their northern allies, while in the case of failure, the Normans would bear the brunt of the repercussions. Moreover, they would end up footing much of the bill for supplying the expedition. Roger had recently concluded a peace with the Zirid ruler, Tamīm ib Mu’izz; he was not going to risk this on a speculative venture.
Ibn al-Athīr’s account was written almost a century and a half after these events, and he confuses certain details. He has the second ruler of Jerusalem, Baldwin of Edessa, as the driving force behind the plan, not the city-states of Pisa and Genoa; and he seems to think that Roger was an independent ruler, not an agent (however nominal) of his elder brother Guiscard. By his description of Roger’s behaviour he clearly means to underscore the barbarism of the new Norman lords of Sicily. Still, al-Athīr did not invent the tale from scratch. He draws on earlier sources, often of high quality. And the attack on Mahdia was very real, representing an important precursor to the First Crusade. We know from other sources that Roger declined an invitation to participate in the invasion. There is, therefore, every reason to believe the gist of his account.1
What al-Athīr’s earthy portrait reveals is the degree to which Roger and Robert had been thrust into the complex political world of the Mediterranean, in which the rulers of North Africa were prominent players. Ibn al-Athīr sought to write a comprehensive history, which from his perspective meant including everything that touched on the world of Islam.2 By the 1080s, this evidently included Roger. And in years to come, Roger and his son and successor, Roger II, would figure ever more prominently in the politics of North Africa.
Sicily and North Africa were of great strategic and commercial significance. Despite the dislocation caused by the fall of the Western Roman Empire in the fifth and sixth centuries, the Mediterranean continued to be an important conduit of trade (and information) between Europe, North Africa and the Middle East. By the eleventh century, the volume (and value) of this commerce was rapidly rising. Italian city-states such as Pisa and Genoa now sought – with considerable success – to control trade across the north-western Mediterranean, while the Byzantines, Venetians and Fatimids cornered the market in the east.3 The importance of Sicily and the Islamic province of Ifrīqya (encompassing modern Tunisia, as well as parts of Algeria and Libya) lay in their position between these two spheres. Whoever could control them, stood to make a tidy profit.
Thanks in part to many years of Islamic rule, Sicily and North Africa were closely integrated. The Sicilian capital, Palermo, lay closer to Mahdia than Rome. And Sicily now supplied grain to North Africa, providing a vital lifeline to a province which suffered increasingly frequent droughts. When Roger began to tighten his grip on Sicily, the natural question became how this trade would be affected.4 As with his response to the Genoese and Pisans, Roger approached the subject with his typical pragmatism. If there was profit to be made, he wasn’t one to rock the boat. Indeed, both he and his son sought to come to terms with the Muslim and Greek communities within and beyond their domains. They favoured Christians over Muslims (and the Latin rite over the Greek one), but welcomed and maintained contacts with people of all backgrounds. One of the leading advisers of Roger II was the Syrian Christian George of Antioch, who’d previously served the Zirid rulers of Ifrīqya; another was Phillip of Mahdia, who succeeded George to the post of amir/admiral (Latin: amiratus).5 Such men were known as Roger’s ‘palace saracens’. Conversion was formally a requirement for service, but this may often have been skin-deep. At Roger’s court, loyalty trumped religious faith – and profit trumped nativism.
As Norman rule became more firmly established, the rulers of Sicily began to assert themselves in the Mediterranean. Roger I had already taken Malta in 1091. Norman control was soon ousted, but was reinstated by Roger II in 1127.6 Malta’s position between Sicily and Ifrīqya made it a natural staging post to these North African domains. But before Roger could build on such success, he was forced to turn his attention to affairs on the Italian mainland.
The fact that an independent Norman strike against North Africa was even conceivable reveals how much had changed over the past half-century. In the 1080s, Roger I was still busy securing Sicily. Any further Norman conquests looked likely to take place in the Adriatic, where Guiscard and Bohemond were now heavily invested. Roger had only countenanced attacking Mahdia as part of a powerful maritime coalition – and soon dismissed the idea. Since then, the balance of power within the Norman domains of southern Italy had changed decisively. Following Guiscard’s death in 1085, Roger lent his support to his nephew Roger Borsa (Robert’s designated heir). But the latter never commanded the same respect as his father. Guiscard had already faced a series of serious revolts in Apulia; and the succession of the young and relatively inexperienced Borsa simply fanned the flames.
Roger Borsa eventually secured his position, but he remained heavily dependent on his uncle. Roger I’s death and the succession of his own young sons, Simon and then Roger II, did little to change this relationship. The mother of the latter two, Adelaide del Vastro – the future queen of Jerusalem – led a successful regency, ensuring that Simon and Roger’s domains survived intact. In 1111, Borsa died, leaving the young William II on the ducal throne. Roger II and William II were roughly the same age, but the former soon emerged as the more powerful political force. William enjoyed precedence as duke, but this was increasingly nominal. And when William died in 1127 without legitimate offspring, Roger was quick to secure direct control of Calabria and Apulia, pre-emptively declaring himself William’s heir.7
The effect of these developments was that power within the Italo-Norman realm shifted south and west, from Bari, Brindisi and Venosa to Salerno and Palermo. If Guiscard had looked across the Adriatic to the Balkans, his nephew gazed south across the Mediterranean to North Africa. Yet the expansion of Roger’s domains also brought considerable complications. His succession to the mainland duchy was hotly contested, not least by the pope, who since the accords of Melfi claimed the right to invest the duke. The situation was exacerbated by Roger’s decision to claim royal status in 1130. If his seizure of ducal authority had been presumptuous, Roger’s claim to kingship was seen as adding insult to injury.
Others outside Italy were also disquieted by these developments. The Byzantine emperors – who had a residual claim to southern Italy – had every reason to fear Roger’s growing power, not least in the maritime sphere. And the German rulers shared their concerns. By dint of the imperial title, they too had an interest in the south. Moves were soon put in motion to secure an alliance between the pope and the two emperors – a resurrection of Leo IX’s old alliance of the 1050s. This helped galvanise dissent within Apulia, where not all looked kindly on rule from Palermo.
As a result, Roger II spent much of the 1130s seeing off threats, both internal and external.8 Still, he never lost sight of other interests. As count, Roger had quarrelled with the Zirid rulers of Ifrīqya over trading rights, ordering an abortive strike against Mahdia in 1123. In 1135, internal struggles within the province provided him with the pretext for further intervention. The local Zirid amir, al-Ḥassan, had appealed for Roger’s help against the Hammadid ruler of neighbouring Bougie. Roger’s fleet was able to provide the requisite cover. But they also took the opportunity to seize the strategic island of Djerba, just off the North African coast. This was a statement of intent, and its significance was not lost on Roger’s northern counterparts. At an assembly held in Merseburg, Byzantine and Venetian emissaries reported – with a degree of exaggeration – that the Sicilian upstart had seized all Africa, the ‘third part of the world’.9
Pressing affairs on the Italian mainland initially prevented Roger from building on this base. But by the early 1140s, the time was ripe. Divisions continued to wrack Ifrīqya, pitting the traditional Zirid amirs against the Hammadid Berbers to their west. Droughts also made al-Ḥassan increasingly dependent on Sicilian grain supplies. Roger was able to leverage this to secure the amir’s submission in 1141/2. The following year, he initiated a series of naval attacks on the coastal towns of western and eastern Ifrīqya. These were framed as attempts to shore up al-Ḥassan’s rule. But soon enough, Roger pivoted from supporting to replacing his Zirid allies. In June 1143, Sicilian forces struck at Tripoli, which had sought to remove itself from Zirid rule. And while the attack failed, greater success was achieved at Djidjelli (Jījiil), in modern Algeria, which was sacked. The following year saw similar successes, with Barasht (Bresk) falling in the west and the island of Kerkenna doing likewise in the east.
When Tripoli was taken in 1146, the foundations were laid for a North African kingdom in Roger’s name. Hitherto, these expeditions had been little more than opportunistic acts of piracy. Now, Roger began to make the transition from intermittent raiding to permanent conquest.10 His cause was assisted by continuing famines, which weakened his opponents at crucial junctures. And as his father had done in Sicily, Roger was careful to conciliate the local Islamic population. While they were forced to pay an additional tax, Muslims would be free to continue worshipping as they had before. Such treatment took the sting out of Christian rule. And soon enough, Roger started to be approached by dissident elements in other regions of Zirid rule.
For the Zirids, the final blow came in 1148, when a large Sicilian fleet appeared outside Mahdia. Realising that resistance was futile, al-Ḥassan now fled inland. Afraid of the prospect of Christian rule, many of the inhabitants followed suit. But once they heard of Roger’s even-handed treatment of the local population, most returned. Momentum was now decisively with Roger, and his forces soon secured Sousse (Susa) and Sfax, strategic port cities to Mahdia’s south. Elsewhere, Tunis was reduced to tributary status and Gabès (Qābis) acknowledged Roger’s suzerainty. Impressed by this progress, Pope Eugenius III consecrated a new ‘archbishop of Africa’ (i.e. Ifrīqya). The domains of the radical Almohads – who in contrast with other Islamic powers, did not tolerate Christian minorities – had recently been expanding in north-western Africa. Roger’s realm now promised a bulwark against them.
The Sicilian Normans seemed unstoppable, and Ibn al-Athīr asserts that Roger would have conquered all Africa had he not been distracted by conflicts with Byzantium.11 This is an overstatement. At most, Roger might have added a few more coastal cities to his domains. Still, al-Athīr’s report gives a sense of the feeling within Zirid circles. Roger’s men had been making quick work of the region; and the Second Crusade, which saw Roger strike against Byzantium in 1147 and 1148, must have come as a welcome relief.
Roger’s own pride in these conquests is clear. He had coinage issued in North Africa and seems to have experimented with the title ‘king of Africa/Ifrīqya’, particularly in Arabic documents in his name.12 (In Roger’s trilingual realm, documents were issued in Latin, Greek and Arabic.) Yet as with Guiscard’s conquests in the Adriatic, Roger’s gains in Africa were dangerously exposed. To the immediate west lay the Almohads, who were unlikely to tolerate a Christian neighbour for long. To the east lay the domains of the Fatimids, the nominal overlords of the Zirids. The latter enjoyed good relations with Roger, and it was with their tacit approval that he’d established his North African enclave. As Roger began to find his feet, however, cracks started to appear in this relationship. The weak Zirids had been of little use to the Fatimids, but the rapidly expanding Normans now presented a rather different problem.
In spring 1153, a large Arab coalition backed by Roger was decisively defeated by the Almohads at Sétif in modern Algeria. In response, Roger sent a fleet under the command of Philip of Mahdia to take the port of Bône (modern Annāba), in the hope of creating a buffer zone. The longer-term prospects for the new African province were now starting to look grim. What sealed its fate was Roger’s death in early 1154. The ensuing years saw considerable instability, as Roger’s heir, William I, sought to establish himself. William’s priority was to secure his core Sicilian and Italian domains. Only once this had been achieved, could he afford to direct his attention to the situation in North Africa.
By then, it was too late. The mid- to later 1150s had seen a set of major revolts against Sicilian-Norman rule, as native rulers sought to re-establish their independence. Their cause was assisted by the continuing Almohad threat. In the end, it was this which proved decisive. In 1159, a large Almohad force moved against the coastal cities which were the linchpins of Norman authority in the region. Tunis soon surrendered; it was followed by Tripoli, Sfax and Gafsa. Greater resistance was offered at Mahdia, where the Normans had holed up within the walled city. Informed of the situation, William re-directed his fleet – which had just returned from raiding Almoravid-held Ibiza – to break the siege. And though these efforts failed, the city itself held out for another six months.13
Some at the Sicilian court felt that William hadn’t done enough, and there may be an element of truth to this.14 But with Almohad pressure mounting, it simply wasn’t worth the time, money and manpower to maintain this precarious toehold in North Africa. Roger’s conquest had always been an opportunistic affair, assisted by Zirid weakness and Fatimid acquiescence. There was little sense in maintaining it in the face of a sustained resistance.
In this respect, Roger’s conquest proved little different from the many regime changes North Africa had experienced since first falling to the Vandals in the early fifth century. The head of state may have been replaced, but the structures of rule – inherited ultimately from the Roman and Byzantine empires – remained intact. When the Almohads overwhelmed the region, they were able to rule it much as it had been for almost a millennium. How far Roger’s African domains were meaningfully Norman (or even Sicilian) may therefore be questioned. Yet from the start a key feature of Norman rule – be it in North Africa or northern France – was its adaptability. The Normans were a product of settlement and acculturation, processes they continued to undergo within the widely dispersed diaspora. On the death of Roger II, their domains in Italy were as much Sicilian, Calabrian and Apulian as they were Norman. Still, a sense of shared culture was maintained with the homeland. Roger is said to have preferred French and Anglo-Norman courtiers, on account of their shared heritage.15 And the names chosen by the Hautevilles for their sons are telling. Robert, Roger and William are not Italian designations. Even as they adopted local customs, the Normans didn’t forget their roots.
The eclipse of Norman rule in Africa was by no means inevitable. The odds may always have been against a sustained presence in the region. But it was only the confluence of three factors – Almohad pressure, Fatimid opposition, and internal problems in Sicily and Italy (occasioned by Roger’s death) – which proved decisive. Had Roger lived longer or the Fatimids proved more amenable, there’s no way of knowing how much longer the Norman kingdom might have survived.
Had it done so, Norman influence on the region would doubtless have been far greater. Compare the fate of Islamic Sicily. The Hauteville rulers’ recruiting of Muslim administrators here was pragmatic toleration rather than self-conscious multiculturalism. Christians were preferred for senior roles – and where possible, Latin Christians. The result was a slow but steady (and cumulatively significant) population shift away from Islam (and also, to a degree, from the Greek Orthodox rite). Lombard settlers from the mainland were encouraged, as was conversion. In the 1160s, toleration had already begun to give way to coercion, and these efforts were stepped up in the following century. As the Latin Christians established themselves more firmly, they begin to display the chauvinism so characteristic of the era. Pressure mounted on the remaining Muslim enclaves, and massacres and expulsions became common.16
In 1148, all this, potentially, lay in store for the inhabitants of Ifrīqya. Fortunately for them, Sicilian-Norman rule did indeed prove fleeting. And so it remains little more than an intriguing footnote in the long history of Norman conquest and colonisation. False starts are, however, every bit as interesting as successes. And Roger’s short-lived kingdom of Africa is a salutary reminder that the Norman influence on Europe, Africa and Asia could have been greater still.