The most important source for the conquest of Sicily is Geoffrey Malaterra’s De Rebus gestis Rogerii Calabriae et Siciliae comitis. He implied that the project for extending Norman authority to Sicily was conceived after the subjugation of Calabria was complete in 1060. However, from 1050 the Pope had appointed an archbishop of Sicily, and Robert Guiscard’s brother held the title of ‘future duke of Sicily’ after the synod of Melfi in August 1059. The island still had to be conquered and clearly, this had already been planned before Calabria was totally conquered. Malaterra chose to emphasise the importance of Count Roger’s role in the enterprise from the outset.
Sicily: the context
Sicily had been ruled by emirs of the Kalbite dynasty since 948 and the country had enjoyed a long period of domestic stability and prosperity. The emirs were, in theory, the representatives of the Fatamid Caliphs, the rulers of Egypt from 969 onwards but in practice they were an independent and hereditary ruling dynasty. During this period, Sicily was effectively brought under Islam. Christians were tolerated, as elsewhere in the Islamic world but they were an inferior class, legally disadvantaged and subject to special taxation. The Muslims were a superior social group and conversion to Islam meant that by 1060 two-thirds of the population were Muslim. The majority of Christians were concentrated in the north-east of the island in the mountainous Val Demone south of Messina.
In the early eleventh century, the political consensus of Islamic Sicily began to break down. The emir Ja’far was overthrown in 1019 by a rising caused by his attempt to alter the traditional system of taxation. At the same time, there was an influx of Berber immigrants from North Africa who brought with them the religious conflict between the Sunni and Shia branches of Islam. The relatively aggressive policies pursued by Ja’far’s successor Ahmed al Ak’hal resulted in an increase in raids on the mainland of Italy and on Byzantine Greece can be seen as an attempt to divert the attention of an increasingly divided society. Al Ak’hal’s assassination in 1037 resulted in a Byzantine expedition to Sicily in 1038 that aimed to exploit the internal divisions on the island but which was itself undermined by the Norman attack on Apulia in 1041. By the 1050s, political power on the island had fragmented into three contending principalities: the west was ruled by Abd-Allah ibn Manquet, the south and the centre by Ibn al-Hawwas and the east by Ibn Maklati based in Catania. Ibn-al-Hawwas was probably the most powerful of the emirs but none was in a position to dominate the whole island. The position was further complicated when Ibn al-Timnah took over Syracuse and then deposed and killed his neighbour in Catania.
The pretence for an invasion was supplied when the emir of Catania, Ibn al-Timnah came to Reggio to seek help from Roger de Hauteville in 1060-1 in order to get rid of his rivals. There are several reasons that could justify this conquest in the eyes of Robert: ensure the security of Calabria, end the Arab piracy and clear away the humiliation inflicted by George Maniakes on the Norman mercenaries, including the elder brothers of the Hauteville family, twenty years earlier, and more prosaically enjoy the spoils of war. Unlike other Norman conquests, the idea of conflict between Christianity and Islam added a religious dimension to the adventure, but the Normans proved afterwards that they never wanted the total eradication of Islam from the island.
The initial campaign of 1061
A further reconnaissance raid took place in March 1061. It was clear this was already being prepared when al-Timnah arrived. The force of 150 knights was still small but was larger than the force used the previous year. Amatus adds an interesting detail saying that Robert Guiscard made an experience soldier Geoffrey Ridel, a Norman whose family came from the Pays de Caux, joint commander. Whether this was to keep an eye on Roger or to act as a restraining influence is difficult to say but there is little doubt that this was a potentially hazardous mission. The Normans landed on the north coast and again defeated a force sent from Messina but then had some trouble withdrawing because bad weather prevented them getting to their ships. The two expeditions provided important intelligence on the Muslim defences and inflicted heavy casualties on their forces in the two battles near Messina.
The Muslim emirs were now aware of the threat from the Normans and took steps to prevent a further crossing by patrolling the seas with a significant naval squadron. In May 1061, Roger eluded the force and crossed with an advanced guard of 300 knights while Robert remained in Calabria with the substantially larger main force. The Muslims evacuated Messina and fled perhaps because losses in the earlier battles made an effective defence impossible. The capture of Messina provided a secure base for Norman naval operations and the short sea crossing from Calabria no longer posed a problem. In the summer of 1061, Robert Guiscard brought the main force over from Calabria mosing along the northern side of the Monti Peloritani and then southwards around the western side of Mount Etna into the land west of Catania. According to Amatus there were 1000 knights and 1000 infantry and it is not surprising that both Roger and Robert sought a battle with the Muslims. This occurred when the Norman encountered the army of al-Hawwas on the banks of the River Dittaino to the east of Castrogiovanni which it completely defeated despite being heavily outnumbered.
The aftermath of this victory appeared disappointing. Roger and a force of 300 young knights raided south-west towards Agrigento but Castrogiovanni proved impregnable and barred among permanent advance to the south of the island. In late 1061, Guiscard returned to Apulia essential to reassert his authority there and a recurrent problem for his operations in Sicily. Roger remained in Calabria. Despite this, the campaign in 1061 had proved extremely successful. Messina had been secured and the whole of the area to the north and east of Mount Etna was in Norman hands. In addition, they now enjoyed two important advantages. First, the Christian enclave in the Val Demone welcomed the Normans as liberators and provided them with much-needed supplies. Secondly, they had a valuable local ally in Ibn Timnah who had been restored to his lordship at Catania by the time Roger returned to the mainland. Early in 1062, he helped Roger to besiege and capture Petralia, twenty miles north-west of Castrogiovanni, which became an important advanced base for the Normans
From Messina to Palermo: Norman operations 1062-1072
Even though Robert had been involved in the first attempts to conquer the island (until 1064 and in 1072), the principal protagonist was Roger, his younger brother. He took advantage of the division of Sicily into three or four emirates, that were practically independent from each other, and the antagonism between the Berbers and Arabs, by leading a guerrilla war with from fifty to a few hundred men. This explains the length of time needed to conquer the island: thirty years. After the first invasion, the taking of Messina and several mountainous areas of the Val Demone in 1061, and the sustained operations from 1061-1064, the conquest continued, albeit laboriously. There were important victories at Cerami in 1063 and Misilmeri in 1068. Palermo was besieged and taken on 10th January 1072, thanks to the help of the fleet captured in Bari in 1071. This established the Normans as a maritime force.
Completing the conquest 1072-1091
The capture of Palermo marked a very clear stage in the conquest of Sicily. It represented the last direct involvement of Robert Guiscard in the conquest of the island and he never returned to the island after he left in early 1072. Guiscard granted Count Roger the fiefdom of Sicily, except Palermo, Messina and the Val Demone, which were possessed jointly. Although Roger was nominally subject to his elder brother, in reality he was left to govern the island and to continue the conquest as and when he could. By 1072, the Normans had conquered about half of the island. They had displayed considerable flexibility in exploiting Muslim divisions and in granting lenient surrender terms and toleration for Muslim worship.
This pragmatic policy towards non-Christians was marked although the contemporary chroniclers were at pains to stress the religious nature of the struggle and the overtly Christian mission of the Normans. William of Apulia wrote that Roger “fought continually against the Sicilians, enemies of the Dibvine Name’ and later, about the siege of Palermo that “This city is an enemy to God and knowing nothing of the Divine worship is ruled by demons…Christ makes difficult work easy.” Malaterra decribed how Roger encouraged his army at Cerami, “Our God, the God of Gods is all-powerful…with God going before us we shall be irresistible.” The chroniclers’ words reflect how they viewed the struggle and both William and Malaterra were writing in the 1090s around the time of the First Crusade. It is easy to dismiss the Holy War component of the Sicilian conquest as a later construct. However, Amatus of Montecassino, writing at the latest before 1078 also viewed events in an expressly religious context. He too saw Robert and Roger as leadingly an expressly religious fight. While the Normans were conducting what was perceived, at least by their own historians, as a holy war recovering land that was rightfully Christian, in practice the means for achieving this was pragmatic. The Normans were numerically inferior. They had considerable difficulty in capturing well-fortified towns and hilltop strongpoints and there were divisions among the Muslims that could be exploited. Flexibilty and toleration coexisted with the ideology of the holy war.
The conquest of the remainder of the island was to last a further twenty years. The main reason for this was the shortage of men at Roger’s disposal. When Guiscard left, the majority of his army went with him. Furthermore, when Malaterra wrote of the two brothers helping each other, this generally meant that, after 1072, it was Roger who was summoned to the mainland to help his brother and these absences not only prevented further conquest but put already conquered areas under threat. Neither could Roger afford to neglect Calabria as there was always the danger of external intervention. In June 1074, a Zirid fleet from North Africa sacked Nicotera in southern Calabria. Raids like this were of considerable nuisance value but early the following year a more substantial Zirid force landed at Mazara, an area taken in 1072 but on the extremes of the Norman conquests. Though the Zirids broke into the town, they did not take the citadel and were routed by a relief force. This episode demonstrates how fragile Norman control was over Sicily especially in their more advanced bases.
Roger was then summoned to the mainland to Guiscard to help him in his campaign in Calabria against his rebellious nephew Abelard, son of his elder brother Humphrey. For much of 1075 and probably into the early part of 1076, Roger besieged Abelard’s base of San Severina in northern Calabria. In his absence, Roger left Sicily under the command of his son-in-law Hugh de Gercé, a noble from Maine who held the lordship of Catania with strict instructions to remain on the defensive. However, Hugh accompanied by the count’s illegitimate son Jordan went on the offensive, were ambushed and Hugh and many of his men killed. The leader of the Muslim army that defeated Hugh was Ibn-el-Werd, emir of Syracuse who was to be a major stumbling block to continued Norman expansion for the next ten years. During this time, he was the most effective Muslim leader on the island, sometimes taking the war to the enemy and possessing in Syracuse one of the best harbours on Sicily.
Once Roger return to Sicily in 1076, the offensive was resumed and there were significant Christian gains in the next five years. A punitive expedition against Ibn-el-Werd’s land in the south-east resulted in widespread destruction and famine in the Val di Noto. In 1077, Roger captured Trapani bringing the west of the island firmly under his control. The following year, Castronuovo, one of the strongest fortresses in central Sicily was betrayed to him. In early 1079, the siege of Taormina, the one remaining Muslim fortress north of Mount Etna began and lasted six months. Roger’s control of the sea was important and both Trapani and Taormina were blockaded from the sea and his fleet was large enough to compel a Zirid squadron of fourteen ships to abandon its attempt to relieve Taormina.
However, fresh problems emerged in 1079 with a revolt of Muslims in western Sicily at Jato and the surrounding districts, not much more than twenty kilometres from Palermo. Malaterra stated that there were 13,000 Muslim families in this area who was previously acknowledged Norman overlordship and paid taxes to the count. The rebellion may have been caused by setting tax levels too high because of spending on mercenaries to supplement Roger’s on occasions, rather meagre force. Roger sought to persuade the Muslims to submit. When this failed, he established new fortresses at Partinico (to the north) and Corleone (to the south) but it was only after a long campaign and the destruction of the Muslim harvest that the area became quiet. Revolts remained a problem. Despite this, the Norman had made considerable progress by 1080 when again events on the mainland intervened. Roger’s presence was necessary to shore up support for his nephew Roger Borsa while Guiscard was fighting the Byzantine Empire. In his absence, Catania was betrayed to Ibn-el-Werd but was quickly recovered by the count’s son Jordan when Ibn-el-Werd made the mistake of fighting a pitched battle against the Normans.
When Roger returned to Sicily, he immediately strengthened the defences of Messina. Malaterra was right to regard the city as the key to Sicily but it had been under Norman control for twenty years and by the early 1080s was well behind the front line. Roger appears to have been very uneasy about the situation on the island perhaps because he recognised that he would almost certainly be needed on the mainland again. In addition, Roger faced attempts by his subordinates to entrench their own authority in particular areas. The first challenge came from a knight called Ingelmarius whom Roger had favoured. He fell out with the count in the early 1080s when he built a castle at Gerace without comital approval. More serious was the rebellion in north-east Sicily led by Jordan while his father was on the mainland in 1083-4. It is difficult to explain why Jordan rebelled but as the count’s only healthy adult son, albeit an illegitimate one he saw himself, though he had not yet been designated as Roger’s successor. It is also important to recognise that this rebellion can be sen as an attempt to consolidate lordships by people who had previously been loyal but who had not gain immensely from the conquests in Sicily. Roger dealt with the matter quickly and firmly. Although Jordan was soon restored to favour, several of his associates were blinded.
While the Muslim revival under Ibn-el-Werd and rebellions may have hampered the completion of the conquest, there seems little doubt that it was Roger’s absences on the mainland that were the major difficult. After Guiscard’s death in 1085 and the succession of Roger Borsa as duke, count Roger was able to turn his complete attention to the island and the conquest was then completed quite rapidly. He was aided in this by a peace treaty concluded before 1087 with Temin, the North Africa ruler that meant the Sicilian Muslims would not longer get any external help. In 1086, Roger besieged and took Syracuse and Ibn-el-Werd was killed. The following year, he besieged and took Agrigento and the impregnable Castrogiovanni (now Enna) was taken, in the centre of the island.
By now only the extreme south-east of the island was in Muslim hands. Its conquest was delayed when, in the summer of 1088 Roger againt returned to Calabria to help Roger Borsa against his half-brother Bohemond. In April 1089, his army and siege engine appeared round Butera and it surrendered after a short siege. Finally, the last Saracen stronghold at Noto surrendered voluntarily in February 1091. The military phase of the conquest had been achieved though it was far from the end of the process of consolidation.
Consolidating Norman rule
Once the military phase of the conquest was completed, Roger was faced with several problems that needed solution. The Muslims remained the largest ethnic group on two-thirds of the island. Although, in 1064 Robert Guiscard had transferred the entire population of a captured community from Sicily to Calabria and Roger imitated this precendent when he sent to leading citizens of Butera to Calabria in 1089, ‘ethnic cleansing’ was never really an option. More important in terms of keeping the Muslim population quiet was the emigration of many Muslim aristocrats and intellectuals to live in North Africa or Spain as it removed potential leaders of rebellion. The long-term solution was either to encourage widespread conversion among the Muslim population or significant Latin Christian immigration from the mainland. This meant creating a Christian Church structure on the island almost from scratch.
The creation of a new Christian Church had already been begun before the conquest was completed. When Palermo was captured, the principal mosque was converted into the city’s cathedral and the existing Greek archbishop Nocodemus was installed there though his successor was a Latin. It remained the only see on the island until 1079 when Roger founded a bishopric at Troina in the Val Demone though it was not operational until two years later. This was approved by Pope Gregory VII who considered the whole process irregular and warned Roger not to see this case as a precedent. The creation of further sees had to wait until the rest of the island was conquered but a number of monastic foundations were established in the interim. Once the conquest was completed, Roger acted quickly establishing four further bishoprics at Catania, Syracuse, Agrigento and Mazara between 1091 and 1093 with the full approval of Pope Urban II.
Roger needed to administer the areas he had won and reward his supporters. The rebellions of the 1080s showed what could occur when his principal followers felt that they had not received their just rewards. The northern part of the island was already being enfeoffed during the 1080s and the distribution of land was greatly extended after the fall of Noto in 1091. However, Roger retained much of the island in his own hands, especially the centre and the west. Few substantial lordships were established and these were either for close relatives of the count or for favoured churchmen. These were largely in the east and along the north coast of the island. His illegitimate son Jordan was given Syracuse and extensive lands in the south-east of the island including Noto and Lentini though he only enjoyed this for a short time as he died in the autumn of 1091. He then granted his barony to his nephew Tancred, a younger son of his elder brother count William of the Principate. Another son, Godffrey, also illegitimate was given the seignory round Ragusa. In the early twelfth century, there was a third important lordship in this south-eastern region around Butera and Paterno held by Roger’s brother-in-law Henry. With the exception of a handful of non-Norman Frenchmen who were given land, count Roger used his numerous offspring as a means of controlling the new Sicilian establishment. Most of the principal towns and fortresses (apart from Syracuse, Catania and the halves of Palermo and Messina still subject to the duke of Apulia) remained under his direct control. The same applied to a vast area of land in the centre and west of the island, almost all of which remained under the direct control of the rulers of Sicily until late in the twelfth century.
Count Roger stressed his dominance over his followers by personal example and by title. He was incensed when his son Jordan suggested that he, rather than his father, should command the expeditionary force that conquered Malta in 1091. Roger wished to be seen as a conqueror, the man who reclaimed places for Christianity and resented any suggestion that he was too old. His role, as the restorer of Christianity to the land polluted by the infidel was stressed both in his own documents and in papal bulls sent to Sicilians. From 1092, he sometimes styled himself as the ‘Great Count’ of Sicily and Calabria to differentiate himself from other counts on the mainland rather than to stress his superiority over his own vassals. One of the marked features of Sicily is that there were no other counts in Sicily until shortly before Roger II accepted the royal crown in 1130. Roger refused to allow any of his subordinates in Sicily, even his sons to call themselves count. The reason was very simple: Roger did not want the new Sicilian barons, however closely they were related to him to have any claims to independent authority or to a rank that challenged that of the island’s ruler.
The most significant problem facing Roger after 1091 was demographic not vassalic. The Normans and the French made up a small and potentially vulnerable ruling class that dominated an entirely alien subject population largely by coercion or its threat. Culturally, Sicily was a problem. Its Christian population was almost entirely Greek and only in the north-east of the island were they in the majority. Apart from in Messina and Troia, Muslims formed the overwhelming majority of the urban population. Christian conversion did take place but it was overwhelmingly to the Greek rite rather than the Latin probably reflecting the imperfect conversion of the indigenous population to Islam in previous centuries. Efforts were made to attract Christian settlers to Sicily in an attempt to remedy this situation. Roger offered the prisoners he found when he conquered Malta the opportunity to settle in Sicily as free tenants. In 1095, Abbot Ambrose of Lipari also offered land on generous terms specifically to ‘men of the Latin language’. Efforts to encourage immigration continued throughout the twelfth century in a slow though eventually successful attempt to change the population structure of the island.
The demographic imbalance on the island was also reflected in how Sicily was governed. The early governors were Norman and Normans or Frenchment made up most of the castellans but most of the other officials were Greeks and Arabs who were often Christian converts. Existing Arabic land divisions continued to be used. Bilingual documents, including Arabic rights continued to be used throughout the twelfth century and most of Roger II’s charter were written in Greek. Muslim troops were used as early as 1076 in the siege of Salerno, in the march on Rome in 1084 and in most of Roger’s campaigns in the 1090s. Greek monks remained in the majority: between 1084 and 1101 Roger endowed three Latin monasteries but in the same period he endowed fourteen Greek ones. Roger’s religious policy largely followed the demographic structure of his domains.
After 1091, a Norman count ruled the whole of Sicily but changes on the island were limited. The count relied heavily on native officials and practices in his government of the island. Alterations to the overall balance of the population were slow and only began to have an effect on its social structure in the mid-twelfth century. The Normans were a small ruling minority on Sicily and even more so than on the mainland.
 Geoffrey Malaterra, monk of the monastery of Saint-Evroult in Normandy seems to have moved to southern Italy in the period when Robert de Grandmesnil was abbot of the monastery. He was instructed by Roger I count of Sicily (1062-1101), to write his biography. The four books of De rebus gestis Rogerii Calabriae et Siciliae comitis are written in prose with the exception of a few paragraphs in Books III et IV that are in verse. In the Book I, Malaterra tells of the exploits of the Hauteville family and especially those of Robert Guiscard and then those of Roger. Book II presents the early phase of the conquest of Sicily from the beginning of the enterprise in 1061 until the capture of Palermo with the help of Robert Guiscard in 1072. In Books III and IV, the conquest of Sicily is continued dealing with the war against the emir of Syracuse to the submission of Val di Noto in 1091. Certainly Geoffrey Malaterra does not neglect the help Roger gave to Robert Guiscard, each time that the duke needed it to maintain his domination on the Italian mainland. He describes the campaign of Robert Guiscard in Greece and his quick return to help Pope Gregory VII besieged in Rome by Emperor Henry IV. The work ends with the reorganisation of the Sicilian church and the concession made in 1098 by Pope Urban II to Roger. Malaterra provides a valuable complement to the poem by William of Apulia Gesta Roberti Wiscardi, dedicated to Robert Guiscard, brother of Roger. Even if Malaterra is a little too ready to celebrate the courage and the piety of the Norman princes, the De rebus gestis Rogerii is a unique source without which we would know little of the Norman conquest of Sicily in the second half of the eleventh century. The most accessible editions of his work are: L.A. Muratori (ed.) Gaufredi Malaterrae monachi Benedictini Historia Sicula, in Rerum Italicarum Scriptores, volume v, pages 537-605, Milan, 1724, reprinted by J.P. Migne (ed.) in Patrologie latine, volumn 149, columns 1093-1210 and E.G. Pontieri (ed.) Gaufredo Malaterra, De rebus gestis Rogerii Calabriae et Siciliae comitis et Roberti Guiscardi ducis fratris eius, in Rerum Italicarum Scriptores, volume v, 1 (fasc. 211 et 218-219), 2nd ed., Bologna, 1927-1928, pages 1-108.
 Roger was born in 1031, Normandy and died on June 22nd 1101 at Mileto in Calabria. He was the last son of the second marriage of Tancred of Hauteville. Roger went to Italy in 1057 and aided his brother Robert Guiscard in his conquest of Calabria from the Byzantines, something achieved by 1060. They began the conquest of Sicily from various Muslim rulers in 1061 with the capture of Messina but it was only completed in 1091. The turning point of the struggle was the capture of Palermo in 1072, when Robert invested Roger as his vassal with the county of Sicily and Calabria with a limited right to govern and to tax. After Robert’s death Roger acquired full right to govern from Robert’s son and in 1098 received the title of apostolic legate from Pope Urban II, which gave him control of the church in Sicily. At his death, Roger had created a centralised, efficient government, where the authority of the count was unchallenged.
 The accounts of both Malaterra and Amatus of Montecassino mention al-Timnah’s appeal to Roger though Malatera suggests this occurred in late February 1061. By then, the Normans had already decided to launch an invasion of Sicily and in late 1060 Roger had already made a preliminary reconnaissance in force, landing at Messina with sixty knights and inflicting a sharp defeat on a force sent from the town to attack him.
 Malaterra alleged that 700 Normans defeated 15,000 Muslims killing 10,000. The number of Normans may be credible and must have been smaller than the original force that crossed (the need to garrison captured areas) but the Muslim figures are almost certainly an exaggeration.
 Jean Deuve Les Opération Navales Normandes au Moyen Âge (900-1200), Editions Charles Corlet, 2000, pages 11-20 and 59-78 covers Norman naval operations in the Mediterranean in the eleventh century. Les héritage maritime des Vikings en Europe de l’Ouest, University of Caen, 2002 provides valuable context especially pages 101-118.
 In 1083-4, after his return from the campaign in the Balkans, Robert needed Roger’s aid to restore his authority on the mainland and to aid the papacy against the Emperor Henry IV.
 This repaid the debt Roger owed to his long-dead brother who had supported him in his disputes with Guiscard in the late 1050s and early 1060s.
 Henry was in Sicily from c.1094 and may have been married to one of Roger’s daughters by a previous marriage but there is no direct evidence for his lands until 1113. It is possible that he was granted the lordship not by Roger I but by Adelaide, the latter’s widow (and Henry’s sister) during her regency following the count’s death in 1101.