Part II
6
The administration of the Norman Kingdom of Sicily had been regarded to be highly specialized and bureaucratized in comparison with other monarchies in medieval Europe, and was treated as one of the most remarkable institutional achievements of that era. Heinrich Mitteis thought that the administrative institution of medieval Sicily had affected those of England, France and Germany,1 while Albert Brackmann,2 David C. Douglas3 and Antonio Marongiu4 argued that the efficient and rigorous administration of Norman Sicily had been a forerunner of the secular modern administration. When and how this highly bureaucratized administration of Sicily was formed has been a controversial topic among scholars for some time. Some scholars insisted on the influence of the Islamic world5 or the Byzantine Empire,6 while others emphasized the English influence.7 However, I have proposed a new understanding, suggesting that the administration of the Norman Kingdom of Sicily had been far simpler than previously assumed, just a patchwork of the preexisting systems and organizations.8 The purpose of this article is to elucidate the characteristics and changes of the administration of Count Roger I of Sicily (1072–†1101), which has been regarded as the foundation of the administrative system of the Norman Kingdom of Sicily.
When we examine the administration of Roger I, we have to bear in mind the fact that it was able to change greatly over the course of time. It cannot be held that he kept the same policy and the same administrative system for more than forty years, from 1058 when he obtained Mileto in Calabria until 1101 when he died. In fact, a very significant change occurred around 1086, when the powerful Muslim leader Ibn al-Ward died and Agrigento and Castrogiovanni fell. It is true that Muslims resisted in Noto until 1091, but the completion of the conquest of Sicily was thought to be sure a few years before.9
Therefore, in this article I would like to show how Roger I’s administration changed around 1086 and what characteristics the newly formed administration had. The following argument consists of three parts. First, I will show the nature of the administration of Roger I before 1086. Second, I will examine the change caused after 1086. Third, I will show the characteristics of the newly formed administration.
I
Roger I’s administration before about 1086 was basically a wartime one. There is very little information regarding the administration of this time. We have few comital documents, and thus we have to depend on contemporary narrative histories such as those of Malaterra, Amatus, Romuald and William.10 These narrative sources are useful to know the process of the conquest of Roger I, but they offer little information about his administration.
One thing we can learn from these sources is about the generals who commanded the troops. One of them was Roger I’s illegitimate son, Jordan,11 who played a very significant military role under his father. He directed troops in the war and was entrusted with Sicily when his father was on the mainland in 1083.12 Besides Jordan, we know of several magnates of Roger I. In the expedition against Taormina in 1079, the four troops of his army were under the directions of Jordan, Otto, Arisgot of Pozzuoli and Elias of Cartomi (†1081).13 Arisgot worked for Roger I for a long time and was given a large fief after the fall of Palermo.14 Elias was a Christian converted from Islam and was killed in the war against Catania in 1081.15 In this war, Robert of Surdavalle directed a troop together with Jordan and Elias.16 Furthermore, we know from narrative sources that Roger I’s son-inlaw Hugh,17 his nephew Serlo,18 Geoffrey Ridel19 and Ursell (Roussel, Orsell) of Bailleul20 played active roles as commanders of troops.
There is no doubt that they were magnates of Roger I and principal members of his entourage at that time. From narrative sources, we know only those great feudal vassals of Roger I. The comital document of 1085 has a list of witnesses, which also shows only feudal vassals.21
Contemporary narrative histories show that the conquest of Sicily is filled with sieges of cities. In fact, Muslims’ bases of resistance were mainly those cities fortified with ramparts. To capture cities also meant gaining command of their far more extensive neighboring areas. In this time of war, Roger I’s main concern was without a doubt to subject as many cities as possible to his own authority and hold them under his secure command.
When his brother, Robert Guiscard, entrusted Roger I with the conquest of Sicily he left only a small number of knights. Malaterra suggests that Roger I had only some hundred knights under his command.22 And there was always the possibility of rebellions against his and his brother’s authority, especially on the peninsula.23 Therefore, it is quite understandable that he tried to avoid battles if possible and urged Muslims to surrender by negotiation. The case of Palermo illustrates this well.
When the Muslims of Palermo surrendered to Roger I and Guiscard in 1072, their representatives – two qā’ids – together with other magnates negotiated with Roger I.24 Although no sources provide the details of the negotiation at this time, Roger I assured the safety of Muslim residents and allowed them to keep their own faith on the condition that they should pay annual tributes and give service to their new lord.25 According to Malaterra, these representatives negotiated so that “they should not be oppressed by unfair new laws.”26 At this time, as many scholars think, Muslims seemed to be allowed to keep some sort of autonomy, in particular having their own laws, judges and judicial system as they did in a later period.27
A large number of cities probably concluded similar treaties with Roger I when they submitted. Such were the cases of Catania, Mazara, Trapani, Taormina, Syracuse, Castrogiovanni, Butera, Noto and Malta.28 In most of these cities, Roger I possibly kept the old administrative system as it was and did not do more than replace only top administrators with his own.
Therefore, a city under Roger I’s authority could easily rebel against him. Such is the case of Catania. This city was subjected to the rule of Roger I in 1071 (1072), and was given to his son-in-law, Hugh of Gercé.29 After his death, its administration was entrusted to Ibn al-Thumna, a leader of the Muslims of the city. When Roger I was away on the peninsula, the city revolted and took the side of Ibn al-Ward in 1081.30 Roger I managed to regain the control of the city in the following year, but he was forced to spend much of his time and energy on the matter.31
It must be emphasized that in this period of war, Roger I’s top priority was to subject as many cities as possible to his own authority and hold them under his secure command. However, since he could not afford to station many trusted able men in each city, he was obliged to keep an existing administrative system as it was. As the example of Catania shows, he probably appointed leaders of Muslim communities as heads in some cities. In this period it was impossible to build up a centralized governmental system.
II
The situation changed when Ibn al-Ward, the most powerful Muslim leader, died in 1086 and Agrigento and Castrogiovanni fell into Roger I’s hands in the following year.32 Roger I’s brother and feudal lord, Robert Guiscard, had died a little earlier, in 1085. The death of his brother had already changed the situation of Roger I from a vassal of the duke of Apulia to the de facto independent ruler of Calabria and Sicily.33
Around the year 1086, Roger I’s main concern shifted from the war of conquest to the secure governance and efficient administration of his dominion. This change may be reflected in an increase of issuance of his documents in the 1090s. Most of his documents were issued after 1090; very few documents were issued before this time. Furthermore, most of the documents were written in Greek.34 The composition of the entourage of Roger I greatly changed and new governmental officials were appointed after this time.
A Greek document issued in 1117 by Roger II, son of Roger I, includes a Greek and Latin document of Roger I issued in 1090.35 The Latin part and the witness list were written in the same handwriting, which suggests that a scribe copied a document from 1090 by hand in 1117.36 There is also a listing of witnesses in the Greek text,37 which includes clerics and officials as well as feudal vassals.
The people shown as witnesses in the documents were by no means the whole entourage of Roger I, or even all of its principal members, but just a part of this important group at a certain time. If there were a large enough number of lists available, it would be possible to reconstruct a whole image of the magnates surrounding the count. Unfortunately, however, we have very few lists, and it would be misleading to try to reconstruct the entire picture of his entourage based on the limited number of lists, which chronologically span some decades and do not reflect changes in members reliably enough. But the evidence found in these lists is still important even though it is inevitably partial.
The witness lists of the documents issued in 1097 and 1101 include clerics and officials in addition to Norman barons. In the document of February 1097 we see as witnesses the count’s wife Adelasia, his son Malger, William of Hauteville, Jos-bert (Gosbert) of Lucy, Robert Borrell, Paganus de Gorgusio (Gorgiis), Roger of Stilo, John de Traginiis the protonotarius, Nicholas de Mesa, the three chaplains (Girald, his brother Falco and Jeremia de Sancto Egidio) and Hugh of Melfi.38 The document of 16 June 1101 includes as witnesses Countess Adelasia, Robert Borrell, Josbert of Lucy, Robert (bishop-elect of Mileto), Roger (bishop of Syracuse) and Nicholas the chamberlain.39
This implies a change in the composition of the entourage. As mentioned above, Roger I had almost completed the conquest of Sicily in about 1086 and become the sole master of Calabria and Sicily. His priority was to consolidate his position in his territory and organize his administration more effectively. It is probable that he needed administrative officials and clerics more than ever. There was another reason for the increasing number of high ecclesiastics in the witness lists: territorial expansion made it possible for Roger I to found new Latin bishoprics.40 Roger I first created a new bishopric at Mileto with a confirmation by Gregory VII (1073– †1085), probably between 1075 and 1081.41 He also founded several new Latin bishoprics in Sicily. He selected Robert to be the first bishop of Troina in December of 108042 and, together with Robert Guiscard, replaced a Greek archbishop of Palermo with Alcherius, a Latin cleric, before 1083.43 Thereafter he selected Latin clerics to the newly founded or formerly Greek bishoprics: Gerland natione Allobrogum to Agrigento, Stephen of Rouen (Rothomagensem) to Mazara, Roger of Provence (in Provincia ortum) to Syracuse and Anger of Brittany (natione Britonem) to Catania.44 Thus there were drastic changes in the church organization under Roger I, and the witness lists partially reflect these changes.45
The last group of the entourage, who also appeared from 1090 on, was composed of governmental officials. This group included comital household officials like Geoffrey the seneschal and William the seneschal in 1090, both of whom seem to have been Norman vassals. However, the most important officials, such as prōtonotarios (πρωτονοτάριος), prōtokritēs (πρωτοκριτής), and logothetēs (λογοθέτης), were of Greek origin. These titles had been used for the imperial officials in South Italy under the Byzantine rule.
It is certain that Roger I preserved most of the Byzantine administrative offices and officials in Calabria. Under Byzantine rule, the thēma of Calabria had been governed by the stratēgos (στρατηγός) of Calabria. By the tenth century, however, other officials like prōtonotarios and kritēs (κριτής) began to work together with the stratēgos.46
These offices survived the Norman conquest of Calabria, although the title of kritēs had begun to be held by local judges of towns in the course of the tenth and eleventh centuries.47 Under Roger I, Leo the logothetēs held the office of megas kritēs of Calabria (μέγας κριτὴς πάσης καλαβρίας),48 and Nicholas the prōtonotarios, who bore the title of prōtokritēs of the whole of Calabria (πρωτοκριτὴς ἁπάσης Καλαβρίτιδος χώρας) in 1098, seems to have taken over for him.49 These two offices of Byzantine tradition, prōtonotarios and prōtokritēs (or megas kritēs) of Calabria, were significant officials at the comital court and constituted core members of Roger I’s entourage. The Byzantine stratēgos of Calabria disappeared after the Norman conquest, but the title of stratēgos was given to magistrates of towns under Roger I.50
Thus the Greek officials with Byzantine titles were predominant at the court of Roger I after 1086, and Byzantine officials continued to function as local ones in Calabria.51 They were overwhelmingly influential at the time when Roger I’s administrative system was being structured. A Greek document of 1105 mentioned three people as the magnates of the comital court just after the completion of the conquest of Sicily in 1091,52 and all three were Greek. They were kaprilingas (καπριλίγγας) Nicholas, logothetēs Leōn, and Eugenios prosēnestatos tēs eugeneias (προσηνέστατος τῆς εὐγενείας).
III
Thus, after around 1086 the entourage of Roger I came to consist of feudal vassals, clerics and governmental officials, and a new administrative organization began to take shape. Two important features should be pointed out concerning Roger I’s administration in this period. The first one is good use of the former ruler’s governmental units and tools. The second is the predominance of Greeks and absence of Muslims in the central government.
In fact, Roger I seems to have effectively used land registers (daftar in Arabic) and lists of inhabitants (jarīda in Arabic, plateia in Greek and platea in Latin) made by the Muslim rulers in distributing fiefs, and preserved their territorial divisions (iqlīm in Arabic).53 This is suggested by two kinds of sources. First, the following document (a writ of transfer from Roger I to the bishop of Messina), dated as 1094 and transcribed by Pirro, states that a grant of land was made according to the old Muslim divisions:
[Latin] Having heard their petition for the welfare of my soul and of the soul of my brother, the most noble Duke Robert Guiscard,… I gave and in perpetuity granted the village of Saracens called Butahi together with its belongings to the church of St. Nicholas in the bishopric of Messina according to the old divisions of Saracens.54
This document is claimed to be a forgery of the early to mid-twelfth century by Jeremy Johns, but the phrase “secundum antiquas divisiones Saracenorum” was probably taken from older documents.55
Second, the two writs of transfer of 12 and 20 February 1095 consist of a fore-word in Greek, a list of inhabitants in Arabic and an afterword in Greek.56 This suggests that these writs were based on the list of inhabitants of the former Muslim rulers and that Roger I made use of them in granting fiefs.57
Thus the land registers and lists of inhabitants became the most significant land administration tools for Roger I, and keeping these documents and revising them became an important duty of the comital government. Roger I’s policy about land and inhabitants was carried out on the basis of these registers. The afterword of the writ of transfer of 1095 shows this.
[Greek] This plateia (πλατεῖα) was written by the order of me, Count Roger, in Indiction III and AM 6603 (= AD 1094/5) in Messina. However, the other plateiai (πλατείαι, pl. of πλατεῖα) of my land and my feudatories (τερρερίοι) had been written in Indiction I and AM 6601 (= AD 1092/3) in Mazara. Therefore, we order that if anyone of those Hagarites (ἀγαρινοί = Saracens) listed in this plateia should be found in the plateiai of my feudatories the bishop must turn them back without exception.58
While issuing these documents, Roger I’s government had to revise and maintain land registers and lists of inhabitants. They needed skilled officials to do so. In a later period, an organization specialized in such work was created. This was called dīwān al-taḥqīq al-ma‘mūr in Arabic or duana de secretis in Latin.59 However, we do not know whether such an organization existed in this period or not.
The second feature of Roger I’s administration in this period is the predominance of Greeks and absence of Muslims in the central government. Concerning the demographic condition after the conquest, most of the population in Calabria and the eastern region of Sicily was Greek, and in the rest of Sicily Muslim. In order to govern these people Roger I needed experts on their customs and legal systems. It would be quite understandable if he had made use of Greek and Muslim officials of the former rulers. But he did not use Muslim officials in the central government, although they were the majority of the population of Sicily. Most officials were Greeks. Muslims or titles of Arabic origin seldom appear in Roger I’s documents. The only exception is the title of amiratus.60 William of Apulia described that when Robert Guiscard occupied Palermo in January 1072, he appointed his knight to be amiratus to govern this city.61 This word amiratus is a transliteration of Arabic amīr.
Guiscard used this Arabic title instead of a more familiar Latin or Greek title for this new governor, probably with the intention of showing Muslims in Palermo that this knight was their head and representative. The creation of this office symbolically shows the fact that Guiscard simply put one of his vassals at the top of the Muslim population without destroying the autonomy of the existing Muslim administrative units.
When the first amiratus was appointed, Sicily was in the middle of the conquest. Guiscard retained half of Palermo and its administration in his own hands while giving half of it to his brother Roger I. But this city was an exclave for him as the preponderant part of his territory was located on the Italian peninsula. Thus, the amiratus was just a local official of the duke of Apulia who was entrusted with the administration of Palermo and its environs. In its initial stages, the office of amiratus was held by Norman vassals.62
However, when Roger I obtained the other half of Palermo and its administration in the same year of the completion of the conquest of Sicily, he gave this office to a Greek official Eugenios. As shown above, he was one of the three magnates at Roger I’s court.
Thus the office of amiratus, although of Arabic origin, was exclusively given to either Norman vassals or Greeks. It was not given to Arabs at all. Arabic people did not appear in the witness lists of comital documents either. These facts seem to suggest that Muslims were excluded from the central government.
* * *
Although the conquest of Sicily was completed in 1091, Roger I’s main concern seemed to have already shifted from the war of conquest to the secure governance and efficient administration of his dominion some years before. This is a change from wartime governance to peacetime governance. After around 1086 a new administrative organization began to take shape. This change seems to be reflected in the increased number of comital documents issued and in the change of members of witness lists.
In this formation period, Roger I made good use of the former ruler’s governmental units and tools, especially lists of inhabitants and land registers. He kept Muslims’ administrative autonomy in many cities and utilized Byzantine officials together with their local administrative units. But he did not employ Arabic people as high officials in the central government.
This does not mean that there were no Arabs around Roger I. When he began the conquest, his ally was Ibn al-Thumna (Betumen).63 Elias of Cartomi, who directed one of Roger I’s troops, was formerly a Muslim.64 According to Eadmer of Canterbury, a great number of Muslim soldiers had been working for Roger I.65 Ḥamūd (Chamut), once a ruler of Castrogiovanni, converted to Christianity and spent the rest of his life in the region of Mileto.66
Roger I seems to have excluded Arab officials from the central government intentionally. The memory of the long war against Muslims, or the betrayal of Ibn al-Thumna of Catania, might have made Roger I take this policy. There might be other reasons. No matter the reason, however, it was after his death that many Arabs began to play important roles in the central government, and we do not know precisely when this occurred. The transfer of the capital to Palermo would probably have contributed to this change. George, that great minister of Roger II who knew Arabic language and culture well, might have started to employ Arabic people.67 However, the answer to this important question requires further investigation.
Notes
* This is a revised version of my paper, The Administration of Roger I: The Foundation of the Norman Administrative System, which was read at the Congresso internazionale di studi per il IX Centenario in Troina in 2001, and was published in its proceedings: Ruggero I Gran Conte di Sicilia 1101–2001, ed. Guglielmo De’ Giovanni-Centelles (Rome, 2007). Unfortunately there were too many errors in the text in the proceedings, most of which had been caused by the conversion from a Word file to a different type of file for printing in the editorial process, and it remained uncorrected due to the absence of my own proofreading. For the present publication, I have corrected all these errors and added new information, although I have kept revisions to a minimum to ensure the original argument is not altered.
1 Heinrich Mitteis, Deutsche Rechtsgeschichte, ein Studienbuch, new ed. Heinz Liebe-rich (Munich, 1978), p. 186.
2 Albert Brackmann, “The Beginning of the National State in Medieval Germany and the Norman Monarchies,” Medieval Germany 911–1250, trans. Geoffrey Barraclough, 2 vols. (Oxford, 1938), vol. 2, p. 289.
3 David C. Douglas, The Norman Fate 1100–1154 (Berkeley, 1976), pp. 2–3, 120, 217.
4 Antonio Marongiu, “Lo spirito della Monarchia normanna nell’Allocuzione di Ruggero II ai suoi Grandi,” Atti del Congresso internazionale di diritto romano e di storia del diritto. Verona 1948, ed. Guiscardo Moschetti, 4 vols. (Milan, 1951), vol. 4, pp. 315–327 (reprinted in Antonio Marongiu, Byzantine, Norman, Swabian and Later Institutions in Southern Italy [London, 1972]); Antonio Marongiu, “Concezione della sovranità ed assolutismo di Giustiniano e di Federico II,” Atti del Convegno Internazio nale di Studi Federiciani (Palermo, 1952), pp. 31–46 (reprinted in Marongiu, Byzantine, Norman, Swabian); Antonio Marongiu, “La concezione di sovranità di Ruggero II,” Atti del Convegno Internazionale di Studi Ruggeriani, 2 vols. (Palermo, 1955), vol. 1, pp. 213–233 (reprinted in Marongiu, Byzantine, Norman, Swabian).
5 Michele Amari, “Su la data degli sponsali di Arrigo VI con la Costanza erede del trono di Sicilia, e su i divani dell’azienda normanna in Palermo. Lettera del dottor O. HARTWIG e Memoria del Socio Amari,” Atti della R. Accademia dei Lincei, serie 3, anno 275 (1877–78), Memorie della classe di scienze morali, storiche e filologiche, vol. 2 (1878), pp. 409–438; Michele Amari, Storia dei Musulmani di Sicilia, 2nd ed., 3 vols., ed. Carlo A. Nallino (Catania, 1933–1939), vol. 3, pp. 451–473, 498–499, 541–553; Francesco Gabrieli, “La politique arabe des Normandes de Sicile,” Studia islamica, vol. 9 (1958), pp. 83–96; Jeremy Johns, “The Norman Kings of Sicily and the Fāṭimid Caliphate,” Anglo-Norman Studies, vol. 15 (1993), pp. 133–159; Jeremy Johns, “I re normanni e i califfi fāṭimiti. Nuove prospettive su vecchi materiali,” Del nuovo sulla Sicilia musulmana, ed. Biancamaria Scarcia Amoretti (Rome, 1995), pp. 9–50; Jeremy Johns, Arabic Administration in Norman Sicily (Cambridge, 2002), pp. 1–7, 193–300; Adalgisa de Simone, “Il Mezzogiorno normanno-svevo visto dall’Islam africano,” Il Mezzogiorno normanno-svevo visto dal’Europa e dal mondo mediterraneo (Bari, 1999), pp. 281–285; Alex Metcalfe, “The Muslims of Sicily under Christian Rule,” The Society of Norman Italy, ed. Graham A. Loud and Alex Metcalfe (Leiden, 2002), pp. 289–317.
6 Marongiu, “La concezione di sovranità di Ruggero II,” pp. 228–232; Francesco Giunta, Bizantini e bizantinismo nella Sicilia normanna, 2nd ed. (Palermo, 1974 [1st ed. 1950]).
7 Charles H. Haskins, “England and Sicily in the Twelfth Century,” English Historical Review, vol. 26 (1911), pp. 433–447, 641–665; Charles H. Haskins, The Normans in European History (Boston, 1915); Charles H. Haskins, Norman Institutions (Cambridge, MA, 1925), pp. 23–24, 61, 111–112, 232–234; Carmela Ceci, “Normanni d’Inghilterra e Normanni d’Italia,” Archivio scientifico del R. Istituto superiore di scienze economiche e commerciali di Bari, vol. 7 (1932–1933); Dione Clementi, “Notes on Norman Sicilian Surveys,” The Making of Domesday Book, ed. Vivian H. Galbraith (Oxford, 1961), pp. 55–58; Antonio Marongiu, “I due regni normanni d’Inghilterra e d’Italia,” I normanni e la loro espansione in Europa nell’alto Medio Evo, vol. 16 (Settimane di studio del centro italiano di studi sull’alto medioevo, Spoleto, 1969), pp. 497–557; Sally Harvey, “Domesday Book and Its Predecessors,” English Historical Review, vol. 86 (1971), p. 765. Some scholars think the nature of the kingship of Sicily was in substance same as other feudal monarchies in Europe: Léon-Robert Ménager, “L’institution monarchique dans les États normands d’Italie. Contribution à l’étude du pouvoir royal dans les principautés occidentales, aux XIe–XIIe siècles,” Cahiers de Civilisation médiévale, vol. 2 (1959), pp. 303–331, 445–468; Donald Matthew, The Norman Kingdom of Sicily (Cambridge, 1992), pp. 165–206.
8 Hiroshi Takayama, “The Financial and Administrative Organization of the Norman Kingdom of Sicily,” Viator, vol. 16 (1985), pp. 129–157; Hiroshi Takayama, “Familiares Regis and the Royal Inner Council in Twelfth-Century Sicily,” English Historical Review, vol. 104 (1989), pp. 357–372; Hiroshi Takayama, “The Great Administrative Officials of the Norman Kingdom of Sicily,” Papers of the British School at Rome, vol. 58 (1990), pp. 317–335; Hiroshi Takayama, The Administration of the Norman Kingdom of Sicily (Leiden/New York/Cologne, 1993).
9 Concerning the basic information on the administration of Roger I, see Takayama, The Administration, pp. 25–40; Graham A. Loud, The Age of Robert Guiscard (Harlow, 2000), pp. 146–185; Ruggero il Gran Conte e l’inizio dello stato normanno (Rome, 1977). See also Julia Becker, Graf Roger I. von Sizilien (Tübingen, 2008).
10 Gaufredus Malaterra, De Rebus Gestis Rogerii Calabriae et Siciliae Comitis auctore Gaufredo Malaterra, ed. Ernesto Pontieri (Bologna, 1927–1928) (hereinafter Mala-terra); Amatus Casinensis, Storia de’ Normanni di Amato di Montecassino, ed. Vincenzo de Bartholomaeis (Rome, 1935) (hereinafter Amatus); Romualdus Salernitanus, Chronicon sive Annales, ed. Carlo A. Garufi (Città di Castello, 1909–1935); Guillaume de Pouille, La Geste de Robert Guiscard, ed. Marguerite Mathieu (Palermo, 1961) (hereinafter Guillaume de Pouille).
11 Jordan was a natural son of Roger I and a brother of Geoffrey born of the same mother. Malaterra, Lib. III, Cap. XXXVI, p. 78: “Erat autem Jordanus ex concubina, tamen magnae viris animi et corporis et magnarum rerum gloriae suae dominationis appetitor.”
12 Malaterra, Lib. III, Cap. XXXVI, p. 78. This promising successor-to-be of Roger I died young in 1092, just in the same year of the completion of the conquest of Sicily (ibid., Lib. IV, Cap. XVIII, pp. 97–98). According to Necrologia Panormitana (ed. Eduard Winkelmann, Forschungen zur Deutschen Geschichte, vol. 18 [1898], pp. 473, 475), Jordan died on 17 (15 kal. oct.) or 18 September (14 kal. oct.) of 1091. But, as Pontieri (Malaterra, p. 98, note 1), editor of Malaterra, suggests, Roger I, together with Jordan, made donation to the monastery of St. Agatha at Catania in a diploma of 26 April 1092 (Catania, Archivio Capitolare della Cattedrale di Catania, Pergamene latine, n. 1: 26 April 1091, Ind. XV [= AD 1092]. Cf. Carl A. Kehr, Die Urkunden der normannischsicilischen Könige [Innsbruck, 1902], p. 14).
13 Malaterra, Lib. III, Cap. XVIII, p. 67: “Primus ad excubias Othonus, alter Elias, Tertius Arisgotus, Jordanus abinde remotus esse recusavit.”
14 He fought in the battle of Cerami in 1063. Malaterra, Lib. II, Cap. XXXIII, p. 42; Lib. II, Cap. XLVI, p. 53.
15 Malaterra, Lib. III, Cap. XXX, p. 75. Elias of Cartomi was probably a native of Cartomi in Spain. See Amari, Storia dei Musulmani, vol. 3, p. 156, note 1.
16 Malaterra, Lib. III, Cap. XXX, p. 75: “Porro Jordanus, filius comitis, et Robertus de Surda-valle et Elias Cartomensis – qui ex Saracenis ad fidem Christi conversus, postea apud Castrum-Johannis a sua gente hostiliter interfectus, quia negando apostata fieri noluit, martyrio vitam laudabiliter finivit – exercitu commoto, versus Cathaniam iter intendunt.”
17 Malaterra, Lib. III, Cap. X, p. 61.
18 Malaterra, Lib. II, Cap. XLVI, p. 53.
19 Amatus, Lib. IV, Cap. VIIII[sic, = IX]–X, pp. 231–3; Cap. XVIII, p. 237, note 2. Geoffrey Ridel became duke of Gaeta in 1068. Cf. Loud, The Age of Robert Guiscard, p. 153; Amatus, p. 274, note 1.
20 Malaterra, Lib. II, Cap. V; Lib. II, Cap. XXXIII, p. 42, note 2; Amatus, Lib. I, Cap. VIIII [sic, = IX]–XV, pp. 17–20. Roussel of Bailleul, who was active in Sicily in 1063, came to serve the Byzantine Emperor in the 1060s, and became an independent lord in Asia Minor after the battle of Mantzikert in 1071. Malaterra, Lib. II, Cap. XXXIII, p. 43; Amatus, p. 17, notes 1, 2, and p. 18, note 2; Loud, The Age of Robert Guiscard, p. 153; Jonathan Shepard, “The Uses of the Franks in Eleventh-Century Byzantium,” Anglo-Norman Studies XV: Proceedings of the Battle Conference 1992 (Woodbridge, 1993), pp. 299–302.
21 Kehr, Die Urkunden, p. 412: “†EGO ROGGERIVS COMES ME INTERSCRIPSI. †Signum Roberti Burrelli. †Signum Willelmi de Altavilla. †Signum Willelmi de Monte Piloso. †Ego Radulfus de Monte Piloso idem testor. †Signum proprie manus Goffredi filii comitis. †Signum Willelmi Culchebreti. †Signum Robberti de Bubone. †Signum Roberti Britonis filii Willelmi de Altavilla. †Signum Guidardi Orbec.”
22 Malaterra, Lib. II, Cap. XVII, p. 34: “Comes vero Rogerius, quietis impatiens et laboris avidus, trecentos juvenes secum ducens, usque Agrigentum praedatum et terram inspectum vadit, totam provinciam incendio concremando devastans.” Malaterra, Lib. II, Cap. XVIII, p. 35: “Media vero hieme, videlicet ante natalem Domini, cum ducentis quinquaginta militibus iterum mare transiens, usque ad Agrigentinam urbem, totam patriam sollicitans, praedatum vadit.” Malaterra, Lib. II, Cap. XXIX, p. 39: “iterum Siciliam cum trecentis debellaturus aggreditur.” Cf. Ferdinand Chalandon, Histoire de la domination normande en Italie et en Sicile, 2 vols. (Paris, 1907), vol. 1, p. 328.
23 In fact, a large revolt of Norman barons broke out on the peninsula in 1064 and 1072.
24 Amatus, Lib. VI, Cap. XVIIII [sic, =XIX], p. 281: “Et, en celle nuit, se esmurent o tout li ostage, et manderent certains messages liquel doient dire coment la terre s’est rendue. Et puiz, quant il fu jor, dui Cayte alerent devant, loquel avoient l’ofice laquelle avoient li antique, avec autrez gentilhome.” Cf. Amari, Storia dei Musulmani, vol. 3, pp. 130–131.
25 Malaterra, Lib. II, Cap. XLV, p. 53: “Quandoquidem fortuna praesenti sic hortabantur, urbis deditionem facere, se in famulando fideles persistere, tributa solvere: et hoc jura-mento legis suae firmare spopondunt.” Guillaume de Pouille, Lib. III, p. 182: “Cuncta duci dedunt, se tantum vivere poscunt. Deditione sui facta meruere favorem Exo-rare ducis placidi; promittitur illis Gratia cum vita. Nullum proscribere curat, Observansque fidem promissi, laedere nullum, Quamvis gentiles essent, molitur eorum. Omnes subiectos sibi lance examinat aequa.” Cf. Amari, Storia dei Musulmani, vol. 3, pp. 130–131, 277; Chalandon, Histoire de la domination, vol. 1, p. 208; Loud, The Age of Robert Guiscard, pp. 161–162.
26 Malaterra, Lib. II, Cap. XLV, p. 53: “Proximo mane primores, foedere interposito, utrisque fratribus locutum accedunt, legem suam nullatenus se violari vel relinquere velle dicentes, scilicet, si certi sint, quod non cogantur, vel injustis et novis ligibus non atterantur.”
27 Amari, Storia dei Musulmani, vol. 3, p. 132; Chalandon, Histoire de la domination, vol. 1, p. 208; Francesco Gabrieli, “La politique arabe des Normands de Sicile,” Studia Islamica, vol. 9 (1958), p. 93.
28 Amari, Storia dei Musulmani, vol. 3, p. 277.
29 Malaterra, Lib. III, Cap. X, p. 61: “Comes vero, quibusdam necessitatibus se vocantibus, a Sicilia versus Calabriam digrediens, Hugonis de Gircaea, cui, propter strenuitatem, quam habebat – nam et praeclari generis a Cenomanensi provincia erat – cum filia sua de priore uxore Cathaniam dederata, totam Siciliam servandam delegavit, interdicens ne, si Bernarvet, quia vicinius sibi Syracusis morabatur, aliquem incursum versus se faceret, callidas eius versuitias cavens, nusquam urbe digregiens, hostem persequeretur.”
30 Malaterra, Lib. III, Cap. XXX, p. 75: “Hic quendam paganum, nomine Benthumen, quem comes apud Cathaniam majorem urbi praefecerat, callidis circumventionibus aggrediens, ad tradendam urbem multis munerum, possessionumve pactionibus sollicitabat. Paganus vero nominis sui competens imitator, avaritia coecatus, fidei sacramentorumque, quae comiti dederat, oblitus, statuto termino, infra urbem illum cum multitudine suorum fraudulenter de nocte accipiens, traditionis nomen sibi perpetuo vindicavit.”
31 Malaterra, Lib. III, Cap. XXX, p. 76.
32 Malaterra, Lib. IV, Cap. II, p. 86; Lib. IV, Cap. V–VI, pp. 87–88.
33 Robert Guiscard died on 17 July 1085 in Cephalonia during his expedition against the Byzantine Empire. Guillaume de Pouille, Lib. V, pp. 252–254. Cf. Loud, The Age of Robert Guiscard, pp. 222–223.
34 Becker, Graf Roger I., pp. 245–259.
35 Palermo, Archivio di Stato, Tabulario di Abbazia di S. Filippo di Fragalà e di S. Maria di Maniaci, Pergamene, n. 1. Facsimile: Salvatore Cusa, I diplomi greci ed arabi di Sicilia pubblicati nel testo originale, vol. 1 (Palermo, 1868–1882), Tav. III (Diploma of Count Roger II, 7 May, AM 6625, Ind. X [= 1117], Mileto, Greek. Edition: Cusa, pp. 383–385; Giuseppe Spata, Le pergamene greche esistenti nel grande archivio di Palermo [Palermo, 1862], pp. 245–248, which has errors concerning the dates on p. 703). This includes the privilege granted by his father Roger I (June, AM 6598, AD 1090, Ind. XIII, Greek and Latin).
36 Cusa, p. 385; Spata, p. 248: “huius rei sunt testes. gofridus filius comitis rogerii. gofridus stratigotus. paganus de gorgusio. willelmus capriolus. willelmus de surdavalle. hugo de puteolis. gofridus senescalcus.”
37 According to the Greek text, this document was issued in the presence of Geoffrey, son of the count; his real brother Jordan; Robert Borrell; William the seneschal; Stephen the iatros; Stephen the chaplain; Basil, son of Tricari the archon of Demena; Nicholas the prōtonotarios, chamberlain and prōtospatharios; Lord Ūrsinos the notarios; and Nicholas, son of the prōtospatharios Garzēfa. Cusa, p. 384; Spata, p. 247: “γέγωνεν δὲ τὸ παρὸν χρυσωβούλλιον κατενώπιον γιοσφρέδα υἱοῦ κόμητος, καὶ ἰορδάνου ἀπταδέλφου αὐτοῦ καὶ ῥουμβέρτου βουῤῥέλλου καὶ γουλιάλμου συνεσκάλκου καὶ στεφάνου ἰατροῦ καὶ στεφάνου καππελλάνου καὶ βασιλείου υἱοῦ τρηχάρι ἄρχοντος δεμέννων καὶ νικολάου πρωτονοταρίου καὶ καπριλλίγγουα καὶ πρωτοσπαθαρίου καὶ νοταρίου κυροῦ οὐρσίνου καὶ νικολάου υἱοῦ πρωτοσπαθαρίου τοῦ γαρζήφα καὶ ἑτέρων πλήστων.”
38 Francesco Trinchera, Syllabus graecarum membranarum (Naples, 1865), doc. LX, p. 78: “Quod actum est teste et concedente. Adelaide. coniuge. mea et Malgerio filio meo. Willelmo. de alta villa. Iosberto de luciaco. Roberto borrello. Pagano de gorgiis. Rogerio de stilo. Iohanne prothonotario de traginiis. Nichola de mesa. Giraldo capellano meo. et Fulcone fratre eius capellano meo. Hugone de melfia. Jeremia de sancto egidio capellano meo.”
39 Trinchera, Syllabus, doc. LXIX, p. 87: “κομιτίσσης αδυλασιας καὶ απεδωκα τον χαρτὶον εὶς χειρὰς του κυρου λανυινου ενοπιον ρονκεριου επισκοπου συρακουσιου καὶ ρουμυερτου δε παρις καὶ ροὺμυὲρτου υουρελλου καὶ γιοὺσυὲρτου καὶ λοὺτζι†καὶ νικολαου μεσῶν”; “Comitissa adelaide. et roberto borrello et iosberto de luciaco. et roberto electo melitensi atque Rogerio siracusano episcopo. et nicolao canberlario”; “hanc chartulam tradidi in manus domini Lanuini, coram Rogerio Syracusano Episcopo, et Roberto de Paride, et Roberto Borello, et Giusberto, et Lutzi † et Nicolao cubiculariis.”
40 Salvatore Fodale, “Il gran conte e la sede apostolica,” Ruggero il gran conte, pp. 27–32; Cosimo D. Fonseca, “Le istituzioni ecclesiastiche dell’Italia meridionale e Ruggero il gran conte,” Ruggero il gran conte, pp. 46–58; Chalandon, Histoire de la domination, vol. 1, pp. 342–347. For Roger’s church policy, see also Lynn T. White Jr., Latin Monasticism in Norman Sicily (Cambridge, MA, 1938); Mario Scaduto, Il monachesimo basiliano nella Sicilia medievale. Rinaschita e decadenza (sec. XI-XIV) (Rome, 1947); Tommaso Leccisotti, “Ruggero II e il monachesimo benedettino,” Atti del Convegno Internazionale di Studi Ruggeriani, vol. 1, pp. 63–72; Léon-Robert Ménager, “La ‘byzantinisation’ religieuse de l’Italie méridionale (IXe–XIIe siècles) et la politique monastique des normands d’Italie méridionale,” Revue d’histoire ecclésiastique, vol. 53 (1958), pp. 747–774; vol. 54 (1959), pp. 5–40.
41 Fonseca, “Le istituzioni ecclesiastiche,” pp. 46–47. The diocese of this bishopric was defined by the diploma of Roger I in 1087.
42 Malaterra, Lib. III, Cap. XIX, p. 68, note 2; Lib. III, Cap. XXIII, p. 101; Lib. III, Cap. XXIX, p. 107. Roger I defined its diocese in February of 1081, and gained the papal confirmation from Gregory VII in 1082.
43 Malaterra, Lib. III, Cap. XLV, p. 53. Alcherius received a bull of Gregory VII dated on 16 April 1083 (Philipp Jaffé, Samuel Löwenfeld, Wilhelm Wattenbach, Ferdinand Kaltenbrunner, and Paul Ewald, eds., Regesta pontificum Romanorum ab condita ecclesia ad annum post Christum natum MCXCVIII, 2 vols. [Leipzig, 1885–1888], vol. 1, pp. 644–645, no. 5258; Rocco Pirro, Sicilia sacra disquisitionibus et notitiis illustrata, 2 vols., reprint ed. Antonino Mongitore [Palermo, 1733], vol. 1, p. 70). Cf. Fonseca, “Le istituzioni ecclesiastiche,” p. 52; Scaduto, Il monachesimo basiliano, p. 51; Dieter Girgensohn, “Dall’episcopato greco all’episcopato latino nell’Italia meridionale,” La chiesa greca in Italia dall’VIII al XVI secolo. Atti del Convegno storico interecclesiale (Bari 30 aprile-4 maggio 1969), vol. 1 (Padova, 1973), p. 30.
44 Malaterra, Lib. III, Cap. VII, p. 89: “Ecclesias passim per universam Siciliam fieri imperat; ipse pluribus in locis de suo sumptus, quibus facilius fiant, attibuit. In urbe Agrigentina pontificalibus infulis cathedram sublimat: terris, decimis et diversis copiis, quae pontifici et clero competenter designata sufficiant, haereditaliter chirographis suis dotat, ornamentis et sacri altaris utensilibus ed plenum consignatis. Huic ecclesiae Gerlandum quendam natione Allobrogum, virum, ut ajunt, magnae charitatis et ecclesiasticis disciplinis eruditum, episcopum ordinans, praefecit. Haud secus apud Mazariam facere addens, omnibus quae rite sufficienter praelato et clericis ad plenum designatis, Stephanum, quendam Rothomagensem, honestae vitae virum, episcopum ordinavit. Apud Syracusam vero idem adjicens Rogerium, decanum ecclesiae Traynensis, hones-tae eruditionis clericum et boni moris et affabilitatis virum, in Provincia ortum, pontificalibus infulis sublimavit…. Apud Sanctam Euphemiam vero, monachum quendam, natione Britonem, virum religiosum, post abbatem totam ecclesiam prudenti mode-ramine audiens, ut hunc ecclesiae Cathaniae – si impetrare queat – episcopum ordinet, intendit…. Sicque solemniter episcopatum concedens, quod nulli episcoporum fecisse cognoscitur, totam urbem sedi suae cum omnibus appendicis suis sub chirographo et testibus haereditaliter possidendam assignavit.” See Fodale, “Il gran conte e la sede apostolica,” pp. 31–32; Fonseca, “Le istituzioni ecclesiastiche,” pp. 49–50. In Catania, however, a Greek bishop kept his office at least until 1103, and in Syracuse a Greek cleric was still working under the Latin Bishop in 1093. In 1096 the diocese of Messina was combined with that of Troina, and entrusted to Robert, bishop of Troina. In Ca labria the first bishop of Squillace, John Nicephoros, was appointed by the papal legate in 1096. Giunta, Bizantini e bizantinismo nella Sicilia normanna, p. 31; Fodale, “Il gran conte e la sede apostolica,” p. 32. The foundation of the Latin bishoprics of Syracuse, Catania and Agrigento has been placed between 1086 and 1088 by Chalandon (Histoire de la domination, vol. 1, p. 344).
45 Robert, bishop-elect of Mileto, appeared in 1095 and 1101; Roger, bishop of Syracuse in 1101; and Anger, bishop of Catania, in 1095. Besides these high ecclesiastics, comital chaplains seem to have played important roles as members of the entourage as well as scribes. We see several chaplains in sources, such as Stephen in 1090, and Girald, his brother Falco and Jeremia de Sancto Egidio in 1097.
46 Jules Gay, L’Italie méridionale et l’empire Byzantine, depuis l’avènement de Basile Ier jusqu’à la prise de Bari par les Normands, 867–1071 (Paris, 1904), pp. 556–560; George Ostrogorsky, History of the Byzantine State, trans. Joan Hussey, rev. ed. (New Brunswick, NJ, 1969), p. 247. For the Byzantine officials in Southern Italy, see also Vera Von Falkenhausen, Untersuchungen über die byzantinische Herrschaft in Süditalien vom 9. bis ins 11. Jahrhundert (Wiesbaden, 1967); Vera Von Falkenhausen, La dominazione bizantina nell’Italia meridionale dal IX all’XI secolo (Bari, 1978).
47 Gay, L’Italie méridionale, pp. 559–560. We have much information concerning the kritēs under Byzantine rule. A certain kritēs Eupraxios appeared in Rossano in the latter half of the tenth century. In 1026, Leo, kritēs of Langobardia and Calabria, presided over a court at Taranto. In 1048 Cricorius, an imperial kritēs of Italy, appeared in Bari. See Evelyn Jamison, “The Norman Administration of Apulia and Capua, More Especially under Roger II and William I, 1127–1166,” Papers of the British School at Rome, vol. 6 (1913), pp. 302–303; Gay, L’Italie méridionale, p. 556.
48 We do not have much information about Leo the logothetēs, but a later document suggests that he concurrently held this office and that of megas kritēs of Calabria (μέγας κριτὴς πάσης καλαβρίας). Bernard de Montfaucon, Palaeographia graeca (Paris, 1708), Lib. VI, p. 402, a document dated September, Ind. V (= AD 1126), AM 6639 (= AD 1130) (Montfaucon attributes the date to AD 1131); cited by Jamison, “The Norman Administration of Apulia and Capua,” pp. 303–304.
49 André Guillou, “Lo svolgimento della giustizia nell’Italia meridionale sotto il Gran Conte Ruggero e il suo significato storico,” Ruggero il Gran Conte e l’inizio dello stato normanno, p. 72, note 12; Montfaucon, Palaeographia graeca, Lib. VI, p. 394.
50 Les actes latins de S. Maria di Messina (1103–1250), ed. Léon-Robert Ménager (Palermo, 1963), pp. 27–42.
51 Cusa, p. 390 (December 1094); Cusa, p. 643 (20 August 1099). Besides stratēgos, we find viscounts (vicecomites, βεσκόμητες), foresters (φορεστάριοι), exousiastai (ἐξουσιασταί), which were described as baiuli in Latin documents, and tourmarchoi (τουρμάρχοι).
52 Palermo, Archivio di Stato, Tabulario di Abbazia di S. Filippo di Fragalà e di S. Maria di Maniaci, Pergamene, n. 8 (Original. May, AM 6613 [= AD 1105], Indiction XIII. Edition: Cusa, pp. 399–400; Spata, p. 203): “κόμιτος ὡς (Spata, κόμητως) ὁδηγούμενος παρὰ τοῦ ἁγίου πνεύματος ὁμοίως καὶ παρὰ τῶν ἀρχόντων τῶν τιμίων λέγω δὴ νικολάου τοῦ ἐνδοξοτάτου καπριλίγγα καὶ τοῦ ἐνδοξοτάτου λέοντος τοῦ λογοθέτου, οἵτινες καὶ αὐτὸς καὶ ὁ προσηνέστατος (Spata, προσηνιστατος) τῆς εὐγενείας εὐγένιος (Spata, εὐγὲνιος).”
53 Amari, Storia dei Musulmani, vol. 2, p. 34; Amari, “Su la data,” p. 430; Mario Cara-vale, “Gli uffici finanziari nel Regno di Sicilia durante il periodo normanno,” Annali di storia del diritto, vol. 8 (1964), pp. 185–187; Chalandon, Histoire de la domination, vol. 1, p. 348.
54 Pirro, Sicilia sacra, vol. 1, p. 384: “Unde audita ejus petitione pro salute animae meae, et fratris mei nobilissimi Ducis Roberti Guiscardi… dedi, et in perpetuum concessi Ecclesiae S. Nicolai Episcopii Messanae, casale Saracenorum, quod dicitur Butahi cum omni tenimento, et pertinentiis suis secundum antiquas divisiones Saracenorum.”
55 Johns, Arabic Administration, p. 41.
56 The writ of transfer of 12 February, which was written in Palermo, has a foreword in Greek, seventy-five names of inhabitants in Arabic and an afterword in Greek (Cusa, pp. 1–3). That of 20 February has a foreword in Greek, 398 names of inhabitants in Arabic and an afterword in Greek (Cusa, pp. 541–549).
57 In Calabria too, as Caravale suggests, there seem to have remained lists of inhabitants similar to those of the Muslims in Sicily. The list of inhabitants in the writ of transfer of the village of Laco in Calabria to the archbishop of Palermo issued in December, AM 6601 (= AD 1092), had been written in Greek. See Caravale, “Gli uffici finanziari,” pp. 187–188; Pirro, Sicilia sacra, vol. 1, pp. 77–78.
58 Cusa, pp. 548–549: “Εγράφη οἱ τιαύτη πλατεῖα τῆ προστάξη ἐμοῦ κόμητος ῥωγερίου τῆς γ’ ἰνδικτιῶνος τοῦχγ’ ἔτους ὄντος μου ἐχ τὴν μεσσίνην, αἱ δαὶ ἄλλαι πλατείαι τῆς ἐμῆς χώρας καὶ τῶν ἐμῶν τερρερίων ἐγράφησαν ἐχ τὸ μαζάρρη τοῦ χα’ ἔτους τῆς α’ ἰνδικτιῶνος. καὶ διὰ τούτω προστάττομεν ὅτι ἐάν τις εὐρέθη ἐχ τὰς ἐμὰς πλατείας ἤτε ἐχ τὰς πλατείας τῶν τερρερίων μου ἐκ τοὺς ἀγαρινοὺς τοὺς ὄντας γεγραμμένους ἐχ τὴν τοιαύτην πλατεῖαν ἵνα ἀντιστρέφη αὐτοὺς ὁ ἐπίσκοπος ἄνευ πάσης προφάσεως.”
59 As for the dīwān al-taḥqīq al-ma‘mūr or duana de secretis in a later period, see Takayama, The Administration, pp. 81–84; Takayama, “The Financial and Administrative Organization,” pp. 129–157.
60 Concerning the office of amiratus, see Hiroshi Takayama, “Amiratus in the Norman Kingdom of Sicily: A Leading Office of Arabic Origin in the Royal Administration,” Forschungen zur Reichs-, Papst- und Landesgeschichte, ed. Karl Borchardt and Enno Bünz (Stuttgart, 1998), pp. 133–144.
61 Guillaume de Pouille, Lib. III, vers 340–343, p. 182: “Obsidibus sumptis aliquot castrisque paratis, Reginam remeat Robertus victor ad urbem, Nominis eiusdem quodam remanente Panormi Milite, qui Siculis datur amiratus haberi.”
62 Cava de’ Tirreni, Archivio della Badia della Santissima Trinità, Arca magna, Armarium C, nos. 5, 6 (Edition: Léon-Robert Ménager, Recueil des actes des ducs normands d’Italie (1046–1127), vol. 1: Les premiers ducs (1046–1087) (Bari, 1981), nos. XLIV, XLV). See Léon-Robert Ménager, Amiratus – Ἀμηρᾶς. L’Émirat et les origines de l’amirauté (XIe–XIIIe siècles) (Paris, 1960), p. 25, and Appendice II, nos. 1, 2, pp. 167–168. Ménager thinks at least one of the two diplomas is a falsification based on an original document.
63 Malaterra, Lib. II, Cap. III, IV, XVI–XXII, pp. 30, 34–36.
64 Malaterra, Lib. III, Cap. XVIII, p. 67.
65 Eadmer, Vita Sancti Anselmi: The Life of St. Anselm, Archbishop of Canterbury, ed. Richard W. Southern (London, 1963), pp. 111–112; Chalandon, Histoire de la domination, vol. 1, p. 304.
66 Malaterra, Lib. IV, Cap. V–VI, pp. 87–88.
67 For George, see Ménager, Amiratus, pp. 44–53; Takayama, “Amiratus,” pp. 138–140; Takayama, The Administration, p. 53.