Malta in the High Middle Ages (original) (raw)

Godfrey Wettinger

[1] The assault on Malta and Gozo by the Muslims of Tunisia in 870 AD has always been seen as marking the end of one era and the beginning of another, the first going back so far as one can see in an unbroken line to the late Roman and classical and even earlier times including in its embrace both Latin and Punic and perhaps even earlier cultural traits. The second era, on the other hand, is signalized by the existence and persistence throughout the subsequent thousand years, at least after the year ca.1000 AD, of the Arabo- Maltese language, itself naturally slowly developing and changing down the centuries until it became what we now call our own Maltese language. But it has always been difficult to define exactly how one period passed to the other; in other words, was there an overlap, or was there instead a pretty well-defined gap between the two periods?

I was first directly confronted by the problem almost thirty years ago when I wrote a paper on ‘The Arabs in Malta’ first published in 1984. I had to admit my ignorance through lack of information. Practically nothing was known then about the events of 870. Kitab al-Uyun had come to be known and its reference to destruction in the Maltese islands in 870 and the transport to Tunisia of sundry architectural features of Maltese churches, but practically nothing else. The contemporary Chronicle of Cambridge which has survived, partly both in Greek and in Arabic versions, unfortunately does little better than establishing the date, partly confirmed by al-Athir. But Al Himyari, who remained unpublished until 1975 having been discovered in the inter-war period by the French scholar of Jewish extraction, Evariste Levi Provencal, is much more generous with information on the events of 870. According to him,

‘[Malta] was attacked by Khalaf Al-Khadim, the mawla of Ziyadat Allah ibn Ibrahim in the time of Abi Għabdalla Muhammad ibn Ahmed, the son of Ziyadat’s brother, with the assistance of Ahmed ibn Ghomar bin Ghabdalla bin Al-Aghlab and he reduced it to destitution at his [the latter’s] command. And this Khalaf is remembered as the builder of mosques, bridges and reservoirs. He besieged Malta but he died during the siege. So they wrote to Abu Għabdalla about his death. Abu Għabdalla wrote to his provincial governor in Sicily, Muhammad bin Khafaga, that he should send them a commander. He sent them Suwaada bin Muhammad, and they conquered the fortress of Malta, and they got possession of its king Ghamrus whom they enslaved, they demolished the fortress of Malta, and they plundered and stole whatever they were able to take. And he transported to Ahmad from the churches of Malta the material that he used to build the castle at the sea entrance to Susa and the road to it across a bridge [and that was the year 255]. And the island of Malta remained after that a ruin without inhabitants. Shipbuilders used to enter [the island] as its timber was of the toughest kind, fishermen of the abundant and tasty fish of its shores [went there], and collectors of honey, the commonest thing there.’

It is very well-known that Al Himyari lived several centuries after the events of 870. It is generally believed that his immediate source was a very well-informed and reliable Arab historian, most probably Al-Bakri himself. Unfortunately, the corresponding passage in Al-Bakri’s account has not been found or, anyway, traced so far and it is not possible to check on it. Of course, Al-Bakri himself lived almost a couple of centuries after 870 and one would also like to find the next intermediate text.

Very little further reliable information is available about what happened in the taking of Malta in 870 AD. Kitab al-‘Uyun speaks of the conquest of Malta in 256/870-71, attributing it to the 28 August, only one day’s difference from that of the Chronicle of Cambridge, a remarkably close correspondence.[2] Kitab al’Uyun cites the inscription on the castle of Habashi at Susa which was recorded by Ibn al-Ġazzar (ob. 1004): ‘All the hewn stones and marble columns in this castle were brought here from the conquest of Malta by Habashi son of ‘Umar’’, identified by M. Talbi with Ahmed bin ’Umar bin ’Abd Allah bin Ibrahim bin al-Aghlab, the commandant of the Muslim fleet.[3]

For the Maltese historian, myself included, until we knew of the details of the assault on Malta in 869-870, it was possible to argue that, in 870, any survivors of the catastrophic defeat of the island’s Christian defenders whether Byzantine Greeks or indigenous islanders, having survived the massacres that could have accompanied the Tunisian victory, would probably have gradually and progressively and inevitably been assimilated to their victorious foes both in culture and religion. That was the basis of my interpretation of Maltese medieval history for about twenty years. In fact, I was surprised how calmly my students and the general public accepted my words though the Maltese public thinks of itself as more Catholic than the Pope himself. Perhaps the suggestion that, in later times after the capture of the island by the Normans and as a result of heavy-handed state pressure, the Maltese again switched over from Islam to Christianity, deadened the shock of the first religious change, that from Christianity to Islam. But, after the revelations of Al Himyari’s account, it became increasingly inevitable to think that that same year 870 had actually witnessed an outright ethnic break and not merely a cultural and religious switch on the Maltese Islands.

The next clear evidence is that provided by Ibn Hauqal who wrote a short paragraph about Malta after dealing with Sicily in the Chapter on the Mediterranean Sea. This formed part of his descriptive account of the face of the earth but he was himself personally interested mainly in commerce and trade and apparently not at all in history. Thus he omits any information on what happened in Malta in 870, but records that Malta was uninhabited, but was occasionally visited by those who rounded up the donkeys then running wild, and hunted and slaughtered an occasional sheep for their own food, but otherwise took care to take provisions with them on their visits. The donkeys were sold abroad because there was a market for them there as draught animals, but not for the sheep. Honey was also collected and exported. It has been objected that Ibn Hauqal, in one surviving text, writes Malta with the spelling of Galita (‘G’ here standing for the Arabic sound of ‘rghain’) but spells it correctly only in the map, where it is also rightly placed to the south-east of what must be Italy. There is, therefore, no doubt that Galita is merely a scribal error for Malta caused by the strong resemblance in Arabic script between the letters ‘m’ and ‘rghain’. It has also been observed that Ibn Hauqal had strong Muslim sectarian prejudices, but no one seems to have noticed that his description of Malta as a totally uninhabited island renders any such prejudice completely irrelevant.[4]

‘On compte au nombre des iles connues, quoique inhabitées, l’île de Malte, située entre la Sicile et la Crete où vivent de nos jours des ânes qui sont devenus sauvages, ainsi qu’une immense quantité de moutons, et qui produit du miel: aussi des étrangers y abordent-ils, avec des provisions pour recueillir ce miel et faire la chasse aux ânes et au moutons. Les moutons n’ont guère de débit, mais les ânes peuvent être exportés, vendus et mis en service ailleurs.’

Transl. G. Wiet.

‘Among the famous unpopulated islands is the island of Malta located between Sicily and Crete. Until this day, there are in this island donkeys running wild and many and numerous sheep and goats in addition to honey. People head for the island with provisions to collect the honey and hunt for donkeys, sheep and goats. The sheep and goats are unsold but donkeys may be brought to certain destinations where they are sold and put to work.’

Transl. Adel Bishtawi, a prominent Palestinian writer, 2011.

His chief modern editors, J. H. Kramers and G. Wiet, believe that Ibn Hauqal’s description of Sicily, which included what he had to say about Malta, was recorded after the rest of the book was written, that is not much before the year 988, the date of its final and definitive edition; he was himself in Palermo on 16 April 973. Unlike so many other medieval chroniclers and compilers of voyages, Christian as well as Muslim, he did not have to rely on other accounts to fill in his own description of Malta. He could be immediately contradicted by his contemporaries if he was not faithful to the current facts known to everyone.

It will also be noticed that, except for the remark on Malta’s production of honey, Ibn Hauqal’s account is independent from the one found in Al Himyari and therefore deserves to be considered as corroborative on such an important matter as that of the complete lack of a permanent population on Malta, proving that Malta remained deserted for over a hundred years. Neither Al-Himyari nor Ibn Hauqal record the existence of a Christian dhimma type community on Gozo, like the one that has recently been imagined to have lived there.

It should be added that Maltese historians have tended to be seriously disconcerted by Ibn Hauqal’s unforgiving picture of a deserted island except for donkeys, sheep and goats and have tried to discount his evidence. Predictably, emphasis was laid on the fact that one particular manuscript spells the place-name in the text as Galita and another as Ħalita but always Malita in the map. I was among the first to point out this undoubted fact and explain it away as the result of a scribal confusion in the Arabic script between the characters for ‘rghayn’ and ‘m’.[5] A.T. Luttrell had, however, already studied Ibn Hauqal as early as 1987.[6] He was obviously unimpressed by him, though he was familiar with Ibn Hauqal’s text on Malta if only in G. Wiet’s authoritative translation. He acknowledged that, if Ibn Hauqal was correct, then

‘Malta was largely or wholly depopulated after … 870 … and it could have been resettled with Arabic-speaking Muslims and miscellaneous slaves and captives. In this case there would be no continuity of a Maltese ‘race’ or of an indigenous Christian or Pauline tradition.’

For the rest he dwelt on Ibn Hauqal’s remarks on the low state of religion in Sicily and the dietary and marriage habits of the Sicilian Muslims. Luttrell ends up by stating that

‘Ibn Hauqal’s work cannot be considered a reliable source for Maltese history in the tenth century…. (His) text may serve to remind historians that simple clear interpretations may be far from the truth.’

In his later writings, written after Al Himyari’s passage on Malta became known, Luttrell merely agrees that Al Himyari’s account increases the probability that Malta was abandoned for a considerable time after 870.[7] He records the possibility of confusion in the spellings of Malita and Galita as pointed out by myself in 1990. Brincat dwelt at length on philological aspects of Ibn Hauqal’s travel book in his Malta 870-1054 Al-Himyari’s Account and its Linguistic Implications (Malta1995) but, in his latest massive scholarly work Maltese and other Languages (Malta 2011), he chose to omit Ibn Hauqal completely, as does Charles Dalli who records the alternative spelling in one manuscript in his book in Maltese Iż-Żmien Nofsani Malti (Malta 2002), but omits Ibn Hauqal completely in his later Malta the Medieval Millennium (Malta 2006). Finally, in 2010, the present writer made a much needed public ‘agonising reappraisal’ of Ibn Hauqal’s reference to Malta, which he is now confirming, taking Ibn Hauqal’s brief notice of Malta as excellent corroborative evidence for the complete depopulation of the island that lasted about a hundred years.[8]

May it be remarked that Michele Amari utilized to the full and published in Arabic and Italian translation Ibn Hauqal’s description of tenth century Palermo, all eighteen pages of it.[9] Francesco Gabrieli, in his 1961 paper, concentrated on additional passages available in Kramers’s edition meant to substitute De Goeje’s edition, passages which reinforce Ibn Hauqal’s poor opinion of the Sicilian Muslims with whom he came into contact and it was not a flattering picture. I am suggesting that the sectarian proclivities and antipathies of Ibn Hauqal should be safely disregarded by the historian of Malta since he reported that Malta did not have any people but only donkeys, sheep and goats, no Sunnis or Shiites, or Christians. Gabrieli was still full of admiration for Ibn Hauqal’s ‘la preziosa e da tempo nota descrizione topografica di Palermo’.[10]

Another Arab writer nicknamed Ad-Dimashqi wrote that Galita was an island known as the island of sheep. According to him, there were many sheep living freely there, grazing and multiplying. Rarely was one of them slaughtered. Like wild beasts, they fled man. There was also a convent known as the convent of the sheep and goats.[11] The passage is in part strongly reminiscent of what Abu Hamid al-Andalusi (died 1169) said about Galita: that it was full of fat sheep belonging to no one. People landed there, capturing numbers of them, which they slaughtered and with their flesh filled their vessels, without ever making them scarce, they were so numerous.[12] In another passage preserved by Al-Qazwini (1203-1283), the same Abu Hamid al-Andalusi said that Malta, situated in the seas of the Byzantine Empire, was full of mountain sheep which jumped about like grasshoppers; there was such a large number of them that they could not escape people who, arriving in their ships, took as many as they wished.[13]

Precisely at the time that Al Himyari’s text came to be known in Malta, I remember telling a small class of my History students that archaeologists should easily find out through appropriate excavations the approximate length of time during which, according to the accounts of Arab writers, Malta remained unoccupied. Al Himyari seems to suggest that Malta remained deserted between 870 and 1048 AD, quite a substantial lapse of time which left me frankly stupified. Since then, Mr Nathaniel Cutajar together with Alessandra Molinari, an Italian archeologist, have tried to reach at least a preliminary finding on the matter. Going by what they published in a preliminary report published in 1999,[14]

‘Ceramics pertaining to the Late 10th and to the 11th Centuries have been identified at Mdina, Ċittadella, Tas-Silġ and at San Ċir…. Contrary to what some Arab chroniclers maintain, Mdina is clearly occupied throughout the 11th Century, and possibly in the Late 10th Century as well. It is, however, still not possible to say whether the Islamic urban and rural settlements followed closely on those of the preceding Byzantine ones.’

This is not far from what the Arab writers I have cited have reported, as I have just indicated. One has to remember that, though they all said that Malta was uninhabited, they also all asserted that the island (of course including Gozo) was occasionally visited by foreign ships with the men manning them looking for and collecting honey, fish, timber, donkeys and mutton. The gap not spanned by the archaeological finds between the years 870 and the Late 10th Century amounts to over a century, that is about three generations, as I have recently had occasion to point out.[15]

All these various Arab writers referring to Galita or Malta can surely not be ignored though one has to discount their more fantastic details. That, after all, is the methodology adopted by serious historians who have had to resort to hagiographical records when nothing else is available. All agree in describing an uninhabited island full of sheep. None mention Gozo as an island with a community of dhimma status Christians. And Dimashqi’s reference to a convent of sheep and goats must involve a confusion of the Arabic word standing for monastery and that standing for sheepfold such as have been found by myself among the oldest place-names of Malta (none in Gozo), names like Dellimara, Dejr il-Baqar, Dejr il-Bniet, Dejr iz-Zara’, Dejr il-Ħmir, Dejr Magħtab, Dejr il-Bużbież, Dejr Deru, Dejr Ħandun, Dejr is-Saf, Dejr is-Safsaf. At first, I tried to fit it into the picture by taking it to mean ‘a monastery or monastic type building’, but only at id-Dejr itself at Rabat were there any religious associations.[16] I, therefore, decided that all the other dejr place-names were connected to the practice of pastoralism, which I for one am now tempted to associate with the earliest phase of Malta’s re-peopling after the end of the long period that the island remained completely deserted, except for temporary visitors. It would have been natural for the earliest permanent tentative settlers to build animal farms or even sheepfolds for the sheep they could round up. Of course, monasteries were not involved, and that was one of my earliest conclusions (going back to 1976) on these place-names recorded in the surviving documentation from several centuries later.

It would seem that these incipient attempts at resettling the island of Malta occurred towards the close of the tenth century, soon after the death of Ibn Hauqal. The other two geographers mentioned, as well as Al Qazwini, themselves lived more than a hundred years later than Ibn Hauqal. It would seem that the early settlers were probably the result of individual personal initiatives, but no real information on them is available. Together with the dozen or so dejr place-names, as well as other probables, linguistically closely related place-names like Imrieħel, equivalent to Mnadar, and Merħlet il-Mogħoż, merħlet itself equivalent to mandra, one could add the important place-name Marsamxett, that is, the port of Mxett localities, plural of mixta, really stock-yards or pens for animals, not wintering places for ships, and Tal-Mixta in Gozo, animal pens rather than sheepfolds, and perhaps a few others closely related to animal husbandry like mandra or merħla, mnajdra or mrejħla, and perhaps baqqari and ħmieri, farms specializing in the breeding and keeping of cattle and donkeys equivalent to Dejr il-Baqar and Dejr ilĦmir, dwejra, and sites of basic occupations like mellieħa, later sometimes replaced by salini, forn il-ġir, later replaced by kalkara, and mġiebaħ (or naħlija), as well as others related to the manufacture of pottery, like Tal- Faħħari and Il-Qallelija, might also belong to the same early period of resettlement, together perhaps with torba or grave place-names, as distinct from ċimiterju or zuntier, which came much later. But one must beware of speculating before much further study.16a Most of them survive only in place-names. The connections with Sicilian words like mandra, kalkara, and salini make a Sicilian origin for the earliest Muslim settlers highly probable even though they spoke an Arabic dialect only itself introduced into Sicily in the first half of the ninth century.16b

The next stage in the economic development of Malta, also quite inadequately recorded merely by place-name evidence, was the establishment of arable agriculture signalised by the appearance of rahal place-names, none of them on Gozo. These settlements probably depended on a mixed type of economic activity, arable as well as pastoral. Over a hundred of them have been discovered in the archival records, and probably appeared over a lengthy period, and were closely associated with the appearance of galcae, fields enclosed by rubble walls to protect the crops from depredations by the sheep and goats. Unfortunately, owing to the nature of the surviving documentation, their origin cannot be closely dated. Finally, however, we reach the date of the rebuilding by the Arabs of the first town on Malta, Mdina itself which, according to Al Himyari, occurred in the year 1048. That date might, however, indicate the official rebuilding of the town, perhaps the erection of the town walls. Individual settlers might have already taken over parts of the ruins of the Byzantine city which would be documented archaeologically by the occasional shards so valuable to the archaeologist. But for a proper distinction to be made between signs of a purely temporary stay and those of a more permanent nature, a preliminary report would probably not be sufficient.

At this stage, one could perhaps speculate on the place of origin of the late tenth and early eleventh century settlers on Malta. They were certainly Arabic-speaking Muslims. Malta was surrounded by such populations, and all of them had serious internal and external political, social and religious tensions which could have spurred migration to Malta or elsewhere. On the whole, it seems probable that a distinction should be made, as I have made, between the probable small-scale original new settlers, perhaps on an individual or family basis, and the one of c.1050 when Mdina was refounded. The latter must have been carried out on a relatively large-scale community basis, involving at least a few hundreds of individuals or even families, either of a community of new arrivals or of joint action by the slowly increasing settlers.

It may be asked why it took more than a century for the resettlement and reconstruction of the Maltese Islands even to begin, though it has been shown that the islands had some, if rather limited, assets such as abundant fish, some timber for shipbuilding, much honey, donkeys for export for draught use and certainly an abundance of sheep. But it must be remembered that the attraction of Malta for migrants had to compete with the pull of neighbouring Syracuse which was treated the same way as was Malta only eight years later. Presumably, the resettlement of Malta could not start before the more urgent and better endowed Sicilian city and province (?) of Syracuse were at least partly resettled and reconstructed. In fact, Syracuse never regained fully her former prominence.

In view of all this, one might perhaps now venture the idea that the earliest significant permanent settlers in Malta could most probably have come from the nearest parts of Sicily, that is, that it was part of the revival and re-population of South-Eastern Sicily all the way from Syracuse round the coast to Gela and similar localities, all places that had been devastated in the Muslim campaigns of the 870s, themselves Muslim peoples who had been advancing all the way from Western and Central Sicily, having crossed over from Northern and Central Tunisia not much more than two or three generations before.

However, there is documentary evidence of other Arab-speaking people arriving directly in Malta from Tunisia during the first half of the eleventh century. Thus, in 1040, the Idrisid rulers of what is now Tunisia sent an army to Malta: ‘the Żirid amir al-Muiżż ibn Badis sent a large army there’ and, eleven or twelve years later, when the Banu Hilal or Sulaym Bedouin tribes even sacked Qairouan, large numbers of the inhabitants of the towns along the Tunisian coast fled to Sicily and Andalusia. Although Malta is not mentioned, ‘there is a very good probability that many fled to Malta as well as Sicily.’ That largely coincided with the reputed rebuilding of Mdina, Muslim Malta’s only town.16c

That explains the most important phenomenon of the medieval micro placenames of Malta and Gozo, namely that they were prevailingly of Arabic origin with just a sprinkling of those of a non-purely Arabic origin, place-names like Mandra, Mnajdra, Latnija, Lwatan, Qortin, Qraten, Qrajten, Rinella, Stabal or Stabar, Wardija, and possibly Nigred as well as Qala. Deeper study should reveal that some of these, like Wardija, were clearly of a somewhat later origin, a possibly earlier equivalent of the latter could have been Nadur.[17] Another prominent characteristic of these thousands of local minor place-names is that so many of them refer to individual persons by name, nickname or family association, thus preserving in a way for hundreds of years a record of their existence and their otherwise unknown connection to the locality: eg. Salama (Sliema), Ġagħfar, Wied Għomor. Presumably, this was the result of the nature of the arrival and settlement of a new people on an empty island. Particularly potentially significant is the place-name Ta’ Xiberras, in later times a prominent surname of the family which had rights of ownership over the hill stretching from the site of St. Elmo to Ta’ Braqxija (Braxia) and originally a personal name meaning Xeheb ir-Ras, Xeheb the Chieftain. On the same peninsula was Misraħ Mawwija, Mu’awwija’s rough grazing ground, recalling one of Islam’s earliest figures.18

Then, during the middle years of the eleventh century, occurred one of the most notable events of Maltese medieval and Arab history, narrated both by Al Himyari and Al Qazwini. Both of these Arabic writers, born at the two extremities of the medieval Arabic world, lived at a much later date and must have obtained their information from other writers. Both versions really complement each other but Al Himyari’s account is by far the fuller and more reliable one. It starts off by saying very clearly that, after 1048, Malta was inhabited by the Faithful who were free and the Għabida or slaves:

‘After the year 440 [after 5 June 1048 AD] Malta was inhabited by the Muslims who [re]built its town, which was restored better than it had once been. The Byzantines in innumerable ships attacked it in the year 445 [1053-54 AD] in great numbers, besieging the Muslims in the town. The Muslims were pressed hard and [the Byzantines] had great hopes of [success]. The Muslims asked them for the amman [conditional surrender], but they refused except for the women and goods. The Muslims counted the number of combatants among themselves and found that they were approximately 400; then they counted their slaves and they discovered that they were greater in number than they were. They therefore gathered them together and told them: ‘If you are faithful to us in the battle and follow us in our fate and finish where we finish, you shall be set free and join us and we shall give our daughters in marriage to you and share our wealth with you. But if you temporize and leave us in the lurch, you shall attain captivity and the same bondage as ourselves. You shall be worse in your situation than we shall be, since in our case one may be redeemed by an intimate friend, or freed from captivity by his governor, or his community will sustain his rescue.’ The slaves promised more than they expected of them and they found them in the fight with the enemy more prompt than they were themselves, for when the people awoke on the second day, the Byzantines came early at them as is their practice, and they had hoped that day to overcome and enslave them. But the Muslims had made their dispositions well for meeting them, and they awoke with a clear understanding of their struggle, and they asked God for assistance to be strong and exalted on their behalf. And they marched and they dashed towards them, knocking them down with their lances and piercing them with their swords without fear or falter, sure of one of two good things, either speedy success or the triumph of death. God the Exalted provided them with help, giving out patience to them and alarm to their enemies; repulsed they fled away without turning round, death exterminating most of them. The Muslims captured their vessels, none escaping but one. The slaves then attained the status of free men and what was told them was fulfilled; after that, the enemy feared their power and not one of them opposed them for a time.’

Both in this version and in Al Qazwini’s there is no reference to the existence of local Christians, the only Christians mentioned being the invading Rum, who were repelled in both accounts. Both versions describe the inhabitants as divided into free persons and Għabida or slaves.19 Michele Amari, first brought Al Qazwini’s account to the notice of Maltese historians, most or all of whom subsequently came to believe that the għabida were Christian descendants of survivors of the massacres of 870 A.D.20 Before Al Himyari’s detailed account of what happened in that calamitous year came to be known to all, I toyed with the same interpretation believing however that, by 1048, they had completely succumbed to the blandishments and threats of the Muslim authorities and had finally reluctantly accepted conversion to Islam. The details themselves of the combat of 1053 made it unlikely that they were still Christians. They fought really hard to repel the Byzantine Christian invaders and they were eventually granted political and social equality by the Muslim elite, the sheikhs, after the fighting had come to an end.21 There was still the difficulty that if they were not related to the original islanders who had fought even more valiantly in 870 against the Tunisian invaders, they could best be understood to have been black mercenaries of African origin who were so common in the Maghribi Muslim countries and in Central Mediterranean islands like Sicily and the Balearics. They were officially reported to have outnumbered the free Muslims in Malta in 1053. If they were all of African origin, they should surely have left substantial surviving signs of them in the ethnic composition of the Maltese population, which is not the case. Incidentally, any among them of white origin would probably have been Slavs of East European origin, related to the placename Skorba (Scaliba) and the surname Sicilianised as Scavuni or Italianised as Schiavone, extant in Malta or Sicily throughout the Later Middle Ages. Perhaps the historically correct solution of the problem is to be found in twenty-eight lines of the twelfth century poem in medieval Greek recently translated by three Maltese scholars, as we shall be seeing.

Both versions of the events of 1053 make much of the fact that the united inhabitants lived happy ever afterwards, indicating that the remarks were recorded certainly before the unhappy events of 1127 and probably before those of 1091. Recently, a gold quarter dinar coin struck in Malta in 1079 AD has been purchased by the Central Bank of Malta, really a remarkable and unexpected find, leading to much speculation on its significance in the economic life of Malta at the time, precisely when the Muslims were the only inhabitants known then to live in both Malta and Gozo, no doubt in accordance to the terms of the social and political pact of ca.1050. Was their economy based on peaceful pastoralism or, at best, on the production of wheat and barley? Had cash crops like cotton or cumin already been introduced, or was it instead really and largely dependent on the preying activities of local privateers as it was to be for so many centuries? In fact, until the coming of Bonaparte and his French soldiers in 1798, Malta was constantly a base for privateering, at first by Muslim local vessels preying on Christian mainly Greek or Byzantine shipping, then after 1127 by local Christian ships preying on Muslims? On the other hand, was the gold coin merely making a political point and not marking the level of economic activity at all? One thinks of the gold coinage of Merovingian France which misled Henri Pirenne into thinking that the economy of the West had not collapsed disastrously with the political breakup of the Western half of the Roman Empire until the irruption of the Muslims into the West in Carolingian times which was signalised by the switch to a sole silver coinage, a misreading until it was pointed out that the Merovingian gold coinage was largely only intended for political propaganda purposes.22 Up to now, the invaluable deposit of documents found in the Cairo geniza has not revealed anything on Malta. Alone, the coin does not reveal much.

The coin was struck a mere twelve years before the coming of Count Roger in 1091. In fact, the Normans had started the reconquest of Sicily as early as 1061 and they had long been extremely active in destabilizing and occupying various parts of South Italy. In February 1090 ab Incarnacione, they completed their occupation of Sicily by the cession to them of Noto. In the following July 1091 (a Nativitate, which is the way most of the modern world computes the years of the current era), the Normans crossed over to Malta with, ‘a great army gathered from all over Sicily’. Count Roger was personally in command and it was recorded that,

‘The count’s ship, which sailed ahead of the rest thanks to its speed, was the first to make landfall. The count left the ship with only thirteen knights. They mounted their horses and attacked the great multitude of inhabitants who had come to meet them at the shore to prevent them from advancing. The count killed many of them and put the rest to flight, pursuing them for some distance as those who lagged behind were cut down.

When the morning dawned, the count advanced close to the city and set up a siege; he sent his men all around the island to plunder. Gaytus, who ruled the city and the island was, like the rest of the citizens, unaccustomed to the demands of war and were terrified by the enemy presence. He and his citizens asked for a truce so they could speak freely with the count. When this was granted, they came to the count’s tent to sue for peace. After attempting various circumlocutions, they finally realized that they would not be able to deceive the shrewd prince. In accordance with the count’s desires, they first released their Christian captives, a great many of whom they held within the city.

Then they handed over to the count their horses and mules as well as all the arms that they possessed, along with a great deal of money. After determining what would be given in tribute each year, they promised that the city would serve the count. Thus they were joined to the count with oaths given in the manner of their law. Seeing the Christian captives as they left the city – tears of joy at their unexpected liberation flowing from the depths of their hearts – carrying in their right hands crosses made of branches or reeds or whatever else could be found at hand, shouting out ‘Kyrie Eleison’, and bowing down at the count’s feet, our men were themselves covered with tears, touched as they were by the emotion of such a pitiful sight. With the city now subject to him, the count placed the captives on the ships and hastened to return to Sicily, all the while apprehensive that the ships might sink under the greater weight of the additional passengers. But the hand of God, as we believe, was clearly manifest: bearing the ships through the waves, it lifted them a cubit higher on the sea, so that their weight of the ships’ cargo was less of a burden than what it was when they were sailing to the island in the first place.’

‘As the count hastened to return over the sea, he spied an island in the distance called Gozzo and ordered the sails to be adjusted so he could go there and attack it. Landing on Gozzo, he laid it waste, plundering it, knowing that in this way the inhabitants would seek terms with him. Thus he placed the island under his dominion.

He returned to Sicily heavily laden with booty, large amounts of which he donated to the faithful vassals who had missed him so much. He then addressed all the Christians that he had brought back with him, making them free men and offering those who wanted to stay in Sicily the chance to build for themselves villages at his expense anywhere they chose, supplying them with all they needed for making them profitable, declaring such localities to be free, that is exempt of dues and feudal service for ever. To those who wished instead to return to their own properties and to see their friends, he gave them permission to go anywhere they wished, granting them passage as needed from anywhere in his territory across the straits of Faro at no cost. In their great joy, the former captives gave thanks to God and to the count for their liberation. Each (of the ones who chose not to stay) then left for his homeland by various routes according to his nation, spreading the name of the count as they sang his praise.’

Going through the whole of Malaterra’s account, one finds that the only Christians mentioned were the captives who greeted him when they left Mdina through its open doors. No other Christians were involved either in Malta or Gozo. And no provision was indicated for the resumption of the practice of Christianity on either island. This should not occasion any surprise if one follows this paper closely. Only Muslims were mentioned in the two versions of the events of the 1050s; the events of 1127 ‘reported’ in the poem show that it was only then that Christianity was reintroduced, even then only at the point of the sword, except for the purely hypothetical possible reappearance of a new captive community between 1091 and 1127.

In fact, while in innumerable places that he conquered in Sicily the reader is told that he left a permanent garrison, erected a castrum or citadel, or left his wife in charge of some soldiers, in the case of Malta and Gozo no such information is recorded. When the two Norman brothers took Messina in 1062, they spent eight days arranging their affairs and strengthening the city as they saw fit, and when the citizens of Petralia, both Christian and Saracens, surrendered their castle to Count Roger, he garrisoned it with knights and mercenaries,23 and then went to Troina,24 where he left his wife and troops in charge. After taking Palermo in 1071, he strengthened its fortress and disposed of the city as he saw fit, retaining it under his own control.25 In 1077, it is expressly recorded that he organised the defence of Castronovo which he had just acquired, according to his wishes.26 At Trapani, he ordered the construction of a citadel and other fortifications on all sides,27 and in 1086 on taking Agrigento, he furnished it with a very strong castle surounded by a rampart, towers and bastions.28 At Noto, only a few months before the Malta expedition, Jordanes, Roger’s son, after receiving the surrender of the city, arranged to have a citadel built there.29 It will have been noticed that no such precautions are recorded to have been made either in Malta or in Gozo, but much was made of the havoc that his men wreaked in both islands.

Inevitably, this reawakens the old dispute on the precise significance of the invasion of 1091. For some time, students have been told that for most purposes Count Roger’s expedition to Malta and Gozo could be considered as little better than a raid in force. It was ignored by the Arab historians, who did not ignore the invasion by Count Roger II in 1127. We have seen that express provision was made in so many other places in Sicily for the maintenance of a garrison, the building of a castle, towers or a ditch, one sees the expeditionary force withdrawing from Malta ‘after obtaining the cession of the town’, merely taking lots of booty with them and all the freed Christian captives. No garrison is expressly left behind, no new castle or town towers. No prominent Norman was left behind in charge of the new acquisition. Unfortunately, just at this point, complete darkness again falls on the affairs of Malta, without a single contemporary documentary reference to the island for the next thirty-six years. The result has been that it has come locally to be viewed largely as a mere raid, what was then frequently called a razzia.30

However, the recent translation and publication by three Maltese scholars of a twelfth century Greek poem apparently composed on Gozo by an exiled Greek Sicilian writer31 has led to the discovery of a 28-line section purporting to describe what happened on Gozo in 1127. For the rest of this paper, the present writer is dependent on the published version, in fact, on the published translation of the original text. The text, as arranged for us by the three editors, speaks of Christians departing from the Pact of Old.32 Was this not a poetical version for what really happened, that the Pact had been abrogated by the new regime, that is, by Count Roger II and his men not by local Christians? The only local pact we know of is the one of the ca.1050s, and that could certainly be called the Pact of Old but it did not concern Christians at all and was not a dhimma pact. Its abrogation can only be construed as a complete reversal of the social and political structure that had been reached almost eighty years before.

But, after all, here we are dealing with a poem, and not an archival document which would have been a real relic of the past and could be examined as such, work to which historians are well habituated. But there is no reference to a dhimma pact at all, that was a hypothesis of the editors. The facts do not add up. Nor does the poem imply anything as strange as a dhimma pact in 870 AD about which date the poet could not have heard anything. Again, that was an unwarranted assumption of the editors. And what did the newly-found bishop and his native Christians do in the 36 years since the last known visit by the Normans on the island of Gozo? And is there not also the difficulty of understanding why the Greek exile who composed the poem did not find solace in the company of the band of Christians mentioned in the passage on an island full of fanatical Muslims? I myself know no Greek at all and I cannot deal with these matters as well as I would have liked.

It is highly probable that one of the principal aims of the expedition of that year was precisely to bring to an end the long survival of the pact of ca.1050, a social and political arrangement which had been reached exclusively between the two sections of an exclusively Muslim population and which must have been seen as an anachronism at the moment when Count Roger II rose to real power in, or soon after, 1127. Of course, this does not exclude that other factors in the balance of power in the Central Mediterranean were also involved. This is the passage:

‘Did not the great leader of the admirals / Himself open the troublesome gates / of the foreign godless Agarenes? / Do you not even know, Sir, how while the sceptre / Was not yet raised, in so many good [circumstances], / And while it (still) did not even exceed / The limits of the Council of the admirals, / The most resplendent leader of all the leaders, / Having mustered only a small naval expeditionary force / And a host of spear-bearing archer-infantry, / Sailed to Melitogaudos, the country of Hagar, (and) / Not having been dismayed at the impudence of the godless (sons of Hagar), / Having encircled (them) with diverse engines of war, / He subdued (them) with might and main for the Lord? / When he saw, on the one hand, these (inhabitants) / Invoking only the heresiarch, the all-abominable Mohammed, / He banished from the country their sheikhs, / With all their households and (their) black slaves,33 not indeed a few. / He on the other hand, brought out into the open / The pious inhabitants of the place, together with their bishop,34 / Who having departed from the Pact35 of old, / Got rid of the hated things by which they used to invoke Mohammed. / He then established into most sacred temples, / Places (formerly) belonging to the most hated Mouddibi, / Sacred and useful priests / Who were worshipping the Holy Trinity from ancestral times, / Interceding with this (Trinity) on behalf of the sceptre-bearing, / And rather for his purified soul.’

On the other hand, if the passage has been correctly translated, it provides good evidence for the persistence of the internal Islamic regime on both islands right down to the second Norman invasion. It was only then that the sheikhs were worsted, that their households and the black inhabitants, descendants of the għabida, now really no longer slaves, were crushed, killed or expelled, that the practice of Christianity on Malta, including Gozo, was resumed after a break of well-nigh 270 years, and that a different status was established for the remaining Muslims and the newly set up Christians on both islands. It was then, or soon afterwards, that a small band of local Muslim poets was exiled to Sicily never to be allowed back, and some little time afterwards that a Sicilian Greek poet was incarcerated perhaps on Gozo,36 where he spent nine years of his life remonstrating vainly against the Norman authorities for forcing him to live in continual association with Muslims whom he loathed so vehemently. It was only then that the local inhabitants began to lose much of their contact with fellow Muslims and Arabic speakers in the surrounding Arab world, and began gradually to settle down to a troublesome relationship with the new local and Sicilian Christians.37 However, the poem’s account still remains to be reconciled with what has become the normal picture of the two Rogers being relatively sympathetic to Arab or Muslim susceptibilities. Perhaps the harshness revealed by the Byzantine Greek poem in or around the year 1127 was the result of his admiral’s influence rather than Roger’s own personal permanent policy, as he was then still a relatively young person unable to thwart the designs of his more mature counsellers. And for long afterwards the Counts of Malta, themselves usually Lords High Admirals of the Kingdom, and the kings of Sicily were repeatedly at loggerheads, with Malta alternately suffering or benefiting as a result. Malta in the end dropped the practice of Islam but retained the use of the local version of spoken Arabic, unlike Sicily.

Perhaps it was only just before the Norman invasion of 1127 that the large treasure of gold coins that was discovered during the rebuilding of the cathedral after the earthquake of 1693 was hidden away. The finality of the Arab reversal of 1127 accounts for the length of time it took for the treasure to be rediscovered and recovered. The invasion of 1091 had been far less fatal in nature and, had the treasure been hidden away on that occasion, it could have been uncovered and recovered by the same administration that might have hidden it away soon after the departure of the Norman invaders of that year. After all, the signs are that the Muslim administration remained unchanged in 1091. Unfortunately, the coins were mostly melted down and were not examined individually, making it impossible to check whether their dates included some between 1091 and 1127.

Presumably, new settlers were sometimes purposely brought in, now almost exclusively Christians and perhaps partly of a Byzantine Greek cultural background of whom not much is known. But they were never so numerous as to swamp the linguistic predominance of the surviving Arabic speakers. That the latter were relatively very numerous is vouched for by the poem itself. The Greek liturgical words alleged to exist or to have existed in the Maltese language must belong to the post-1127 period. Since the composition of the Byzantine Greek poem has been dated by its three editors, or rather by Tsolakis their predecessor, to ca. 1135-51, the poet himself was not a witness to what happened when Malta (or was it Gozo?) was re-conquered by Count Roger II, but his poem is the only direct evidence of that event and he is loud in his complaints at the abundance of Hagarenes. Of course, further parties of settlers could have reached Malta or Gozo in subsequent times from nearby Sicily or Calabria or from as far away as the Middle East after the disastrous failure of the Crusades, from Antioch or Aleppo or other places. This eventually created another possible curious tenuous direct and probably continuous link with the Christianity of the First Millennium that is to be found in the thousands of persons with surnames such as Grech, Armeni, Darmanin and Darmenia, Calabachi (from Aleppo), etc., probably descendants of Christian refugees from Greece and the Levant in the same post-1127 period.38 Their historical continuity was part of the history of the countries from which they came whether Greater or Little Armenia, Asia Minor, Syria or elsewhere. But the contemporary Maltese, descendants of the Muslim settlers of the 10th-11th centuries, were themselves all Muslims, except for those among them who chose to conform to the new political and religious situation under the [later?] Normans or Suabians by becoming Christians. That process of Christianization was not to be completed before round about 1250 A.D.

The next Arab writer who wrote about Malta and Gozo was commissioned by King Roger himself in 1154 to write his description of the whole known world of his time. His account is therefore strictly contemporary to the poem composed by the Sicilian Greek person exiled to ‘Gozo’ if we are to accept the main conclusions of the three editors, but Idrisi’s account was of a totally different nature, the Greek poem being purely a work of literature while Idrisi’s was a careful compilation done after the most detailed enquiries concerning the various countries he was commissioned to describe. It was meant to be factual, informative and authoritative, not personal and emotional. This is what Idrisi wrote about the Maltese Islands:39

‘To the East of this island [Pantelleria] at a distance of one hundred miles there is the island of Gozo which has a secure harbour.

From Gozo to a small island called Comino… [distance not given in any manuscript]

To the East of this lies Malta, a large island with a safe harbour opening to the East. Malta has a city. The island abounds in pastures, flocks, fruit and above all in honey. Between this island and the closest point of Sicily there are eighty miles.’

My translation from Michele Amari’s Italian version.

In another passage Idrisi points out that the minor Sicilian port of Scicli was visited by vessels from Malta as well as from Calabria, Africa and other places. It will be noticed in the published version, which is easily available, that Idrisi does not make any reflection on the religious or ethnic composition of the population.

In 1175, the secretary of Bishop Burchard travelling through the Straits of Messina while on a diplomatic mission to Saladin on behalf of the emperor reported that the island of Pantelleria on the other side of Sicily was inhabited by Saracens who lived and hid in caves, while Malta itself was also inhabited by Saracens but no caves are mentioned.40 Nothing is said specifically about Gozo but it is probably included under Malta.

The next surviving bit of information comes from a royal document of the Empress Costanza dated November 1198 addressed to the Christian and Saracenic inhabitants ‘of the island of Malta and Gozo’ by virtue of which, in her own name and that of her son the Emperor Frederick, still a minor, she informed them that they were being returned to the royal domain and that she was forgiving the Christians themselves the annual penalty to which they had been subjected by her late father King Roger for the killing of a Saracen.41 Here we have the first definite confirmation that Malta and Gozo had two religious communities living side by side not without some conflict. King Roger is shown to have protected the Muslim community. In 1198, his daughter the Empress Costanza seems to be on the point of distancing herself from that policy. But Islam itself did not disappear completely from Malta before about the middle years of the thirteenth century.42

In the meantime, there appear the first clear signs of the setting up of a diocese in Malta with a Latin rite bishop.43 He is listed by Falcandus in 1168 as a member of a royal inner council in Sicily of ten familiares regis together with Bishop-elect Richard of Syracuse, Bishop Gentile of Agrigento, Archbishop Romuald of Salerno, three counts, the notary Qaid Richard, Walter Dean of Agrigento and the king’s tutor.44 Malta must at the time have been considered as a see in partibus infidelium, and the bishop normally resided in Sicily where he frequently occupied prestigious official and administrative posts, as we have seen. Actual work on the building of the cathedral is not heard of before 1299.45

Of course, the Maltese language by then was already in existence, having developed from the dialectal Arabic brought to Malta from Sicily late in the tenth century.46 It no doubt originated in the spoken Arabic which was taken to Sicily from Tunisia early in the ninth century, surviving in the Maltese islands to the present day, while Islam both in Sicily and in Malta was forcibly replaced by Christianity.

* * * * * *

During the last thirty years, two outstanding historical texts have come to light which took everyone by surprise, and enabled historians to reach at least much better tentative conclusions on important historical points. The first was the arrival in Malta in 1987, some twelve years after its publication in Beyrouth, of the text of Al Himyari concerning Malta, which should have immediately led to the proper and permanent re-evaluation of the text of Ibn Hawqal, which had been known for at least a century already, but was regularly largely discounted by local and even foreign historians. Then in 2010, three Maltese scholars translated and published the twelfth century Sicilian Greek poem which contained a twenty-eight line section concerning an alleged Norman invasion. This can best be regarded as related to the second Norman invasion as recorded in Ibn al-Athir, clearly indicating that Ibn al-Athir’s Norman invasion of c.1127 has to be taken seriously after all and would seem to confirm Anthony Luttrell’s dismissal of the significance of the first invasion, that of l091, as merely that of a raid in force, a razzia.

____________

Notes

[1] This paper was read publicly in the Ambassadors’ Hall, Auberge de Castille, on 7 December 2010. Some slight changes have been made to the text to clarify some points and extend others.

[2] Anonymous manuscript entitled Kitab al-‘Uyun wa-l-Haqa’iq in Staatsbibliothek Preussischer Kulturbesitz, Orientableitung, We 342 Bl.6a, fol. 6; published in Omar Saidi, ‘Kitab al ‘Uyun’, Les Cahiers de Tunisie, xx, No. 79-80 (1972), 51.

[3] M. Talbi, L’Emirat Aghlabide, Paris 1966, 475.

[4] F. Gabrieli, ‘Ibn Hawqal e gli Arabi di Sicilia’, Rivista Studi Orientali, xxxv (1961), 245-53.

[5] G. Wettinger, contributions to The Sunday Times, 26 August 1990, 30 September 1990; Joseph M. Brincat, Malta 870-1054 Al-Himyari’s Account and its Linguistic Implications, Malta 1995, 41-43; C. Dalli, IŻ-Ûmien Nofsani Malti, Malta 2002, 31, but omits Ibn Hauqal completely in his later Malta The Medieval Millennium, Malta 2006.

[6] A.T. Luttrell, ‘Ibn Hauqal and tenth-century Malta’, Hyphen, vol. v, no. 4, Malta 1987.

[7] A. Luttrell, ‘Slaves and Captives on Malta: 1053/4 and 1091’, Hyphen, vii, no. 2, Malta 1992.

[8] In Sunday Times (of Malta), 2 and 16 May 2010.

[9] M. Amari, Storia dei Musulmani di Sicilia, Edizioni Elefante, II, 336-54; Biblioteca Arabo-Sicula, Edizioni Dafni, 1982, I, 9-27.

[10] Francesco Gabrieli, ‘Ibn Hauqal e gli Arabi di Sicilia’, Rivista degli Studi Orientali, xxxvi (1961), 245-253.

[11] Ad-Dimashqi, ‘Kitab nuhbat ad-dahr’, Ms. at the Paris Bibliotheque Nationale, No.2187, fol. 72.

[12] Abu Hamid al-Andalusi, ‘Tuhfat al-albab’, ed. G. Ferrand, Journal Asiatique, t. 207 (1925), 105.

[13] Al-Qazwini, ‘’Aga’ib al-mahluqat, (Cairo, n.d.), 123.

[14] Alessandra Molinari and Nathaniel Cutajar, ‘Of Greeks and Arabs and of Feudal Knights. A preliminary archaeological exposure of Malta’s perplexing Middle Ages, Malta Archaeological Review, 3 (1999), 9-20.

[15] As I indicated in my letter to the Sunday Times, 2 May 2010.

[16] G. Wettinger,’Some Maltese Medieval Place-names of Archaeological Interest’, Atti del Colloquio Internazionale di Archeologia Medievale, II, Palermo, 1976, esp. 344-46, 361.

16a But not Mtaħleb since that seems instead to be related to the surname Calabachi, ‘the young man from Ħalab or Aleppo’, in Northern Syria.

16b For the place-names in this paragraph and elsewhere in this paper, see G. Wettinger, Placenames of the Maltese Islands ca. 1300-1800, Malta 2000.

16c Leonard C. Chiarelli, A History of Muslim Sicily, 130, citing Ibn ’Idhârî, Kitab al-Bayân al-Maghrib, 1:275, and ’Abd al-Wahid al-Marrakushi, al-Mu’jib fi talkhis akhbar al-Maghrib, ed. Muhammad Sa’id al-‘Iriyan (Cairo, 1963), 441. Also, Chiarelli, 131.

[17] G. Wettinger, Place-Names of the Maltese Islands, ca. 1300-1800, xv.

18 Ibid.

19 The relevant sentence in Al Qazwini says that “The Muslims counted themselves and discovered that the number of their Għabida was greater than that of the free men.” It is not necessarily implied that the għabida were not counted among the Muslims, rather the opposite. Għabida in the eleventh century were more often than not black slaves of African origin.

20 Thus A. A. Caruana, Frammento critico della storia Fenicio-Cartaginese, Greca-Romana e Bisantina, Musulmana e Normanno-Aragonese, 428-29. He was apparently misled by Amari’s reference to the united mass of the free men of Malta and their għabida as ‘i Maltesi’ in M. Amari, Biblioteca Arabo Sicula, i, 241.

21 G. Wettinger, ‘The Arabs in Malta’, in Malta. Studies of its Heritage and History, Malta 1986, 91.

22 Henri Pirenne, Mohammed and Charlemagne, 1939, esp. 107-117, 293; A.F. Havighurst, The Pirenne Thesis, 1976, 65-66, etc.; Richard Hodges and David Whitehouse, Mohammed, Charlemagne and the Origins of Europe, London 1983, esp. 6, 90-91, 109-110.

23 G. Malaterra, Res Gestae…, transl. by Kenneth Baxter Wolf (University of Michigan), 95.

24 Ibid. 106.

25 Ibid., 125.

26 Ibid., 143.

27 Ibid., 142.

28 Ibid., 181.

29 Ibid., 190.

30 A. T. Luttrell, Medieval Malta. Studies on Malta before the Knights (London, 1975), 31.

31 J. Busuttil, S. Fiorini, H.C.R. Vella, Tristia Ex Melitogaudo – Lament in Greek Verse of a XIIth- Century Exile on Gozo, Malta 2010. This publication has already been fleetingly referred to supra.

32 ‘Pact of Old’ would seem to exclude a simple reference to the pact of 1091. Ant it will be seen that the details of what happened in the Maltese Islands in that year are totally at variance with the details in the poem and with the known attitude of the two Rogers to their Muslim subjects.

33 The original Greek text refers to Ethiopians but without any indication of a servile status (this has been pointed out to me by my colleague Professor A. Bonanno, who has examined the original Greek text) which is added in the translation by the editors. The original Għabida normally understood in the contemporary writings to have been black mercenaries achieved their liberty after the repulse of the Byzantine invaders in c.1050s, and were still enjoying their social freedom and equality in 1127.

34 Possibly relatively recent captures in the 1091-1127 period.

35 That it was a dhimma pact was another intervention of the editors.

36 Not everyone agrees with Stanley Fiorini’s Gozitan identification.

37 I have never really had any reluctance to find some real historical information in the twelfth century Byzantine Greek poem recently published in Malta; see my contribution to the Sunday Times (of Malta), 21 March 2010. I was sure it could not have had any reference or relevance to the events that occurred in Malta or Gozo in 870 A.D. Throughout, I must have been influenced, even if unconsciously, by the finding of some reliable information about the slaves’ prisons and the slaves’ life in Malta in general as well as the inevitable personal details in a late sixteenth century unrelated string of poems composed by Mustafa, cadi of Paphos in Cyprus, who had been captured at sea by a French privateer and brought to Malta where he spent more than two years living as a slave. When he received his ransom money, he returned to Constantinople as a free man. During that time, he wrote a series of qasidas in Ottoman Turkish bewailing his ill-luck and pleading for a rapid resolution of his fortunes. The document bears an uncanny resemblance to the much earlier Byzantine Greek poem to which it is wholly unconnected. References in G. Wettinger, Slavery in the Islands of Malta and Gozo ca. 1000-1812, Malta 2000, 87, 177; W. Schmucker, Die Maltesischen gefangenschaftserinnerungen eines Turkischen kadi von 1599’, Archivum Ottomanicum, ii (1970).

38 But no doubt some were descended from Muslims captured by privateers based on Malta on accepting baptism, perhaps years after their capture.

39 Michele Amari, Biblioteca Arabo-Sicula, Edizioni Dafni, 1982, vol. primo, 55, 75.

40 . Text in Monumenta Germaniae Historica: Scriptores, xxi, Hanover, 1869, 236.

41 Charlotte Schroth-Kohler, Theo Kolzer, Herbert Zielinski, ‘Zwei staufische Diplome fur Malta aus den Jahren 1198 und 1212’, Deutsches Archiv fur Erforschungen des Mittelalters namens der Monumenta Germaniae Historica, Deutssches Archiv xxxiii/ii, 501-521.

42 See my Kliem Malti Qadim, Malta 2006, for the various ways the Maltese language was referred to by the Maltese themselves in the fifteenth and later centuries.

43 Of course one has first to discard the false lists drawn up centuries ago based on the inaccurate identification of Malta with the Calabrian town of Mileto: B. Fiorini, ‘Il comm. Abela e la cronologia episcopale di Malta’, G.F.Abela. Essays in his honour…, Malta, 1961, 82-110; A.T. Luttrell, Studies on Malta before the Knights, 33.

44 H. Takayama, The Administration of the Norman Kingdom of Sicily, Leiden 1993, 118.

45 H. Bresc, ‘Malta dopo il Vespro Siciliano’, doc. 1, will of Guglielmo da Malta, 8 February1299, Melita Historica, vi, no. 3, 1974, 317-20.

46 This was just the time that Ibn Hauqal was describing the swarms of schoolmasters in three hundred Palermitan schools who were busy teaching their children how to speak correctly in Arabic much to the scorn of Ibn Hauqal himself who could not abide their innumerable solecisms of language.