Source: PROCEEDINGS OF HISTORY WEEK. (1993)(1-29)
The official report on Malta and Gozo made in about 1241 survives only in the form of a letter from Frederick II to Giliberto Abbate, his governor in Malta, in which the emperor’s chancery recapitulated information sent to him by Giliberto. Frederick was concerned primarily with the incomes of the Sicilian royal curia to which the Maltese islands belonged, and the result was a document now of fundamental importance in providing not only the earliest surviving written statistics for Maltese history but also an outline of the way in which the islands of Malta and Gozo fitted into the administrative pattern of the Sicilian kingdom. This study presents a revised and annotated edition of the text together with a discussion of the technical difficulties involved and an estimation of its implications for the history of Malta and Gozo and for the structure of their society at a key stage in their Latinization and Christianization. There are very few documents which illuminate the history of Malta in the time of Frederick II and it should be emphasized that interpretations of the period remain tentative and hypothetical, especially as evident corruptions in the surviving version of this text impose complicated considerations of the population and taxation details it includes.
The text printed below does not contain Giliberto’s original report. Nor does Frederick II’s chancery’s summary of it survive in its original form or in the register in which it was originally recorded, but in the so-called Excerpta Massiliensia. That codex contains documents from the royal registers for the years 1230 to 1248 which were recopied in an Angevin milieu, presumably for the use of the royal curia, at the end of the thirteenth or the beginning of the fourteenth century. The extracts it contains were apparently taken from a whole range of materials rather than from a particular series of registers; there were numerous omissions and abbreviations, while of ten there was no date. Discrepancies in the taxation statistics in the surviving version of Giliberto’s report, which do not add up correctly, mean that both the taxation and the population figures, which result from successive copying and recopying, should be used with considerable caution; furthermore, a number of words and figures are clearly corrupt while some incomes are given in round figures which suggest approximations or estimations rather than genuine statistics.
[p.2] The document is in Marseilles, Archives départmentales des Bouches du Rhóne, B 175, f. 14 verso - 15 recto (no. 78). It was published by E. Winkelmann in 1880,  and subsequently republished, without comment, first by A. Mifsud  and again, with a number of omissions and an occasionally unsatisfactory English translation, by C. Dessoulavy.  The problems of the Excerpta, and the relevant literature, are surveyed by W. Hagemann,  while E. Sthamer made a study of the documents in the register on the basis of the handwriting but added nothing with regard to this Maltese document.  The question of the document’s date is especially difficult but, to judge merely from its place in the codex, it belonged either to about 1241 or to 1245/6 as it was placed between two groups of documents of those dates;  Frederick II died in 1250. 1241 is probably, but by no means conclusively, to be preferred on account of the reference to quondam Paolino de Malta who is known to have been Frederick’s governor of Malta in 1239 and 1240, and possibly in 1235.  The report has been discussed by many [p.3] authors, including R. Valentini,  A. Luttrell,  H. Bresc,  A. Pertusi,  I. Peri  and G. Wettinger. 
The present edition contains a number of variations with respect to Winkelmann’s text. Capital letters, but not punctuation or grammar, have been rationalized. The manuscript itself provides uncertainties and inconsistencies concerning the expansion of words such as tar’ and in the case endings of numerals. The text at Marseilles sometimes gives these in the nominative and sometimes in the accusative, but contractions are here expanded in the accusative following the practice in a surviving fragment of the register of Frederick II for 1239/40. Where the Marseilles text gives the nominative it seems likely that the scribe would have copied what was a Roman numeral rather than a number which had been written as a word in the original version. 
The Giliberto who governed Malta and Gozo has often been considered to have been an abbot, though it might seem unlikely that a religious would have been appointed to rule the islands. In reality, he came from an important and wealthy Sicilian family, the Abbate of Trapani. Giliberto was very probably the brother of Enrico Abbate, who was a leading functionary of Frederick II in 1239 and consul in Tunis in 1240, and who was secreto in Sicily from about 1253 to 1256.  Giliberto was dead by 1269, but his son Palmerio Abbate, who had been [p.4] governor of Pantelleria and of Favignana in the Egadi islands,  was a leading Sicilian who seems to have been one of the promoters of the Vespers uprising of 1282 when, according at least to Lu Rebellamentu di Sichilia, he was among those who moved to Malta to continue negotiations with Giovanni of Procida for the overthrow of the Angevin regime.  As governor of Malta, Giliberto Abbate fitted into a pattern of family predominance in Trapani, Erice, the Egadi isles and the waters between Sicily and Tunisia, a cultural and political zone which in his time extended to the Maltese islands. 
Any consideration of the structure of Maltese society in the thirteenth century must depend upon an interpretation of Giliberto’s report and of the related problems of its population statistics, remembering that his letter to Frederick II mentioned the quaternitiones or registers kept by his predecessor as governor and that Giliberto himself presumably maintained administrative records; he thus had access to reliable written information.  Such bureaucratic procedures were in the tradition of Norman practice in Sicily where the government kept tax registers containing the narnes of the villani; originally these were written in Arabic but later they were in Latin as well.  There may have been such registers on Malta, possibly in Arabic, but virtually nothing is known of the island’s administration in the twelfth century. The terminology of Giliberto’s text was basically Sicilian but, as the document itself explicitly emphasized, Maltese customs and institutions differed fom those of Sicily. It seems probable that Malta, unlike Sicily, was almost exclusively Muslim during the twelfth century,  though the presence of at least a small group of Christians on Malta is documented [p.5] for the reign of Roger II who died in 1154.  Assuming that these Christians were not the descendants of indigenous islanders who had somehow survived Muslim rule, they may have been merchants and administrators who possibly settled only in the port or in urban areas, perhaps following the second Norman “conquest” of Malta by Roger II in 1127.
The situation on Malta may, in some ways at least, have been like that of Pantelleria where the tolerant arrangements introduced by the Normans continued into the fifteenth century. On Pantelleria the population paid tribute but remained Muslim; the island’s Muslim governor, appointed by the Sicilian crown, governed according to Muslim laws; and in 1282 the king had to send an Arabic-speaking notary there from Palermo. There is some uncertain evidence that in 1115 the Normans restored an ancient Greek monastery on Pantelleria and that in the fifteenth century the Greek rite was in use there.  It is significant that Roger II imposed on the Christians of Malta and Gozo an annual fine paid in consequence of their having killed a Muslim,  and a Maltese legend of uncertain antiquity which concerns a massacre of Muslims by Christians may constitute a distorted reflection of some such event. 
It is uncertain to what extent Giliberto’s report, essentially a bureaucratic and in no sense a “feudal” document, was concerned exclusively with royal matters such as curial incomes and possessions or the royal garrison and servants; it seems likely, though not certain, that the numbers of families given included all, or almost all, those on the islands rather than only those who were directly dependent on the crown and curia. In Sicily the distinction between what belonged to or was owed directly to the royal curia or demanium and what was let out by the curia, in fief or otherwise, was often unclear. In Malta the land must initially have remained in Muslim hands after 1091, and those Christians who arrived [p.6] during the twelfth century may for a time have lived in the towns without occupying any land. In 1198 the men of Malta and Gozo received a royal charter incorporating their islands to the royal demanium in perpetuity,  yet from about 1192 to about 1222 Malta was a county under a series of counts, and in 1212 Enrico Count of Malta was granted permission to strike his own coinage.  These counts may well have taken over the royal position on the islands without seriously altering their taxation or land-holding systems. In 1222 Frederick II imprisoned Count Enrico and removed him from his countship, and though by 1223 he had been restored to royal favour, he was not given back the castle: nulla restitutione facta de castro Malte.  Enrico’s son Nicoloso later called himself count and he retained claims to the islands but in 1243 he was not, apparently, using the title.  The crown had evidently taken back direct control of the islands. Paolino de Malta was royal proctor there by 1239 and in 1241 circa his successor Gilberto Abbate controlled the royal demanium and worked it with royal villani or villani curie. What lands were then held by the Christians, whether Latin immigrants or converted Muslims, is not known.
There was no sign in Giliberto’s report of a count, of fiefs or of an ecclesiastical or monastic establishment. The document did mention the homines and fideles of the islands who had their own customs and who gave consilium to the governor; many of them had usurped curial rights. Some decades later, in 1270, the governor was to act cum consilio sapientum.  The islands’ homines or fideles - the terms seem to have been interchangeable -certainly included Muslims when the royal privilege or incorporation was granted in 1198 to universus populus totius insule Malte et Gaudisii tam Christiani quam Saraceni fideles nostri.  Conceivably, some of these homines held lands and maybe some had [p.7] villani who were not royal villani curie. The surviving version of the text of Giliberto’s report does contain one use of the phrase feudales nostri Sicilie, but it seems likely that feudales was a copying error for fideles and that the creation of a county had not led to enfeoffments or sub-enfeoffments within the Maltese islands. Nothing else is known of local laws and customs, except that a document of 1270 claimed that in the time of Frederick II certain homines of Malta and Gozo, whose status is not clear, had been obliged to watch over the swine on the royal massarie or estates. 
The most straightforward approach to the population question is to assume that, with certain statistically minor exceptions such as the garrison and slaves, the number of families reported did in fact represent all the settled inhabitants of the islands. It is curious that the population statistics in the text appear among the details concerning the servientes who depended on the curia, perhaps as a result of some rearrangement in the process of summarizing the original report, but the main difficulty is that the apparently unlikely figure of 47 Christian families for Malta does not seem to fit with the taxation statistics or with the proportion of Christians to Muslims on Gozo; it has usually been considered an error. It could in theory be that on Malta there were other Christians who were not listed for the reason that they paid no taxes to the curia, but that would seem unlikely and it would leave it unclear why so many Christian families were listed on Gozo. If the text is assumed to be corrupt, one answer would be to amend quadraginta septem to quadr[ingentas] septem or 407.  Palaeographically this would be quite likely, but it would give only 407 Christian families in Malta as against 203 in Gozo; it would leave the Christian families on Malta in a minority of 407 against 681 Muslim families there; it would give a hypothetical total population for the islands of only some 4500; and it would provide a set of ratios unlike the more regular series tabulated below as the result of the alternative assumption that 47 should be amended to 1047. It is certainly important to establish these ratios and proportions, but deductions from them may be suspect in so far as they are based on an assumption, which is not necessarily entirely justitied, of close similarities between Malta and Gozo. It could be argued that, as has been the case in all recent centuries, there would have been a preponderance of administrators and immigrants on Malta rather than on Gozo in the thirteenth century; but in that case 407 Christian families for Malta would be too low.
[p.8] Yet another possibility is that there were 747 Christian families on Malta, since that figure would produce very satisfactory ratios of 3.7 Christians on Malta to each Christian on Gozo, of 4 families on Malta to each family on Gozo, and of 1.1 Christians to each Muslim in Malta as against 1.3 Christians to each Muslim in Gozo; that would result in a total of 1453 families in Malta and a Christian majority in Malta, making a total of 1819 families or, hypothetically, around 7276 people for the two islands. In that case the situation would be as follows:
Such an argument, derived strictly from a mathematical consideration of ratios, may well provide an acceptable solution,  but it is not explicable palaeographically. Though less statistically defensible, at least on the basis of the text in its surviving form, it seems easier to envisage the conflation of an m for mille with the abbreviation m of christianorum resulting, during a subsequent copying, in its complete disappearance; that is christianorum mille or christianorum m becoming christianor’ m, then christianor’m and finally christianor’, as in the surviving manuscript.  The resulting figure of 1047 [p.9] Christian families on Malta would produce a slightly higher proportion of Christians to Muslims on Malta, at 1.5 to one, than on Gozo, at 1.3 to one, as could indeed have been the case:
A total of 1047 Christian families on Malta would imply a settled population on the two islands of nearly 8500 if a multiplier of 4 were used. Peri suggests a multiplier of 3.5 or 4 for thirteenth-century Sicily  but Bresc’s multiplier of 5.3 for 1277 seems very high.  The fifteenth-century Maltese statistics might imply a multiplier of 5 or even 6, and a total population in 1419/20 of around 10,000.  A multiplier of 6 would give a remarkably high total population of around 12,700 for 2119 families in about 1241. Unfortunately, all such calculations ignore a number of complicating factors and are based partly on guesswork.
This discussion raises the question of the date or dates at which groups of Muslims were expelled from Malta. There had been a major Muslim revolt in Sicily in 1189, and from 1220/1 there was a series of Sicilian wars, mainly in the mountains; the final campaign there began in 1243 and ended in 1246.  There were certainly expulsions of Muslims from Sicily to Lucera in Apulia in 1223.  In 1224 there was also a transportation to Sicily of Italian rebels from Celano in [p.10] the Abruzzi. The main evidence is provided by Riccardo di San Germano who wrote two versions of his reliable work, the first being essentially a monastic chronicle of Monte Cassino and the second a history of Sicily; he died in December 1243 or early in 1244.  The first version reported that in May 1224 Enrico de Morra, acting on the orders of the Emperor Frederick II, sent the rebels from Celano to Sicily with their wives and children, and that in 1227 Frederick II had them freed: Celanenses omnes, qui captiui in Sicilia tenebantur, liberi dimittuntur, Imperatore mandante. This version made no mention of Malta. The second version stated instead that, at the emperor’s command, the Celanesi were in 1224, apparently in May, moved to Sicily by Enrico de Morra, and that Frederick II sent them to Malta: Henricus de Morra iussu imperiali Celanenses reuocat ubique dispersos, ut ad propria redeant, et redeuntes capit et in Siciliam mittit, quos apud Maltam dirigit Imperator. The second version was not continued beyond 1226 and so did not cover the liberation of 1227 mentioned in the first version.  Enrico Count of Malta may in some way have been involved; he lost control of Malta itself in about 1222, but in the following year he was active in the suppression of the Muslim revolt in Sicily; and, significantly perhaps, he was at Celano in April 1223 and at Catania during 1224.  There may have been people from Celano who went to Sicily but not to Malta and who were freed by Frederick II in 1227. That at least some Celanesi were at some point sent to Malta seems certain, since in 1252 Marino, son of Andreas Bacconensis of Celano, addressed a petition to the pope in which he claimed that his father had, on account of his devotion to the Roman Church, been captured by followers of Frederick II and imprisoned on Malta, where he died after much suffering.  There is, therefore, some evidence for the expulsion of Muslims from Sicily and for the transportation of Christians to Malta and Gozo in about 1224.
The so-called Annales Siculi recorded further expulsions of Sicilian Muslims to Lucera in 1245: de mandato domini imperatoris comes Riccardus de Caserta [p.11] ejecit omnes Saracenos de Sicilia.  Furthermore Ibn Khaldun, who was a great scholar but who was writing over a century later and was not well informed about Malta, stated that following the death of Abu Zakaraya Yahya, the Hafsid Sultan of Tunis who died on 2 October 1249, the Muslims of Palermo revolted and Frederick II besieged them and then sent them to Lucera: When the news of the death of the Emir Abu Zakaraya reached Sicily, there were Muslims living in the city of Palermo. The sultan had negotiated on their behalf with the lord of the island [Frederick II] an agreement concerning the city and its environs. They lived side by side [with the Christians] until [the news of] the death of the sultan reached them. Then the Christians attacked them furiously, and they took to their strongholds and to the wilds. They gave command over themselves to a rebel from the Bani Abs. The tyrant of Sicily [Frederick II] surrounded them in their stronghold in the mountains and besieged them until he forced them to climb down. He carried them over the sea to the mainland and granted them Lucera as their dwelling. Then he crossed to the island of Malta and expelled the Muslims who were in it, and joined them with their brothers [in Lucera]. The tyrant of Sicily conquered Sicily and its islands, and he drove from them the word of Islam and replaced it with the word of his own infidelity. But God has triumphed over his power. 
The elements of this account ring true, but its chronology seems confused. Frederick was not in Sicily in 1249 or the years before it and there is no record whatsoever that he went to Malta at any time; and in December 1250 he died. There were expulsions from Sicily in 1245 and probably in later years as well, and there were references, but in documents dated well after 1250, to expulsions of Muslims from Malta, to the confiscation of their Maltese properties and to Maltese Muslims at Lucera.  Giliberto’s report shows that there was still a considerable number of Muslims on Malta in about 1242, and an expulsion from Malta in 1245, which possibly continued in later years, seems quite possible.  [p.12] Ibn Khaldun was evidently wrong in placing such an expulsion in 1249 and in writing of an imperial visit to Malta made after October 1249, unless it was Frederick’s son Manfred who was involved. Apparently, however, Ibn Khaldun confused Abu Zakaraya with Muhammed Ibn Abbad who led the Muslim revolt in Sicily which ended in 1225. It was presumably then that the “tyrant” Frederick II effected expulsions from Malta, though without going there himself. This was the one single reference Ibn Khaldun made to any Muslim revolt in Sicily. Unsure where to place various stories of such a revolt, but having perhaps some indication that Abu Zakaraya had helped the Sicilian Muslims, Ibn Khaldun may have decided to associate the incident with Abu Zakaraya’s death. The sequence of a Tunisian ruler intervening on behalf of the Sicilian Muslims in and around Palermo, of a revolt following his death, of Christian repression and Muslim resistance, and of the defeat of the Muslims and their exile to Lucera is all acceptable. In Ibn Khaldun’s text this sequence was followed by the mention of the Bani Abs; that is likely to have been a reference to Muhammed Ibn Abbad which probably involved a confusion, explicable in the Arabic, between the forms Abbad and Abs. The mention of the Bani Abs was followed by that of an attack on Malta and of expulsions and conquest there. The sequence of violences in Sicily did continue until 1246 or later, but the story of the Bani Abs and the attack on Malta probably belonged to an earlier period between about 1221 and 1225. 
If such an explanation is broadly correct, then it seems likely, though not certain, that there were expulsions from Malta and transportations there from Celano, perhaps in about 1224, with further expulsions in about 1245. Such expulsions, coming both before and after 1241, would have been consistent with roughly half the population of the two islands being Muslim in that year. The presumed 1250 Christian families on Malta and Gozo in about 1241 could have included Italians from Celano, other Latin immigrants and converted Maltese Muslims in proportions which cannot be established. The Latins transported from Celano could have been intended to replace Muslims who had been expelled, but that too is hypothetical. It seems that down to 1241 the king was prepared, perhaps because there were still insufficient Latins on the islands, to countenance [p.13] a Muslim presence on them in 1241. Giliberto was, therefore, instructed to allow the islanders to live according to their own mores and constituciones as long as that appeared to be in the royal interest. Later, probably in coincidence with the Sicilian oppressions of 1243 to 1246, there were quite possibly renewed expulsions of Muslims from Malta, presumably accompanied by further conversions from Islam.
The governor’s report of about 1241 was concerned with the royal or curial establishment. The records of Giliberto’s predecessor Paolino de Malta showed that the garrison of the three royal castles had numbered 150 servientes, 25 of whom were the sailors of two small vessels; their pay had amounted to 11,000 tareni a year. The grammar is somewhat confused but, unless the report was somehow adding 70 wives to 150 servientes to make 220, the garrison appears to have risen to 220 by Giliberto’s time; and he asked for another 20. The 220 men received a total of 35 salme of wheat per month, and seventy of the garrison had wives who were given a further 12 salme a month. In addition there were 5 granary men, 60 male and female servi and ancille who worked on walls, buildings and other services for the curia, and 4 curatuli or estate managers, who between them received 12 salme monthly; that made a total of 59 salme, to which were added cheese, butter, meat and other foodstuffs.  There were also 84 servi from the island of Gerba off the Tunisian coast who were presumably captives and who worked the curial lands of the demanium; these too must have been fed. The estates farmed directly by the curia consisted of 55 paricole or about 9 percent of the two islands’ total surface; presumably these estates did not include the grazing lands of the curia, or common lands or fields cultivated either by the villani who owed their produce to the curia or by other landholders. The jardena or fields of the curia produced 19 percent of those curial incomes paid in money rather than kind. In addition there were 55 cowherds, 10 shepherds, 2 mule or horse keepers, 3 donkeymen and 12 other male and female serui. These servants, said to number 144 though the figures as given amount to 166, received a monthly total of 50 salme of barley; they received considerably less barley per person than the garrison and others who received wheat. The serui and ancille were presumably slaves.  Many slaves were taken in a raid on Gerba [p.14] in 1223.  Probably all or most of the serui were Muslims; the curia provided them with a food allowance in kind, and they were not listed as paying taxes in money or produce.
|paricole of curia||40||15||2.7||55|
|INCOMES (in tareni)|
|baiulacio et uxorte||1390||600||2.3||1990|
|gisia villanorum curie||2516||584||4.3||3100|
|Total as given||10901||3781||2.9||14681|
|[Total by addition]||||||[3.2]|||
[p.15] Of the incomes listed, the dohana involved general customs dues, probably on both imports and exports; the tintoria, tube, barbaria and apothece taxed dyers, musicians, barbers and shopkeepers; the uxorte derived from petty fines; the gisia was paid by Muslims; centimula, jardena and orti were incomes from mills, gardens and orchards; the madia was the Gozo boat; and corbinorum remains obscure.  There was no mention of grain, of wine  or of cotton which became vitally important;  all three may have paid taxes reported under the dohana. In 1283 the crown confirmed to the men of Malta and Gozo their existing permission to import 300 salme of grain from Licata in Sicily without paying tax. 
The curia did not necessarily manage or benefit from all taxes or gabelles, some of which might have been alienated or farmed out and thus have been omitted from Giliberto’s list. In 1281, for example, there were the decime sclavorum, presumably tenths levied on the sale of slaves and captives.  The various taxes enumerated by Giliberto, most of which were Sicilian in origin, brought the curia nearly 15,00 tareni; the villani sarraceni also paid a fourth of their produce in kind. The garrison of 150 had cost 11,000 tareni just a little earlier when Paolino de Malta was governor, so a garrison of 220 may have cost some 16,000 tareni, while wheat, barley, cheese and so on were provided for men, wives and beasts; the types and numbers of animals were also reported. In the report the quantities provided for men and animals were added up, correctly, to 59 salme of wheat and 84 salme of barley a month. Malta and Gozo clearly had a money economy and they may have produced some wealth in the form of lucrative cotton exports; the report suggests that the islands may roughly have paid for their own defence and administration. A royal letter of 1277 claimed that in earlier reigns, which would have included that of Frederick II, Malta customarily brought to the crown an annual proventus, presumably income [p.16] calculated before expenditures, of more than 30,000 tareni.  If the report of about 1241 was correct, this would seem to have been an exaggeration. 
The ruling group on the islands was composed of an essentially Sicilian administration consisting of a royal governor, who may also have held the military command, the officials of the curia and the garrison; probably there were a few Sicilian judges and notaries, a number of Genoese or other foreign merchants, and some clergy. At some point shortly before 1257 certain Genoese merchants on Malta were in revolt against the crown.  More permanently established were the homines or fideles of the Maltese islands, a number of them possibly converted indigenous Muslims and others presumably Christian immigrants and settlers; in about 1241 it may still have been the case that some of these homines were Muslims, as in 1198. The homines had their own customs and they gave consilium to the governor; presumably they held lands, though there was no indication of that in the report. Some of them had usurped unspecified royal raciones and jura which were not necessarily lands. A royal letter of 1272 claimed that the fideles had long held possessions, while another letter of 1270 showed that various inhabitants of Malta had asserted that Frederick II had despoiled them of their lands on that island.  Very little is known of this group, and much of what has been written about it should be treated with caution. 
[p.17] Possibly there were still a number of free Muslims on Malta in about 1241, but most of the Muslims there were villani who paid the gisia or jizya to the curia. The gisia was originally a Muslim tax imposed on non-Muslims, but in twelfth-century Sicily it was owed by all Muslims, whether villani or not, to their Christian rulers; in 1239 even the Muslims at Lucera in Apulia were paying the gisia.  The Normans also imposed it on the Muslims of Tunisia.  In Sicily the Normans normally made a Muslim-style pact or aman with subject groups who were assured certain freedoms as ahl al-dhimma or “people of the pact” and who were bound, probably by a written treaty, to pay a tributum or censum. Something similar evidently occurred on Malta in 1091 when the Muslims agreed to pay an annual data according to their own Islamic law: more legis suae.  In Sicily such a pact was the basis for the payment of the gisia, which was in part a religious penalty which could be avoided through conversion to Christianity; it set people apart by law as well as by religion. The tax could be assessed as a collective tribute payable by a community which raised it through its elders, but increasingly it became a personal tax collected with the help of a system of tax registers which listed the names of those who were due to pay it. In twelfth-century Sicily there was much diversity in the status of the villani, some of whom also paid part of their produce in kind as a form of the Islamic tax known as the kharadj. The standard villani on the royal estates were bound or “adscripted” to the land and their service was heritable. Other unattached villani, known as muls, commended themselves to a lord or to the church and owed gisia by virtue of the position or tenure they thus acquired. Another class which paid gisia was the rijal al-mahallat who were possibly Muslim burgenses from the villages. 
After 1091 Malta may have paid its data as a collective sum owed by all or most of the inhabitants. Then, as the island was Christianized, there would increasingly have been Christians who did not owe the gisia with the result that those who did, the non-Christians, paid as individual families of villani curie who were listed, or at least counted, according to Sicilian practice, as in Giliberto’s [p.18] report. The fourth part of their produce which Giliberto reported that the Muslim villani of Malta and Gozo owed to the curia evidently corresponded to the kharadj.  Giliberto gave no figure for this income from produce, probably because it was paid in kind and consumed on the islands. There was again no indication that this quarta on produce was paid by non-Muslims or by any Muslims who were not villani curie. There seems, however, to have been some doubt in Sicily whether even notable Muslims were fully free. 
Given the ferocity of Frederick II’s persecution of the Sicilian Muslims, it is possible that some of them fled to Malta. Conceivably there were also in Malta a few Muslims who had become Greek rather than Latin Christians, as did occur in Sicily.  The words ego and martur, for “I” and “witness”, were written in Greek for a witness, who did not actually sign his name, at the bottom of a Maltese document of 1274, and in 1299 arnaton pousale pete witnessed an act in Gozo by writing his name in Greek characters.  There is little other explicit evidence for Sicilian or Maltese Greeks or for Greek-speaking or Greek rite converts from Islam on Malta at that time, presumably because all or most of the Greek Christians on Malta had, unlike those of Sicily, been eliminated long before the thirteenth century. Though Frederick II continued the Norman tradition [p.19] of favouring the Greeks in Sicily, Basilian monasticism declined after about 1270. The number of Greek rite Sicilians, many of them not connected to the Basilian monks, was apparently limited, while conversions from Islam were probably not numerous and may often have involved Greeks from families which had earlier converted to Islam and had retained their Greek speech; such Muslims of Greek origin were presumably not to be found on Malta.  That folk memories and scraps of vocabulary may nevertheless have preserved an awareness of a Greek past and a feeling that the islanders were neither Sicilians nor Africans is suggested by an extraordinary jumble of ancient mythology, Sicilian toponomy and garbled Byzantine history recounted by Bartolomeo de Neocastro of Messina.  According to this story, presented as an account given by an old Muslim on Gerba who was asked in 1284 to explain the origins of the inhabitants of Gerba, Kerkena, Pantelleria, Lipari and Malta, they were the children of the black Egyptian god Amman and a group of Byzantine women from Sicily who had been exiled in 1043 for their support of the rebellious Byzantine general Georgios Maniaces.  The islanders were said to have been their offspring who subsequently adopted Arabic speech and Muslim religion but retained some memory of their mothers’ Greek tongue: quaedam a graeco sermone memoriter teneamus. Some Sicilian Muslims converted to the Greek, rather than the Latin, rite probably because they already knew some Greek and because Sicilian Greeks were familiar to them while the Latins were unknown conquerors and invaders. Maltese Muslims, having no Greeks to whom they could assimilate, were presumably converted to the Latin religion but they would have found the Italian language unfamiliar and so they retained their own Arabic speech. 
There were a few Jewish families on Malta and Gozo but their status is uncertain. In Sicily Jews paid the gisia from Muslim times until the fifteenth century, and Sicilian Jews were technically servi or slaves of the royal curia, [p.20] though that was largely a point of jurisdictional significance.  In 1211 Frederick II granted the Palermitan Jews and their taxes, including the gisia they paid, to the Archbishop of Palermo, while in 1239 other Jews in Palermo who were known as de Garbo were paying the gisia to the crown.  Giliberto’s report listed only 33 Jewish families in Malta and Gozo without making it clear to whom they belonged or whether or not they paid the gisia; probably they did belong to the curia and did pay the gisia in about 1241 just as they did in the fifteenth century. 
The rest of the analysis of Giliberto’s report largely depends upon the interpretation of two ambiguous passages. The first, concerning Malta, reads in the surviving version: Gisie quingentorum quadraginta villanorum curie tarenos duomilia quingentos sedecim. The second, concerning Gozo, reads: Gisie tarenos centum quadraginta quinque. Villanorum curie tarenos quingentos octoginta quatuor. This is most likely to mean that on the island of Malta there were 540 villani curie who paid the gisia tax of 2516 tareni, while in Gozo 145 villani curie paid the gisia tax of 584 tareni.  Neither 2516 nor 584 tareni seems like a round figure fixed as a collective tribute, while both are divisible by four, which might suggest 629 and 146 households respectively paying four tareni each; on Gozo there were in fact said to be 145 villani curie. This would mean that in Malta each gisia-payer paid approximately 4.7 tareni while in Gozo each paid approximately 4 tareni; and that in Malta there were 540 Muslim and Jewish villani curie to 706 Muslim and Jewish families, and in Gozo 145 Muslim and Jewish villani curie to 163 Muslim and Jewish families, assuming that the Jews were villani curie and paid the gisia. In this case the ratio of tareni paid by villani curie in Malta to tareni paid by villani curie in Gozo, that is 2516 to 584 or 4.3, is precisely the same as the ratio of Muslim and Jewish families in Malta to that of Muslim and Jewish families in Gozo, that is 706 to 163 or 4.3; the result is very similar if the Jewish families are ignored. This again suggests that the figures for Muslim families are correct. Most of the Muslims counted in the report would, therefore, have been villani curie paying the gisia and also paying taxes in kind, since the report stated that the villani sarraceni curie paid an [p.21] annual fourth part of their produce; the use of the term villani sarraceni might mean that Jews, probably town-dwellers, were villani curie but did not owe this annual fourth. This interpretation would involve a discrepancy - ignoring the Jews - in the region of 150 for the two islands between the number of Muslim families, given as 836, and the number of gisia-paying villani, which would have been 685. Conceivably there were some free Muslims, perhaps in the towns; or there were villani on those lands of the curia which Giliberto mentioned as having been usurped; or, as was common in Sicily, there were villani who did not belong to the curia but to ecclesiastical or other lords; or some Muslim slaves may have been counted as families. There may also have been widows, the blind, cripples or fugitives who did not pay the gisia, and that may have produced apparent statistical discrepancies. Maltese customs probably differed from those of Sicily and Sicilian practices varied greatly. As a general rule, all Sicilian paid the gisia, whether they were villani or not; if they were villani curie they could hold land and would pay the kharadj on it. Maybe there were Muslims on Malta who were villani sarraceni curie and who paid gisia and kharadj in cash and who also owed a quarta or kharadj in kind which was not included in Giliberto’s accounts. All this would, at least, have been consistent with these two passages simply meaning that given numbers of villani curie paid given sums in cash as gisia.
This is in many ways the most satisfactory solution and it may well be the correct one. It would match the general ratios of inhabitants and payments in Malta to those in Gozo, as shown in the table above. The financial totals in the report are probably almost correct, since 10,901 plus 3781 amounts to 14,682 instead of 14,681 as given, a discrepancy of only one tarenus. The total for Malta is 10,901 while the figures reported actually add up to 10,917, again a small difference of only 16 tareni. For Gozo, however, one or more items have apparently been omitted or miscopied because the individual taxes given amount to 3384 tareni which is 397 less than the total recorded in the document. It also has to be assumed that for Gozo the word tarenos after Gisie and the full-stop, or more precisely two points, in the manuscript after quinque were inserted incorrectly; since the text survives only in a copy, this is quite possible. Furthermore, it supposes that most Muslims paid the gisia and were villani curie; that the considerable number of Christian families paid no direct tax, as is likely since no such tax was recorded in Giliberto’s report; and that there were no Christian villani curie, which is also quite possible since there had been so few Christians on Malta in the twelfth century, since there were few, or almost no, Christian Greeks on Malta in the thirteenth century, and since Latin villani were by then almost unknown in Sicily.
[p.22] The two passages might, however, be interpreted quite differently as meaning that in Malta the gisia produced 540 tareni and the villani curie paid 2516 tareni while in Gozo the gisia produced 145 tareni and the villani curie paid 584. In this case the various ratios and proportions between the two islands would be little changed, but it would have to be assumed that in the surviving version the word quingentorum was wrongly placed in the genitive and the word tarenos was dropped out after the Gisie for Malta; and whereas the totals in tareni reported in the surviving text are Malta 10,901, Gozo 3781 and the two together 14,681, the figures given would on this interpretation amount to 11,457 and 3529, making an overall total of 14,986. These discrepancies make this interpretation seem an unlikely one. It might be suggested that there were villani curie who, because they were Christians, paid a tax which was not the gisia; those might have been indigenous Muslims converted to Christianity. It is true that it was generally the case in Sicily that a Muslim villanus curie who converted to Christianity remained a villanus curie but paid the gisia at a lower rate or escaped it entirely; however, there is no evidence for the existence on Malta of a class of Christian villani, and the statistical discrepancies resulting from such an interpretation would remain without explanation. None the less, that there was eventually a considerable group of converts seems almost certain in view of the survival of the indigenous Arabic language in the islands, but nothing definite is known about this process of conversion. The apparent absence on Malta and Gozo of a class of Latin or Greek villani which might have absorbed converts from Islam presumably helped to ensure the survival of an Arabic speech.
Preliminary studies of land tenure and related matters show that, unlike Sicily, there was a considerable extension of peasant land ownership, of leaseholds and paid labourers in Malta during the fifteenth and the sixteenth centuries,  but that proves nothing conclusive in view of the profound changes which occurred on Malta between the thirteenth and fifteenth century.  As in Sicily, many free Christians on Malta and Gozo would have probably been Italian rather than Sicilian in origin, whether or not they came from Celano, and they would conceivably have held lands, following the Sicilian pattern, by non-feudal tenure, owing military service but not direct taxes.  Maltese society in about 1241 apparently included a group or groups of free Christians who paid no direct tax to the curia, an advantage which possibly attracted Muslims on Malta to convert [p.23] to Christianity; a class of Muslim villani belonging to the curia who paid the gisia and contributions in kind; and a few Jews. Hypotheses about these and other possible groups must largely be based on known situations outside the Maltese islands.  Particularly relevant are considerations concerning villani in Sicily. Unfortunately, variations in the status of the Sicilian villani were so considerable that it is impossible to be sure which of its elements might have been transferred to Malta. Christian villani were mostly indigenous Greeks or converted Muslims. Some Muslim villani had been landless, almost slaves, while others were relatively free and held lands which they farmed; some were villani by virtue of their tenure, others by personal status; some villani owed produce as well as money. By 1241 the Muslim villani of Sicily were disappearing as they converted, fled, or were deported or otherwise eliminated during the ruthless repressions of Frederick II. and by the second half of the thirteenth century Sicilian ecclesiastical lands had to be let out in emphyteusis as there were so few villani left. 
In the absence of further documentary evidence, there is no way of knowing how these Sicilian practices and developments were assimilated to the customs and situations of Malta. There had been a group of Christians who were foreign slaves or captives on eleventh-century Malta but Count Roger is said to have removed them all in 1091. In the twelfth century Malta was largely Muslim and presumably there were few, if any, indigenous Christians who could have become non-Muslim villani. The process of Latinization must have been furthered by the transportation of rebels from Celano in 1224. The Christian families listed on Malta and Gozo in 1241 circa were Latins, or converted Muslims or conceivably Greek immigrants from Sicily or Italy, or they were converted Muslim from Malta and Gozo. Indigenous Maltese and Gozitan Muslims, whether of town or country, may have gone into exile and some were expelled, though such expulsions may have taken place only after the time of Giliberto’s report. On small flat islands it would not have been easy to flee from the curial estates and the temptation to convert would have been strong; many must have done so. The curia still possessed serui, possibly converted Muslims, in 1271, 1273 and 1281.  Seruus did not always means slave, and the serui who worked the royal [p.24] estates on Malta may well have been “serfs” or villani of the curia. A seruus curie was documented on Malta in 1372 and another on Gozo in 1373/4.  However, the ordinary tax-paying villani may have disappeared, or largely disappeared, in the late-thirteenth century. A text of 1271 referred to Muslims who had been expelled from the islands - it is not clear when - and whose possessions had passed to the royal curia, and in the same year the king ordered that uncultivated curial lands on Malta and Gozo were to be leased at an annual census to agricultores who were presumably not villani. 
Giliberto’s report was a technical and statistical document concerned above all with taxation, and it classified people into categories which were probably not so clearly distinguished in real life as in the bureaucratic text in the form in which it survives. In Sicily, where there were Muslim revolts which received external support and where the complete Christianization and Latinization of the Muslim population proved impossible, Frederick II resorted to brutal repressions and expulsions. Something similar occurred on Malta, but there a section of the Muslims, unable to retreat to the mountains as did the Sicilian Muslims, preserved their language and to some extent their economic and social position by accepting conversion; in a sense they resisted Latinization by accepting Christianization. There must have been tensions between converted Muslims and Christian immigrants. Thirteenth-century Malta and Gozo evidently constituted a society in a state of political, religious, economic and social upheaval in which many men remained unsure of their status as the indigenous inhabitants changed their religion and condition or left the country, as Latin immigrants established themselves on the islands, and as the officials of the royal curia sought to squeeze the maximum profit from the land and people.
[p.25] Fridericus etc, ad Gilibertum Abbatem87 etc, Litteras et capitula que nostro culmini destinasti etc quod autem in ipsis capitulis vidimus contineri quod jura dohane Malte terre et maris tam vendencium quam emencium88 ualent tarenos auri89 mille annuatim et nongentos. Tintorie90 tarenos octingentos. Tabule tubarum91 tarenos octingentos. Barbarie92 tarenos trecentos quadraginta. Cabelle Baiulacionis93 et Sorte94 tarenos mille trecentos nonaginta. Gisie95 quingentorum quadraginta villanorum curie tarenos duomilia quingentos sedecim. Apothecarum96 curie nostre tarenos centum. Cabella Corbinorum97 tarenos centum [p.26] quinquaginta. Cabelle decem jardenorum98 curie nostre, tarenos duo milia septingentos quadraginta et unus. Cabella trium ortorum99 centum octoginta. Quorum omnium prouentuum jura sunt in summa, tarenos decem milia nongentos et unus.100
Significasti etiam quod jura dohane maris et terre Gaudisij sunt annuatim tarenos quingentos. Tyntorie tarenos ducentos sexaginta. Barbarie tarenos octoginta. Cabella tubarum tarenos centum septuaginta. Cabelle madie101 tarenos trecentos. Cabelle Baiulacionis et uxorte102 tarenos sexcentos. Gisie tarenos centum quadraginta quinque. Villanorum curie tarenos quingentos octoginta quatuor. Centimulorum103 curie nostre tarenos octingentos nonaginta. Quorum omnium tarenorum summa est triamilia septingentos octoginta unus.104
Et summa omnium prouentuum Insularum Malte et Gaudisij est tarenos quatuordecim milia sexcenti octoginta unus.105
Et continebant etiam ipsa capitula quod statute sunt in agricultura curie nostre Malte paricole quadraginta, et in Gaudisio paricole quindecim.106
Et quod villanj sarracenj curie nostre Insularum ipsarum soluunt annuatim curie nostre de omnibus victualibus eorum quartam partem.
[p.27]Et quod inuenisti per quaternitiones quondam Paulinj de Malta107 quod dabantur annuatim pro solidis et uictu108 centum quinquaginta seruientium, qui statuti sunt in custodia trium castrorum nostrorum insularum ipsarum, tarenos milia, inter quos seruientes computati sunt marenarii vigintiquinque, qui deputati sunt etiam ut cum necesse fuerit armentur ex eis vackette109 due pro seruicijs nostris.
Significasti insuper, in Insula Malte esse familias christianorum110 quadraginta septem. familie Sarracenorum sexcentas octoginta unam. Iudeorum vigintj quinque.
Et in Gaudisio scripsisti esse familias christianorum ducentas tres. Sarracenorum centum quinquaginta quinque et familias Iudeorum viij.111
Et eadem capitula, continebant quod ducenti viginti seruientibus statutis in custodia castrorum nostrorum Insularum ipsarum dantur pro uidenda mense quolibet de frumento curie nostre salme112 triginta quinque et septuaginta mulieres uxores, videlicet septuaginta seruientuum predictorum recipiunt pro uidanda de frumento quolibet mense salmas duodecim et quinque granecterijis et sexaginta seruis et ancillis113 curie nostre deputatis ibidem ad marramata114 curie nostre et alia seruicia nostra in castris ipsis.
Et quatuor curatulj agrorum nostrorum, dantur similiter pro uidanda de frumento curie mense quolibet salme duodecim.
[p.28]Cuius predicti frumentj summam scripsisti esse pro uidanda omnium personarum predictarum mense quolibet salme quinquaginta nouem.
Et quod significasti dari eis quolibet mense de caseo cantaros115 nouem, de butiro, carnibus et alijs pro companagio116 ad sufficienciam eorundem et quod octoginta quatuor serui Gerbinj117 deputati in seruicijs paricolarum curie nostre quinquaginta quinque custodes boum et vaccarum decem pastores ouium et caprarum curie nostre, duo custodes jumentorum tres custodes asinorum duodecim seruj et ancille deputati in seruicijs curie nostre.
Que omnes persone sunt centum quadraginta quatuor,118 recipiunt mense quolibet pro uidanda eorum de ordeo salmas quinquaginta.
Et quod dantur quinque mulis qui sunt in tribus centimulis castri maris Malte,119 asinis quatuor et mulis duobus statutis in furno ipsius castri. duobus stallonis equinis et duobos stallonis asinis pro annona eorum singulis mensibus de ordeo salme tredecim. Et quod in castro ciuitatis Malte120 sunt somerii duo et mulus unus in centimulo,121 quibus datur de mense in mensem pro annona eorum de ordeo salma una.
Et in centimulo castri Gaudisij sunt asinj tres, roncinus unus et asinus unus stallonus.
Et in centimulis quatuor de cabelle muli sex, quibus datur per mensem pro annona ipsorum de ordeo salme tredicim.
Et sunt ibi asini triginta duo deferentes victualia curie nostre quibus statutj sunt per mensem de ordeo salme septem.
[p.29]Et quod tocius ordei quod datur pro nostris seruicijs supradictis. Summa est quolibet mense, salme octoginta quatuor.
Continebatur etiam in transmissis capitulis tuis te necessario statuisse in predictis tribus castris nostris121 de consilio fidelium nostrorum Insularum ipsarum alios seruientes vigintj, quod placet nobis dum modo sint necessarij.
Continebant insuper ipsa capitula quod homines Insularum ipsarum uiuunt122 alijs moribus et constitucionibus quam alij homines Regni nostri Sicilie, et consultasti Excellenciam nostram quid inde placeat culminj nostro, ad quod tibi taliter respondemus quod si uideris redundare ad maius123 comodum curie eos uiuere, iuxta tenorem sacrarum constitucionum nostrarum124 et secundum consuetudines quas habent alij feudales125 nostri Sicilie, eos facias uiuere, secundum quod alij Siculj nostri uiuunt, quod si maiores126 et constituciones eorum redundant ad maius127 comodum curie nostre eos permictas, secundum soliti sunt uiuere hactenus in utroque tamen considera modum128 curie nostre.
Consultasti etiam sunt129 tibi delatum est, per fideles nostros multi de insula130 ipsis detinent raciones et jura spectantia ad demanium nostrum et sine mandato nostro contra eos procedere noluisti. Cum igitur unicuique nostrorum fidelium sua jura illibata seruemus, pati minime uolumus nec debemus nostra ab alijs occupari. Quare uolumus ut inquisitis omnibus plene et cognitis racionibus et juribus nostris, a quibuscumque ipsa inueneris contra justitiam detineri, ad manus nostre curie sicut justum fuerit studeas reuocare.
 E. Winkelmann, Acta Imperii Inedita Saeculi XIII., I (Innsbruck, 1880), 713-715 (no. 938)
 A. Mifsud, “Frammenti Storici”, La Diocesi [Malta], i (1916), 17-19.
 C. Dessoulavy, “Malta in the Middle Ages”, Journal of the Malta University Literary Society, ii nos. 10-12 (1937).
 W. Hagemann, “La nuova Edizione del Registro di Federico II”, in Atti del Convegno Internazionale di Studi Federiciani (Palermo, 1952), 330-331.
 E. Sthamer, “Studien über die sizilischen Register Friedrichs II”, Sitzungsberichte der Preussischen Akademie der Wissenschaften, xxxii (1920); see also J. Mazzoleni, La Registrazione dei Documenti delle Cancellerie meridionali dall’Epoca sveva all’Epoca viceregnale, i (Naples, 1971), 20-23, and tavole x-xii, and T. Kölzer, “Die sizilische Kanzlei von Kaiserin Konstanza bis König Manfred: 1195-1266,” Deutsches Archiv, xl (1984).
 Winkelmann (1880), i. 601, 715.
 A. Luttrell, “Frederick II and Paolino de Malta: 1235,” Quellen und Forschungen aus italienischen Archiven und Bibliotheken, lv-lvi (1976).
 R. Valentini, “Funzione storica e Dispersione del Patrimonio demaniale di Malta”, Archivio Storico di Malta, xiii (1942), 3-6.
 A. Luttrell, “Approaches to Medieval Malta”, in Medieval Malta: Studies on Malta before the Knights, ed. A. Luttrell (London, 1975), 36-40.
 H. Bresc, “The Secrezia and the Royal Patrimony in Malta: 1240-1450”, ibid., 129-131, 134.
 A. Pertusi, “Le Isole rnaltesi dall’Epoca bizantina al Periodo normanno e svevo (secc. VI-XIII) e Descrizioni di esse dal sec. XII al sec. XVI”, Byzantinische Forschungen, v (1977), 282-286.
 I. Peri, Uomini, Città e Campagne in Sicilia dall’XI al XIII secolo (Rome-Bari, 1978), 154-155.
 G. Wettinger, “The Arabs in Malta,” in Malta: Studies of its Heritage and History, ed. Mid-Med Bank (Malta, 1986), 98-99. On the historiographical background, S. Candido, “Contributi alla Storia di Malta dall’Età Araba in un Carteggio inedito di Michele Amari con Francesco Crispi: 1853-1854”, Atti della Accademia Nazionale dei Lincei: Memorie - Classe di Scienze morali, storiche e filologiche, ser. VIII, xxxi fasc. 3 (1988).
 Hans Martin Schaller of Munich most kindly checked the 1239/40 register on this point.
 L. Sciascia, “I Cammelli e le Rose: gli Abbate di Trapani da Federico II a Martino il Vecchio”, in Mediterraneo Medievale: Scritti in onore di Francesco Giunta, iii (Soveria Mannelli, 1989), 1174-1181.
 Ibid.., 1182-1185.
 Luttrell (1975), 43; but Sciascia, 1185-1188, shows that Palmerio’s attitude was ambiguous.
 H. Bresc, “Pantelleria entre l’Ilslam et la Chrétienté,” Cahies de Tunisie, xix (1971 ), 117.
 The discussion in Luttrell (1975), 36-40, is far from exhaustive and requires amendment in the light of the new reading of the word gisie and of its implications.
 J. Johns, “Entella nelle Fonti arabe: il Registro dei Confini (Jarida) di S. Maria di Monreale, 1182 D.C.,” in Alla Ricerca di Entella, ed. G. Nenci (Pisa, 1993).
 A. Luttrell, “The Christianization of Malta,” The Malta Year Book 1977 (Malta, 1977); idem, “Ibn Hauqal and Tenth-Century Malta.” Hyphen [Malta], v no.4 (1987); idem, “Slaves and Captives on Malta: 1053/4 and 1091,” ibid., vii no. 2 (1992); idem, “L’Effritement de l’Islam: 1091-1282”, in Le Carrefour Maltais = Revue du Monde Musulman et de la Méditerrannée. Ixxi (1994).
 G. Schroth-Köhler - T. Kölzer - H. Zielinski, “Zwei staufische Diplome für Malta aus den Jahren 1198 und 1212,” Deutsches Archiv, xxxiii (1977), 519-520; this text of 1198 is also in T. Kölzer, Constantiae Imperatricis et Reginae Siciliae Diplomata: 1195-1198 (Cologne, 1983), 238-240. The Latin text of 1198, now lost, was followed by nine lines in Arabic. S. Fiorini, “Malta in 1530,” in Hospitaller Malta 1530-1798: Studies on Early Modern Malta and the Order of St. John of Jerusalem, ed. V. Mallia-Milanes (Malta, 1993), 112-113, emphasizes the presence of a few exiles, at least one of them Greek, on twelfth-century Malta.
 Bresc (1971), 105-111; idem, Un Monde méditerranéen: Économie et Société en Sicile, 1350-1450, 2 vols. (Palermo-Rorne, 1986), ii. 622-624.
 Text in Schroth-Köhler et al., 519-520.
 J. Cassar Pullicino, “Norman Legends in Malta,” in Medieval Malta, 99.
 Text in Schroth-Köhler et al., 519-520.
 Text ibid., 521.
 Marchisio Scriba, Annali Genovesi di Caffaro e de’ suoi Continuatori, ed. L. Belgrano, ii (Rome, 1901), 192-193.
 In addition to Luttrell (1975), 34-40, see D. Abulafia, “Count Henry of Malta and his Mediterranean Activities: 1203-1230,” in Medieval Malta; E. Mazzarese Fardella, I Feudi comitali di Sicilia dai Normanni agli Aragonesi (Milan, 1974); A. Luttrell, “Malta e Gozo: 1222-1268,” in Jaime I y su Época = X Congresso de Historia de la Corona de Aragón: Comunicaciones 1 y 2 (Zaragoza, 1980).
 Text in V. Laurenza, “Malta nei Documenti angioini del R. Archivio di Napoli,” Archivio Storico di Malta, v (1934), 7.
 Text in Schroth-Köhler et al., 519-520.
 Text in Laurenza, 8-9.
 Pertusi, 282-284; Peri (1978), 154, simply gives 407 without discussion.
 Stanley Fiorini has kindly calculated that if those figures from Giliberto’s text which are applicable to the whole of both Malta and Gozo - that is for the dohana, tintoria, barbaria and baiulacio, but excluding the gisia not paid by Christians - are utilized for a regression analysis, the Malta figures regressed to the Gozo data yield a gradient of 2.42 for the least-squares line; with 95 per cent confidence, the error in this gradient is plus or minus 1.48. With that level of confidence and on the assumption that the number of families on Gozo was 366, the total Malta population lies between 334 and 1428 families. It would follow that the Maltese Christian population was at most 722 (ie. 1428 minus 706), making the 1047 estimate rather unlikely and the 747 guess more acceptable. If a lower degree of confidence were accepted, then the 1047 hypothesis would become less unlikely.
 As already proposed, but without the supporting arguments used here, in K. Beloch, Bevolkerungsgeschichte Italiens, i (Berlin, 1937), 164-165.
 Peri (1978), 108-110, 246-247.
 Bresc (1986), i. 60.
 G. Wettinger, “The Militia List of 1419/20,”” Melita Historica, v. no.2 (1969), 83. Figures from 1419 to the 1480s produce estimates varying between 8098 and 10,633: Fiorini, l22, l26. It is not known whether Muslim and Christian families were of the same average size.
 F. Maurici, L’Emirato sulle Montagne: Note per una Storia della Resistenza musulmana in Sicilia nell’età di Federico II di Svevia (Palermo, 1987), 38-49.
 E. Winkelmann, Kaiser Friedrich II., i (Leipzig, l889), 537-538; M. Amari, Storia dei Musulmani di Sicilia, ed. C. Nallino, iii (revised ed: Catania, 1937/8), 613-614.
 Riccardi de Sancto Germano notarii Chronica , ed. C. Garufi, Rerum Italicarum Scriptores, ns. vii part 2 (Bologna, 1938), pp. xix-xxvi, xxxiii.
 Ibid ., 112-113, 146. Peri (1978), 127, 154, writes of a Sicilian expedition which attacked Malta in 1224; it resulted, he claims, from Frederick II’s desire to exterminate the Muslims on Malta, who were replaced with Latins from Celano. Peri is following Amari (1937/8), iii. 617, who deduces an expulsion of Muslims from Malta both from a passage in Ibn Khaldun and from the evidence for the immigration of Latins from Celano into Malta.
 Abulafia, 121-122.
 Text in Registres d’Innocent IV, ed. E. Berger, iii (Paris, 1897), 50; kindly communicated by Godfrey Wettinger.
 Annales Siculi , ed. E. Pontieri, in Rerum Italicarum Scriptores, ns. V. part I (Bologna, 1927), 119.
 Text in M. Amari, Biblioteca arabo-sicula: versione italiana, ii (Turin-Rome, 1881), 212-213.
 References in Luttrell (1975), 37; Wettinger (1986), 98-99, 103-104.
 There is, however, no proof that any expulsions accompanied the obscure resistance on Malta in or before 1257 of certain Genoese who apparently had connections with the converted Sicilian Muslim Oberto Fallamonaca: Luttrell (1980), 592-597. A 1492 copy of Manfred’s privilege of 1259 suggests that he was then on Malta: ibid., 601. However, a 1284 copy gives felfis (Melfi) not Malte: Palermo, Archivio di Stato, Cancelleria 2, f. 53-55v (kindly communicated by Stanley Fiorini).
 Ibn Khaldun’s passage was kindly analysed in detail by Jeremy Johns. Amari correctly understood Ibn Khaldun’s anachronism, but explained it in a way which Nallino showed to be incorrect: Amari (1937/8), iii, 609-610, 617-620. Various errors in this respect in A. Luttrell, “The Roots of Medieval Gozo,” Al-Masaq, iv (1991), 55, are here corrected. Ibn Abbad probably died between 1222 and 1224: Maurici, 39-45; Johns (1993), 91. However Al-Himyari placed his de ath in 1219/1220: text and translation of excerpt, manuscripts and sources in ibid., 83-88.
 If these figures were reliable and complete, they might indicate the diet of those involved; such calculations are not attempted here.
 Cf. Amari (1937/8), iii, 240-244; I. Peri, Il Villanaggio in Sicilia (Palermo, 1965), 14/15 n. 9, 21 n.6, 77 n.63, 101 n.72.
 Amari (1937/8), iii. 617-618.
 Details, infra.
 Fiorini, 130, incorrectly implies that Giliberto’s report explicitly mentioned an assisa vini on wine and a conciara tax on tanning.
 G. Wettinger, “Agriculture in Malta in the Late Middle Ages”, Proceedings of History Week 1981, ed. M. Buhagiar (Malta, 1982), 15-21; L. Galoppini, “Notizie su Maltesi e il Cotone di Malta a Cagliari nella seconda metà del Trecento,” Melita Historica, x no.1(1988).
 Text in G. Silvestri, De Rebus Regni Siciliae: 9 settembre 1282 - 26 agosto 1283 (Palermo, 1882), 316-317.
 Texts in Laurenza, 51-56.
 Text ibid., 42-43.
 Another form of taxation, the colletta, involved variable sums imposed on all communities in the Sicilian kingdom, including those which did not belong to the roya1 curia. In 1277 Malta and Gozo were assessed to pay a colletta of 350 uncie or 10,500 tareni; Bresc (1986), i. 60-65. That figure may have been based on an earlier assessment and, at a debatable rate of 6 fuochi or hearths per uncia, it would imply 1750 fuochi for the two islands in 1277. That would mean a reduction in the number of families from the total of 2119 reported by Giliberto (assuming that 47 should be amended to 1047), a change which might be explained in part by the expulsion of Muslims after 1241. Note, however, that there never was a census and that the tax lists were affected by all sorts of exceptions and exemptions, while the number of hearths were arbitrarily changed to reflect local wea1th and ability to pay; W. Percy, “The Earliest Revolution against the ‘Modern State': Direct Taxation in Medieval Sicily and the Vespers”, Italian Quarterly, xxii (1981). Furthermore, S. Epstein, An Island for Itself: Economic Development and Social Change in Late Medieval Sicily (Cambridge, 1992), 40, argues for a rate of 3 tareni per fuocho, which would give 3500 families for Malta and Gozo. These calculations all seem inconclusive.
 Luttrell (1980), 591-597.
 Texts in Laurenza, 9-10, 19.
 Discussion in Luttrell (1980).
 Amari (1937/8), iii. 255-264.
 References in H. Idris, La Berbérie Orientale sous les Zirides: Xe-Xlle siècles, i (Paris, 1959), 360. Fiorini, 130, incorrectly implies that the gisia was “only a tax on the Jews”.
 Gaufredus Malaterra, De Rebus Gestis Rogerii, ed. E. Pontieri, in Rerum Italicarum Scriptores, ns. v part 1 (Bologna, 1927), 95.
 J. Johns, The Muslims of Norman Sicily: c. 1000-c. 1194 (unpublished D. Phil. thesis: Oxford, 1983); idem, “Duana Regis”: Arabic Administration and Norman Kingship in Sicily (in press).
 Peri (1965), 14 and n. 9, and C. Cahen, “Kharadj.” Encyclopaedia of Islam, iv (revised ed: Leiden, 1978), 1030-1034.
 Peri (1965), 14/15 n. 9. On Muslim lords and landholders in Norman and post-Norman Siciliy, see Amari (1937/8), iii. 267-276; V. von Falkenhausen, “I Ceti dirigenti prenormanni al Tempo della Costituzione degli Stati normanni dell’Italia Meridionale e in Sicilia,” in Forma di Potere e Struttura sociale in Italia nel Medioevo, ed. G. Rossetti (Bologna, 1977), 347-351; H. Bresc, “Féodalité coloniale en Terre d’lslam: La Sicile (1070-1240).” in Structures féodales et Féodalisme dans l’Occident Méditerranéen (Xe-XIIIe siècles); Bilan et Perspectives de Recherches (Rome, 1980).
 S. Caruso, “Una Pergamena greca del tempo di Frederico II: Nota su una Famiglia arabo-greca di Messina,” Byzantion, xlvi (1976).
 A. Luttrell, “Medieval Malta: the Non-Written and the Written Evidence,” in Malta: a Case Study in International Cross-Currents, ed. S. Fiorini-V. Mallia-Milanes (Malta, 1991), 37-38. The conversion process and its chronology require further study. There is no clear proof for the existence of a Greek rite or of any Basilian monastery on Malta; the archaeological, topographical and toponymic evidence will be presented in Mario Buhagiar’s doctora1 thesis. On Muslim conversions to the Greek rite in Sicily, see Bresc (1986), ii. 587-594, 607, but his reference to Malta as “a point de resistance de la culture greque” rests on slender evidence; see also the articles in idem, Politique et Société en Sicile: XIIe-XVe siècles (London, 1991). Note that in 1336 circa a German traveller spoke of Christians in Sicily who were ad ritum Sarracenorum: Ludolfus de Suchem, De Itinere Terrae Sanctae Libe1, ed. F. Deycks (Stuttgart, 1851), 20.
 M. Scaduto, Il Monachismo basiliano nella Sicilia medievale -Rinascita e Decadenza: sec. XI- XIV (Rome, 1947), 228-232, 240-242; V. von Fa1kenhausen, “Il Monachesimo greco in Sicilia”, in La Sicilia rupestre nel Contesto delle Civiltà mediterranee, ed. C. Fonseca (Ga1atina, 1986); J. Tohns, “The Greek Church and the Conversion of Muslims in Norman Sicily?,” Byzantinische Forschungen, xxi (1995).
 Bartolomeo de Neocastro, Historia Sicula, ed. G. Pa1adino, in Rerum Italicarum Scriptores, ns. xiii part 3 (Bologna, 1921), 63-65.
 Cf. V. von Fa1kenhausen, La Dominazione bizantina nell’Italia Meridionale dal IX all’XI secolo (Bari, 1978), 95-96.
 “Arab” resistance is discussed in Bresc (1986), ii. 624-628.
 Peri (1978), 87, 303 n. 3; Q. Senigaglia, “La Condizione giuridica degli Ebrei in Sicilia,” Rivista Italiana per le Scienze Giuridiche, xli (1906), 77, 79-81, 87-88.
 Texts in B. and G. Lagumina, Codice Diplomatico dei Giudei di Sicilia, i (Palermo, 1884), 12- 15, 19-20.
 G. Wettinger, The Jews of Malta in the Late Middle Ages (Malta, 1985), 17.
 As assumed in Peri (1978), 154.
 Wettinger (1982), 4-15, 43-48, and idem, “The Village of Ħal Millieri: 1419-1530”, in Ħal Millieri: a Maltese Casale, its Churches and Paintings, ed. A. Luttrell (Malta, 1976).
 Cf. Bresc (1975), 127-136.
 Peri (1965), 45, 69-74; (1978), 96-99.
 Eg. H.E. Mayer, “Latins, Muslims and Greeks in the Latin Kingdom of Jerusalem”, History, lxiii (1978).
 In addition to the works of Peri and Johns, see H. Bercher-A. Courteau-J. Mouton, “Une Abbaye latine dans la Société musulmane: Monreale au XIIe siècle,” Annales: Économies, Sociétés, Civilisations, xxxiv (1979).
 Texts in Laurenza, 17-18, 27-29, 51-52.
 Bresc (1975), 130-131; Wettinger (1982), 11 n. 52.
 Texts in Laurenza, 13-15, 18-19.
 Giliberto Abbate.
 Dohana: customs dues on buying and selling by land and sea.
 Sicilian gold tarì varied considerably in weight, but the tareni of this text were evidently a standard unit of value; P. Grierson - W. Oddy, “Le Titre du Tari sicilien du milieu du XIe siècle à 1278”, Revue Numismatique, VI ser., xvi (1974). The tarì may, in fact, have been used as a money of account.
 A tax on cloth dyeing: Acta Curie Felicis Urbis Panormi, ed. F. Pollaci Nuccio - D. Gnoffo, i (reprint: Palermo, 1982), pp. cxi-cxiv, 338.
 For Tabule read cabelle; cabella or gabella meant tax. At Palermo the gabella tubarum was a tax on musicians: ibid., i, pp. xcv-xcvii, 323.
 Barbaria: a tax on barbers. It survived in Malta in 1506, but was not known by that name elsewhere: J. Luca de Barberiis, Liber de Secretiis, ed. E. Mazzarese Fardella (Milan, 1966), 79 n.77.
 Bresc (1975), 134, gives the baiulacio as “petty fines and incomes”. It apparently consisted of incomes from various rights: Luca de Barberiis, 78/9 n.76.
 The xurta or scorta was the city guard or police patrol: Amari (1937/8), iii. 916-917. Bresc (1975), 134, gives xorta as “fines and taxes on games of chance”.
 Gisia: the jizya, a tax paid by Muslims and Jews. This term was identified in Peri (1978), 154-155. The surviving ms. clearly gives quingentor’. The reading Gisie is clear in both cases, but Winkelmann (1880), i. 713, incorrectly gave cusie. The interpretations of “cusie” in Dessoulavy, 539 (mintage), Valentini, 5 (caxia), and Bresc (1975), 134 (tax on tailors), may now be ignored.
 Apothece: shops.
 Cabella Corbinorum: Valentini, 5, suggested it might be a corruption of centimulorum (mills); this may be possible, since the curia had four such mills on Gozo while centimula de cabelle were not mentioned on Malta: infra n. 35. It would follow that mills produced 150 tareni on Malta and 890 tareni on Gozo. Dessoulavy, 539, proposes a tax on basketwork; possibly the tax was on some commodity contained in baskets. Corbisierorum referring to a tax on rope-making, is also a conceivable amendment.
 jardena: possibly fields for trees and fruit with a wall to exclude animals and some form of irrigation.
 orti: possibly enclosed vegetable gardens.
 10,901, but the figures given amount to 10,917.
 The madia was the Gozo ferry boat: Bresc (1975), 133 n. 40, 134 n. 47. Apart from the names Malta and Gaudisio in place of the classical Melita and Gaudos, the Arabic terms xorta (supra, n. 8), marramata (infra, n.28) and madia are in a sense probably the oldest surviving written Maltese words.
 uxorta, given above as Sorta: supra, n. 8.
 Centimulum: mill driven by animal power.
 3781, but the figures given amount to 3384.
 14,681: the two totals given amount to 14,682.
 Paricola, a Sicilian field-measure documented as paricla in 1208 or 1209: Winkelmann (1880), i. 86,89. A text of about 1247/8 read quod feuda Sicilie statuta fuisse dicuntur in triginta periclatis de terris ad triginta salmatas per periclatam: ibid., i. 701-702. The Sicilian salma was 1.75 hectares: Bresc (1986), i. 55. So the curia cultivated about 29 square kilometres, or roughly 9 percent, of Malta and Gozo.
 Frederick II’s governor of Malta in 1239 and 1240, and possibly in 1235.
[108 Ms: uit’ (?) or uin’ (?).
 Vacketta or vacheta: light ship.
 47: possible emendations are discussed supra.
 The small island of Comino appears to have been uninhabited, or at least it paid no recorded taxes.
 The Sicilian salma contained 2.75 hectolitres of grain: Bresc (1986), i. 55.
 Ancille: female serui.
 Marramata: wall or building work. This word, also used in Sicily, certainly derives from the Arabic: A. Varvaro. “Problematica dei Normannismi del Siciliano”, Atti del Congresso Internazionale di Studi sulla Sicilia Normanna (Palerrno, 1973), 364; G. Bautier-Bresc - H. Bresc, “Maramma: i Mestieri della Costruzione nella Sicilia Medievale”, in I Mestieri: Atti del II Congresso Internazionale di Studi Antropologici Siciliani (Palermo, 1983); G. Wettinger - M. Fsadni. L-G’anja ta’ Pietru Caxaru: Poeżija bil-Malti Medjevali (Malta, 1983), 52.
 The Sicilian cantaro or quintal weighing 79.3 kilograms: Bresc (1986), i. 56.
 Companagium: food provision.
 84 serui from the island of Gerba.
 144: but the figures given amount to 166.
 Castrum maris: the earliest clear reference to a castle in the Grand Harbour.
 The castrum of the ciuitas Malte: the modern Mdina.
 in centimulis quatuor de cabella: the incomes given include nothing from mills in Malta but a gabella of 890 tareni from the centimula curie on Gozo, while there were three such mills in the Castrum maris and one in the Ciuitas on Malta and one in the castrum on Gozo. If the document is correct, four other centimula de cabelle, listed after the centimulum in the Gozo castrum, produced 890 tareni on Gozo.
 The third castle is presumably that on Gozo, and not a third castle on Malta as Pertusi, 284, supposes.
 Mss. uiunt.
 Ms. manus.
 Probably the Constitutions of Melfi promulgated in 1231: Pertusi, 285.
 Possibly fideles since all other references are to fideles, while feudales might imply a group of feudatories which is otherwise totally undocumented on Malta in this period.
 Read mores?
 Ms. manus.
 Read consideras comodum?
 Read quod sicut?
 Read insulis?
LIST OF ABBREVIATIONS
|AAF||Archbishop’s Archives, Floriana|
|ACM||Archivum Cathedralis Melitae|
|AGPV||Archives of the Grand Priory, Venice|
|AOM||Archivum Ordinis Melitae|
|CM||Consolato del Mare|
|JHA||Journal for the History of Astronomy|
|MMV||Martime Museum, Vittoriosa|
|NAM||National Archives of Malta|
|NAMCons||National Archives, Mdina -Consolato del Mare|
|NAR||National Archives, Rabat|
|NAR Lib||National Archives, Rabat – Libretti|
|NAR RP||National Archives, Rabat - Register of Passports|
|NAR RSB||National Archives, Rabat - Register of Ships built in Malta|
|NAV||Notarial Archives, Valletta|
|NLM||National Library of Malta (Valletta)|