Cor prudens possidebit scientiam
The following is an updated and revised summary of my journal article on Fraxinetum which appeared in the UCLA Journal Comitatus in 2010 (the full article and the footnotes can be accessed here: http://www.academia.edu/3537846/Fraxinetum_An_Islamic_Frontier_State_in_Tenth_Century_Provence)
According to Liutprand (d. 972), the bishop of Cremona, the history of Muslim Fraxinetum began around 887, when a small vessel carrying about twenty Andalusi sailors landed on the Provençal coast near the modern town of St. Tropez. The Andalusis forcibly seized the neighboring settlement of Freinet, and on the mountain above the town proceeded to occupy the fort, which had been called Fraxinetum since Roman times. The subsequent fortress-city which they established was highly defensible and practically impenetrable, protected on one side by the sea from where the Andalusis drew their reinforcements, and on the other by large forests of thorny trees. Consequently, the fort could only be accessed through a single, narrow path leading up the mountain. Contemporary Latin authors, namely Liutprand of Cremona and the anonymous author of the Life of Beuve of Noyers, emphasize the Iberian origin of the raiders, but differ in naming them; Liutprand calls them “saraceni,” whereas the author of the Life of Beuve refers to them as “hispanicolae.” Tenth-century Arab geographers, especially Muhammad Ibn Ḥawqal in his Surat al-Arḍ (977) and al-Istakhri in his Kitab al-Masalik wa al-Mamalik (951), refer to the fortified port of Fraxinetum as Jabal al-Qilal (“Mount of Lumber/Timber”) and describe it as a vast mountainous region blessed with rivers/streams and fertile soil that takes two days to cross. Ibn Ḥawqal, like Liutprand, emphasizes the virtual impenetrability of the fortress and specifies that it was only accessible through one route on the side of the mountain. He also adds that it was dependent on the Umayyads of Cordova, as implied by his cartographic representation of Fraxinetum as an island at the mouth of the Rhone River and located close to the Iberian Peninsula, similar to the Balearic Islands.
(Remains of the fortress of Fraxinetum over the village of La Garde-Freinet)
Shortly after their establishment at Fraxinetum, the Andalusis called upon their brethren in Iberia and the Balearics to join them; about one hundred warriors answered this call, some encouraged by their religious zeal and others by the prospect of wealth from raids. Within two decades of their arrival, the Andalusis managed to subdue Provence in its entirety and even conducted raids as far as western Italy, where they occupied Acqui and threatened the Abbey of Novalesa in 906. Despite their relatively small numbers, the Muslims conquered the land with relative ease due to the divisions and internecine struggles that had characterized Provence since the disintegration of the Carolingian Empire. Consequently, they did not meet any significant resistance from the Provençals. By 939 the Andalusis had managed to cross the Alps and raided what is today northern Italy as well as southern Switzerland, where they attacked the renowned monastery of St. Gall. They established numerous fortresses—which Latin chroniclers in the raided regions all called Fraxinetum or some variation of the name (Frassineto, Frascendello, etc.)—thus facilitating their domination of Provence and the Rhone Valley. From their principal base at Fraxinetum, the Muslims extended their raids into Alemannia and Rhaetia in the North, Grenoble in the West, and Lombardy in the East. Even though Provence was under the nominal control of the Andalusis, the local administrative (and religious) infrastructure was left intact, so that most Provençal towns were relatively self-governing, provided they paid a tribute or tax to Fraxinetum.
Following the simultaneous destruction of the important Provençal port of Fréjus, the sack of Genoa in 935 (by North African and Sicilian Muslims), and the extension of raids beyond the Alps, Hugh of Arles, King of Italy, resolved to act against the Muslims of Fraxinetum. In 941 he summoned a fleet from the Byzantine Emperor, Romanus Lecapenus (r. 920-944), in order to assault the fortress both by land and by sea, hoping to crush Fraxinetum and break the power of the Andalusis in the trans-Alpine region. At a critical juncture during the two-pronged attack, when Fraxinetum was about to fall to his forces, Hugh decided to halt the offensive and form an alliance with the Muslims. He opted for this change in strategy because he had received word that his rival for the Italian crown, Berengar of Ivrea, intended to cross the Alps with reinforcements from Saxony and invade Italy. Hugh reached an agreement with the Andalusis whereby they would occupy and control the Alpine passes, effectively closing the connection between France and Italy, and thus bar any hostile armies from reaching his kingdom. The Muslims maintained this agreement, as it allowed them to acquire vast amounts of booty through controlling the movement of soldiers and pilgrims traveling through the Alps between Francia and Italy.
In the short term, Hugh of Arles was heavily criticized for his actions by his contemporaries, including Liutprand of Cremona. More significantly, the realpolitik behind his decision to allow Fraxinetum to survive would have dramatic long-term consequences. It was during the period of its control of the Alpine passes that Fraxinetum reached its peak, and the raids by the Andalusis became the most destructive and deadly; according to Latin chroniclers, the Muslims sacked numerous monasteries and indiscriminately killed hundreds of pilgrims on their way to Rome. It was also during this time that Fraxinetum hosted a number of rebels and renegades, notably Adalbert of Italy, son of Berengar of Ivrea, from the neighboring kingdoms, thereby drawing additional hostility from the local and regional authorities in Germania, Francia, and Italy. The Muslims built a long line of defensive fortresses along the mountain range in order to consolidate their control of the passes and to increase the scope of their attacks. However, their confidence—largely a product of the lack of resistance and the repeated successes of their activities in Provence and Piedmont—would also prove to be their downfall. Their first major miscalculation was conducting raids into the Upper Rhine Valley.
(Otto I, Holy Roman Emperor, Manuscriptum Mediolanense [Chronicle of Bishop Otto of Freising], ca. 1200. Milan, Biblioteca Ambrosiana, Cod. S.P.48)
This was the territory of Otto I, who decided to appeal to the Caliph of al-Andalus, Abdurrahman III, whom Otto believed to have the authority to check their activities. This episode of diplomacy between the two most powerful sovereigns of western Europe, Otto I and Abdurrahman III, demonstrates that the significance of the Muslims of Fraxinetum went beyond the limited scope of their raids, and actually had the potential to upset the balance of power and status quo between Muslim and Christian powers in western Europe. The second error committed by the Andalusis, which subsequently triggered a series of events leading to the eventual demise of Fraxinetum, was the capture of Maiolus (d. 994), the Abbot of Cluny—considered to be a living saint by many of the counts, dukes, and kings of western Europe—while he crossed the Alps in 972. His capture provided a unifying factor in the struggle of the lords of Provence against the Andalusis and prompted them to respond collectively to the threat emanating from Fraxinetum. Following his ransom and subsequent release, Maiolus rallied a coalition of nobles in a semi-crusade aimed at removing the Andalusis from Francia.
The expedition was led by Guillaume I of Provence, but aristocrats from northern Italy, Provence, and Septimania also took part. The Frankish forces met the Muslims at Tourtour, in upper Provence, during the summer of 972 and destroyed their ranks before moving on to the main base at Fraxinetum, which did not receive any reinforcements from al-Andalus and, following a short yet intensive siege, fell in late 972 (although a number of sources place the date as late as 990). After the destruction of Fraxinetum, the Muslim inhabitants of Provence—combatants and non-combatants alike—were either killed, enslaved, or exiled, and the lands they had controlled were partitioned among the many lords who had taken part in the expedition to expel them from Provence. This victory of the Provençals over the Andalusis of Fraxinetum effectively ended Muslim control over southern France almost 240 years after Charles Martel’s defeat of Abdurrahman al-Ghafiqi at the Battle of Toursin 732.
Fraxinetum: An Islamic Frontier State in 10th-Century Provence
Traditionally, the establishment of both al-Khandaq in Crete (which I have written about elsewhere on this blog) and Jabal al-Qilal in Provence by Andalusi Muslims has been understood as a manifestation of the phenomenon of what many scholars have termed “Saracen piracy.” Although modern scholars still occasionally apply this definition, several historians of both Fraxinetum and Crete have tended to be more cautious about how they term the movement that facilitated the arrival of the Muslims in the Aegean and Provence. Consequently, the historiography of the Muslim Mediterranean in the early Middle Ages has branched into two lines of argumentation, with more recent scholarship challenging the traditional interpretation of Muslim bases in the Mediterranean as “pirate nests.” The debate has focused on defining the exact nature of Muslim maritime expansion, and has been split on characterizing this phenomenon either as part of the trend of “piracy” or as part of the larger, more complex context of jihad (holy war) on the Mediterranean frontier with Christendom in the early Middle Ages
Although the debate over whether Fraxinetum should be characterized as a frontier state or a pirate base may appear to be a mere argument over semantics, there is some value in thinking carefully about how we should identify the phenomenon that prompted a group of Andalusis to dominate much of southern Provence and the Alpine passes for nearly a century. Scholars of early medieval Muslim maritime expansion have distinguished between organized, dynastic-sponsored ventures, such as the conquest of Sicily by the Aghlabids, and independent, more spontaneous campaigns, including the establishment of Muslim bases at Fraxinetum, Crete, and southern Italy. This distinction has formed the grounds for the debate about whether the actions of autonomous Muslim sailors in the Mediterranean constituted piracy or corresponded to jihad. While the official nature of the conquest of Sicily by the Aghlabids contrasts sharply with the unorganized and spontaneous seizure of towns by Muslims in southern Italy, it is less evident that the general motivating factor of both ventures was entirely different. The use of the word “official” in this context is potentially problematic and needs to be clarified.
“Official” is meant to denote the endorsement, participation, legitimacy, and aid granted by the dominant Muslim dynasties (the Aghlabids, the Abbasids, the Umayyads, the Tulunids, and the Fatimids) to the independent flotillas operating in the Mediterranean. Historically, the authority to declare jihad rested with the amir or khalifa, who sponsored and even participated in such expeditions, which were meant to bestow the ruler with and confirm his Islamic legitimacy, given that jihad was viewed as an obligation. Furthermore, such “official” calls for jihad were launched from fortified coastal/frontier towns, known as ribaṭs, which were usually established, garrisoned, and supported by the troops of major Muslim dynasties. While several campaigns were known to be officially sponsored and labeled as “jihad” by an Islamic centralized authority, as was the case with the conquest of Sicily in 827 and the campaigns of the Abbasids against Byzantium in the early ninth century, it is highly ambiguous whether other expeditions, such as the invasion of Crete, were considered as such, since they were not initiated by any major dynasty. Nevertheless, because these campaigns/incursions were also carried out against non-Muslim territory, regardless of the fact that they were not organized in the traditional sense or coordinated by a centralized authority, they can also be considered jihad according to tenth-century Islamic jurisprudence.
(Ribat, Sousse, Tunisia)
As Deborah Tor has recently shown, beginning in the late eighth and early ninth century a new phenomenon began to take shape in the Islamic world, namely that of “privatized jihad.” This involved the transference of religious leadership of jihad from the Caliph to the mutaṭawwi‘a, or ghazis, volunteer warriors for the faith, and hence the transformation of jihad from “centrally-directed state campaigns to independent, non-governmentally controlled smaller-scale raids.” Although lacking any dynastic endorsement or legitimacy acquired directly from the Caliph, these campaigns were viewed by their participants and other observers in the Islamic world as an exercise of the religious duty of jihad, given that the objective was not merely to plunder or raid, but to defend (or expand) the frontiers of the Islamic world by establishing ribaṭ outposts garrisoned by ghazi warriors.. It is in this latter sense that the establishment of Fraxinetum and the conquest of Crete by the Andalusis differed from the more recognizable jihad launched by Muslim dynasties such as the Abbasids or the Aghlabids.
Although many historians have not distinguished between piracy and jihad, especially within the context of the tenth-century Mediterranean, there is a major difference. In the case of the Muslims of Fraxinetum, it is clear that their raids were aimed largely (if not solely) against non-Muslim shipping in the western Mediterranean and non-Muslim targets in Provence and Piedmont (monasteries, villages, etc.). Therefore, it is more appropriate to speak of their activities/raids within the context of guerre de course or jihad, rather than within the vague context of “piracy.” This argument in favor of viewing the Muslims of Fraxinetum as ghazis engaged in jihad is reinforced by Ibn Hawqal’s characterization of them as “mujahideen” (holy warriors; literally “those who do jihad”), and by al-Istakhri’s description of them as constantly engaged in “holy war” with the “infidel Franks.” This does not mean that “jihad” did not encompass piratical activities such as looting, enslavement, and murder, but only that there was a broader ideological context in which such activities took place.
The creation of the Spanish March in 795 greatly reduced the prospects of the Andalusi Muslims extending their conquests beyond the Pyrenees and induced them to turn to maritime activity. Although there is evidence of land incursions into Aquitaine and Septimania as late as 931, when an Andalusi army sacked Toulouse, land-based campaigns against Francia launched from al-Andalus were quite rare. Fraxinetum, it would seem, provided an outpost to the Andalusis from which they could continue to wage jihad against the Franks. Beyond this ideological justification provided for their activities by the contemporary Muslim sources, Fraxinetum was a major source of wealth in the form of slaves and timber.
It would perhaps be most useful to think of the Muslim sea-raiders in the early medieval Mediterranean as ghazis, a term which allows scholars to characterize their activities without downplaying either the strategic/economic dimension of the raids or lose sight of the ideological framework in which they were taking place. The term “piracy,” while certainly an accurate characterization of some of the activities of the Muslims in places such as Fraxinetum and Crete, is largely unhelpful because it fails to convey the complexity of such frontier establishments. Moreover, this terminology is largely a product of an uncritical reading of the primary sources, which has led to the replication of many of the arguments and assumptions of the medieval Latin authors within modern scholarship. It is undeniable, based on several contemporary accounts, both Christian and Muslim, that the activities of the Muslims in Provence included significant destructive attacks against the regions adjacent to their main base. However, they also engaged in other, less hostile activities, such as trade, agriculture, and industry. The image presented by Latin chroniclers is largely that of Muslim Arab raiders aimlessly pillaging and looting the Provencal landscape. This picture, while certainly reflective of the perceptions of Christian chroniclers, obscures the other reality, represented by the Arabic chroniclers, of a Muslim frontier region also engaged in peaceful activities and fulfilling a larger political, military, and even religious function.
By looking at how the contemporary Arabic sources portray the Andalusis in Provence, and juxtaposing their representation with the portrayal in the traditional set of Latin texts, it becomes clear why modern scholars have described the Muslims of Fraxinetum in different ways. The image presented by the Latin chronicles and that provided by the Arabic sources present a contradictory view of the Muslim settlement and the activities of its inhabitants. Whereas the Latin sources almost unanimously describe the Muslims in unflattering terms, such as “raiders,” “perfidious thieves,” “barbarians,” or “pirates,” the Muslim sources refer to “steadfast souls” and “holy warriors.” While they are informative and descriptive with respect to the raids of the Andalusis in Provence, the Latin documents are less useful in describing the Muslims themselves or their base at Fraxinetum. Similarly, there is no reason to take the Muslim sources at their word when they describe the Andalusis in Provence in such flattering terms. However, considering this textual evidence allows us to counterbalance the one-sided testimony of the Latin chronicles and allows us to appreciate the complexity of Fraxinetum.
It is understandable that the Latin chroniclers portrayed the Andalusi Muslims negatively due to the devastation that accompanied their raids. It is also evident that, apart from experiencing the raids first hand, the ecclesiastical chroniclers, from whom most information regarding Fraxinetum has reached modern historians, had little direct contact with the Muslims in their own environment, and had only a vague idea of their other, less violent activities. Consequently, the natural outcome of scholars giving more weight to the Latin sources, which are more plentiful and presently far more accessible, is that the representation of the Muslims of Provence by the Frankish chroniclers becomes more influential and gradually more acceptable in scholarly discourse about Fraxinetum. In other words, the image represented by writers such as Ekkehard of St. Gall, Liutprand of Cremona, and the author of the Chronicle of Novalesa—who all portray Andalusi raiders pillaging Provence, Piedmont, and the Alps and aimlessly sacking monasteries—has become the dominant depiction of the Muslims of Fraxinetum. As a result of the frequency of such accounts, many historians discuss Fraxinetum solely in terms of piracy and destructive raids, rather than questioning the veracity of the accounts or conceptualizing Fraxinetum in ways other than as a pirate base. Some scholars have even referred to the tenacious persistence of this image as the “black legend” (legende noire).
A clearer view of the extent to which the Muslim raiders of the western Mediterranean were characterized as “pirates” and “barbarians” by the Latin chroniclers can be gained by examining some brief excerpts from their works. Early Latin references to independent Muslim sea-raiders from Iberia appear in the Royal Frankish Annals, the Annals of Fulda, and Einhard’s Vita Karoli Magni, which discuss the activities of the Andalusis against the Frankish and Italian coasts during the late eighth and early ninth centuries. For the years 798 and 799, the Royal Frankish Annals describe how the “Mauri et Sarraceni” plundered the Balearic Islands in what the Annals describe as “praedonum incursione,” which can be translated as “piracy,” while the Annals of Fulda refer to “mauri piratae” and “Mauris praedatum” in the years 798 and 808 respectively. Similarly, Einhard, in his Vita Karoli Magni, explains how the Frankish and Italian coast, from Narbonne to Rome, was fortified by Charlemagne due to the fact that it was constantly ravaged by “Mauros piracticam.” These references clearly show that the actions of the Muslim sea-raiders in the early part of the ninth century were characterized as piracy by the Latin chroniclers.
Contemporary Latin chroniclers cited the resurgence in Muslim sea-raids against the Frankish coast in the mid-ninth century as a continuation of this earlier period of “piracy.” The Annals of St. Bertin, for example, narrate the sack of Marseille in 838 by “saracenorum piratae,” and describe the Arab raid on Arles in 842 by “maurorum piratae.” When discussing the Andalusi Muslim attacks on the Provençal coast, the author of the Life of Beuve of Noyers refers to “paganorum piratarum,” and decries how “the Iberians laid waste to Provence” (Hispanicolae devastant Provinciam). In the descriptions of the activities of the Muslims by the Latin chroniclers, a heavy emphasis is placed on the devastative (maurorum devastant, mauri irruenes), the destructive (destruendos saracenos, paganorum destructum, depopulantes terram), and the deadly (paganorum adnihilatam, saraceni trucidatur, callidus exactor) aspects of their incursions. The way in which the Muslims and their raids are described by Frankish chroniclers from the mid-ninth century onwards is more detailed and elaborate than how they were described in the earlier part of the century by the Royal Frankish Annals and by Einhard. In addition to maintaining continuity in the label they gave to Muslims who occupied Fraxinetum at the end of the ninth century (“Saracen pirates”), the chroniclers described their raids as being more devastating and destructive than earlier “pirate” incursions.
Moreover, the Latin chroniclers not only describe the Muslims and their activities within the context of piracy, but also present their incursions in terms similar to the raids of the Norsemen (piratae Danorum) and Magyars (Ungariorum gens, Hungari), who were concurrently plaguing western Europe in the ninth and tenth centuries. The major source of information about the establishment of Fraxinetum and the political history of the Muslims there is Liutprand, the bishop of Cremona, writing in the mid-tenth century. Liutprand and many other Latin chroniclers of the time portray the Muslims of Fraxinetum as infidel “pirates,” who were threatening the very heart of Christendom, and who raided, desecrated, and burned monasteries because of their “pagan” disdain for Christianity.
During the ninth and tenth centuries, the areas encompassing modern-day France, Switzerland, Italy, and Germany were being assaulted by three main groups: the Vikings, the Magyars, and the Andalusi/North African Muslims. As the main external threat to Latin Christendom, they were depicted almost monolithically as invading barbarian hordes, ravaging western Europe from three different directions. More specifically, the invasions of the Vikings and the Muslims were both represented as manifestations of God’s wrath against the perceived impiety of the Frankish kingdom. The sensitivity of the Latin chroniclers to the Viking and Muslim destruction of monasteries is especially apparent, and seems to inform their theological perspective on the invasions, since they viewed such attacks not only as manifestations of divine displeasure, but also as reflective of “pagan” hatred for Christianity. By recognizing that the Latin chroniclers described both the Vikings and the Muslims in similar terms, as God’s instrument for chastising faithless Christians for their sins, it becomes less puzzling why the Muslims of Fraxinetum were depicted as faceless invaders. The characterization of their raids in the Annals of St. Victor of Marseille as God’s divine rod of chastisement against the Christians (“Deus flagellare vellet populum christianum per seviciam paganorum, gens barbaric in regno Provence irruenes”), and the lamentation by Alcuin of York of the Viking devastations in Northumbria using the words of an Old Testament prophet (“Uae genti peccatrici, populo graui iniquitate, filiis sceleratis; derlinquunt Deum et blasphemauerunt sanctum saluatorem mundi in sceleribus suis”– Isaiah 1:4), are both demonstrative of the fact that the Muslim raids in Provence were viewed in the same light as the depredations of the Vikings in the North, i.e., as punishment for the sins of Christendom.
Hence, the representation of the Muslims as one-dimensional “pirates” needs to be understood within the context of the simultaneous Viking incursions, which made the Latin Christians feel particularly vulnerable to outside attacks, and which they depicted almost unanimously as “piracy.” On another level, the characterization of the Vikings and the Muslims as instruments of God’s anger shows that the Latin chroniclers viewed the invading flotillas as a military threat that would eventually subside, rather than as permanent ideological or religious rivals. Contextualizing the Muslim invasion in this way provides a clearer indication of why the Latin Christian chroniclers were, understandably, more concerned with outlining the devastation and horrors associated with the raids, than with attempting to identify and describe with precision the Muslims of Fraxinetum.
(The area surrounding Fraxinetum is today known as Massif des Maures, an indication of the collective memory of the presence of the Andalusis, or Mauri as they were known in Latin)
In order to more fully appreciate the value of the Latin chronicles as useful sources about Fraxinetum, the polemical content of the sources needs to be separated from the informative and substantive evidence that they provide. Although all Latin documents agree that the Muslims had a negative impact on Provence, and that their base at Fraxinetum was only contributing to the regional turmoil, they differ in describing the nature of the Muslim settlement there. The Chronicle of Novalesa, for example, describes Fraxinetum as “a place on the coast near Arles,” Liutprand of Cremona describes it as “a village between Italy and Provence,” Ekkehard’s Casus S. Galli refers to a “valley,” and another contemporary author, Sigebert of Gembloux, refers to it as a “castle.” Moreover, the fortress itself is described variously as a castrum (fortified outpost), a villa (rural agricultural dwelling), and an oppidum (town).
These divergent terms each carry a completely different meaning; depending on whether Fraxinetum is interpreted as being a castrum or an oppidum, the implications for the question of whether it was a raiding outpost or a civic entity on the frontier between the Islamic world and Christendom are very great. Another anonymous Latin chronicler describes how some Muslims had settled and were living unarmedamong the local Provençal townsfolk in the vicinity of Fraxinetum, and that intermarriage with indigenous Christians had become common practice for the Muslims in the area, a fact that is greatly revealing about the nature of Muslim settlement in the region, and raises questions about their social and ethnic identity. It was not uncommon for significant numbers of Christians to participate in the many raids of Muslims in the Mediterranean during the early Middle Ages, especially since the Latin sources testify to the presence of “renegades” (either Christian converts to Islam or simply rebels against local authority) in Fraxinetum. This further encourages us to think of the frontier as representing a far more fluid political and social reality, encompassing a vast array of activities and possibilities.
In order to identify whether the Muslim settlement at Fraxinetum was limited to a castle, town, or even a region, it is important to take into account the contemporary Arabic sources. Unlike the Latin chronicles, the representation of Fraxinetum in the Muslim sources is far more sympathetic. Most information from a Muslim view comes from Arabic geographical works. Unfortunately, there are only three surviving contemporary Arabic sources that make mention of Fraxinetum, and, unlike the detailed Latin sources, they provide scarce information about it. The two main sources are Al-Istakhri’s Kitab al-Masalik wa al-Mamalik, and Ibn Hawqal’s Surat al-Ard. Al-Istakhri, who was writing during the mid-tenth century, describes Jabal al-Qilal (Fraxinetum) as a mountainous country, which the Muslims inhabited and developed to the dismay of the Franks, with many streams and rivers, and explains that it took two days to cross it by foot. Istakhri also notes that the region surrounding Fraxinetum was a neglected area, and that the arrival of the Muslims and their subsequent settlement there led to its prosperity. Furthermore, he states that the Muslims were in a constant struggle with the local authorities (“Frankish infidels”) of the land, and emphasizes the strategic importance of the base by exclaiming that “if not for this mountain [i.e., Fraxinetum] the lands of Islam would be in grave danger!” This latter statement implies that he, a scholar from the eastern Islamic lands, attaches to an isolated fortress-city located in the western Mediterranean, suggests that he viewed Fraxinetum as a ribat, or frontier state, similar to other frontier fortresses on the frontier between Christendom and Islam in the Mediterranean.
Ibn Hawqal elaborates on al-Istakhri’s writing by describing Fraxinetum as the main stronghold of the mujahideen (holy warriors) who were victorious in the land of the Franks, and observes that it was very productive agriculturally due largely to the fertile soil, multitude of land, and water currents that flowed there. He also re-emphasizes Istakhri’s point that it was the arrival of the Muslims which caused the region to flourish. He explains that for security reasons the Muslims fortified the mountain above the settlement by building a fortress, which was only accessible by one narrow path. It appears that Ibn Hawqal viewed Fraxinetum primarily as a viable agricultural settlement housing frontier warriors, whose military character was necessitated by the hostile environment and security concerns, not to mention their supposed raison d’être in Provence: jihad. His description, similar to Istakhri’s, corresponds very closely to the modern scholarly understanding of a medieval Islamic ribat/frontier state, and allows scholars to draw comparisons with the frontier states located on the border region between the Abbasids and Byzantium in Anatolia in the Near East or even the later ghazi Turkish emirates on the frontiers of the Byzantine and Serbian empires in the fourteenth century.
Another tenth-century geographical work, Hudud al-Alam, written by an anonymous Persian traveler and dated to 982, describes Fraxinetum as an inhabitable mountain on the Mediterranean Sea, which is in close proximity to the Italian Peninsula. The author of Hudud al-Alam also adds that “[to the] west of Jabal al-Qilal is a mountain, whose summit is so high that it cannot be reached, and from [this region] comes game, timber, and fuel,” a rare indication of the economic importance of the region of Provence that the Muslims occupied. Despite the brevity of these three geographical sources, two of which were written in Persian and one in Arabic, they are extremely helpful in describing the nature of the Muslim presence in Provence. Due to the scarcity of the Arabic sources, it is far more difficult to conceptualize Fraxinetum in terms of a cultural or economic center, as recent studies have shown for Andalusi Crete, but archaeological evidence has shown that the Muslim presence in Provence was more multi-faceted than has previously been thought.
Scholars have recently elaborated on the nature of the agricultural, semi-industrial, and other, non-hostile ventures of the Muslims of Fraxinetum. The argument has been made that the introduction of buckwheat, originating from Iran, into Provence was initiated by the Andalusis, a fact possibly indicated by etymology, since it is referred to as “blé sarrasin” in the Provençal dialect of the French language. In addition to the introduction of buckwheat, it has been speculated, based on local legend, that cork tree cultivation and the method of turning pine-resin into pine-tar for the strengthening of wooden ships was begun by the Muslims; indeed, the Arabic name for Fraxinetum, “Jabal al-Qilal,” is a reference to this practice. Moreover, the presence of distinctly Arab and Berber agricultural and pastoral practices can also be inferred from the fact that certain species of goats native to North Africa are herded in Provence, while the raising of pigs is rare, a detail that can be attributed to the period of the Andalusi Muslim presence. The archaeological remains of pottery, metallurgy (mines commonly referred to as le trou de Sarrasins and gallerie sarrasine have been identified near Grimaud, La Garde-Freinet, and Plan de Tour), the manufacture of weapons (forges have been excavated in Tende and La Feriére), and forestry have also been cited as indicative that non-military activities were relatively widespread at Fraxinetum, and that there were possibly artisans and other skilled Muslims among the warriors there.
(The village of Ramatuelle, known in the Middle Ages as Ramatuella, probably deriving from the Arabic Rahmatu Allah meaning “God’s Mercy,” was in the heartland of the Andalusi-controlled region of Provence)
It can be inferred from the works of Ibn Hawqal and al-Istakhri that the Muslims were not only “mujahideen” who held Fraxinetum as a defensible, frontier outpost for Islam, but also engaged in other activities such as irrigated farming and trade with the rest of the Islamic world, exporting their produce as well as timber and fuel. The impression given from Hudud al-Alam is that of an economically viable region, from where timber and combustible fuel could be procured. It is therefore apparent that the Muslim sources depict Fraxinetum not as a mere raiding outpost but as a defensible, self-sustaining Andalusi political entity in the midst of a largely-hostile Frankish/Christian population. Muslim sources depict the region of southern Provence as unproductive and undeveloped until the arrival of the Muslims, who reinvigorated the region economically and caused it to flourish. Therefore, the statements by Syrus and Liutprand claiming that “the Saracens transformed the realm into a desert” is strongly contradicted by the Arabic sources, which boast how the Muslims caused the unproductive and empty landscape of southern Provence to thrive.
Although neither the Latin nor the Arabic sources give a balanced perspective, when read in conjunction with each other they can be greatly useful in providing a clearer view of Fraxinetum. It can safely be asserted, based on the existing evidence described above, that Fraxinetum was a fortified outpost (castrum) populated by Muslim frontier warriors who were in constant struggle (jihad) with the Franks in Provence, and whose activities included not only destructive raids that tormented the Provencal Christians, but also more peaceful activities such as farming and trade. Therefore, it is useful to think of Fraxinetum as an Islamic frontier fortress, which had a military disposition and a socio-economic element. Fraxinetum was settled by the Andalusis and developed into a commercial and military center with the intention of using the base as an outpost with a clearly defined purpose; in fact, Fraxinetum housed hundreds (possibly thousands) of ghazi warriors and possessed a relatively large fleet consisting of an estimated 12 to 15 medium-sized vessels. Although neither the Arabic or Latin sources provide extensive evidence in this regard, it is also plausible that Fraxinetum was a major source of slaves as a result of its relationship with the slave-trading center of Verdun in central Francia during this period. Many of these slaves (originating from Central Europe) would have probably ended up in al-Andalus and formed the basis of a new military and social class of Saqaliba. However, in the absence of more authoritative evidence, the specific role of Fraxinetum in this Western Mediterranean slave trade must remain largely conjectural.
Relationship between Fraxinetum and al-Andalus
The relationship between Fraxinetum and al-Andalus is particularly problematic because it is unclear whether Fraxinetum was as an autonomous Muslim frontier state and staging ground for raids against the Franks, or if it was in fact a thaghr/frontier region of al-Andalus that acknowledged Umayyad political authority. On this question, we are largely in the realm of conjecture and it is impossible to determine with any certainty what the relationship between al-Andalus and Fraxinetum actually was. Many of the Latin sources make a direct connection between the Muslims of Fraxinetum and Umayyad Spain by describing the Andalusis as natives of Iberia or tributary to the rulers of al-Andalus, and explain that they received reinforcements from the Caliphate of Cordoba. The Life of Beuve of Noyers, for example, refers to the Muslims of Fraxinetum as hispanicolae, while Liutprand of Cremona describes how they were tributari regis Abdurrahman, which can be understood to mean vassals or subjects of the Caliph of Spain, Abdurrahman III (r. 912-961). This is the strongest piece of contemporary evidence that directly associates the Muslims of Fraxinetum with the Umayyad caliph.
The association of the Muslims of Fraxinetum with al-Andalus also becomes evident from the Latin sources that describe the raiders as mauri, a designation generally understood to mean Iberian Muslim by the ninth and tenth centuries. Perhaps the most compelling contemporary evidence derived from Latin sources suggesting that Fraxinetum was affiliated with, although not necessarily dependent on, al-Andalus is the diplomatic mission sent by Otto I to Abdurrahman III in 953, demanding that Cordoba halt its support of the Muslims of Fraxinetum, whose raids extended as far north as the Rhine valley by the mid-tenth century. Although Abdurrahman III assured Otto’s emissary, John of Gorze, that he was not supporting the Muslims in southern Provence, subsequently there was a recognizable reduction in the activities of the Muslims of Fraxinetum, whose base fell in 972 to a coalition of Provençal knights without receiving aid from al-Andalus. It is plausible that Otto’s embassy had a direct impact on Abdurrahman III with regard to his policy towards Fraxinetum, although without further evidence this cannot be definitively asserted.
While some scholars have tended to gloss over the nature of the relationship between al-Andalus and Fraxinetum, the French historian Philippe Sénac has attempted to tackle the question directly. His contention has been that the main commodity which Fraxinetum had in abundance, and for which there was a high demand in al-Andalus, was timber. Fraxinetum has thus been hypothesized as a possible outlet of this scarce, but essential, resource for al-Andalus, which was in the process of constructing a navy. The theory that Abdurrahman III needed timber for his fleet, and that Andalusi Muslim ghazis (namely those from the largely autonomous eastern Iberian coast and Fraxinetum) provided him with it, touches upon a central thematic question regarding the relationship between Fraxinetum and al-Andalus. Apparently, official Umayyad endorsement of the base at Fraxinetum in exchange for resources such as timber would have greatly endangered the status quo with Christian powers in western Europe. However, since areas in Provence where timber could be procured were of vital importance to Abdurrahman III, the Caliph supported Fraxinetum logistically. Yet, at a time when Andalusi-Italian trade was flourishing at the expense his Fatimid rivals in North Africa and Sicily, he did not officially provide military or economic aid to the settlement, for fear of an economic and military backlash from western European powers, especially the Ottonians and the Italian city-states.
It is plausible that in exchange for providing the Caliphate of Cordoba with timber, the Muslims of Fraxinetum received a certain degree of endorsement, along with supplies and reinforcements, from al-Andalus. According to the eleventh-century Andalusi chronicler Ibn Hayyan al-Qurtubi, the Umayyad admiral Abd al-Malik ibn Sa‘id ibn Abi Hamama, commanding a fleet of forty ships, led an expedition against the Frankish coast in 935 and possibly interacted with the Muslims of Fraxinetum, coordinating the attack in conjunction with them. Although this is not definitive evidence of Umayyad military support for Fraxinetum, it raises the question about the degree to which al-Andalus actively endorsed, encouraged, and sponsored the Muslims stationed there. In fact, as seen above, the anonymous author of Hudud al-Alam explicitly states that Fraxinetum was an important site from which timber (and fuel) was exported to the rest of the western Islamic world. In support of this claim, several archaeological digs in southern Provence near the site of Fraxinetum have recently uncovered Arab tools, such as axes, nails, saws, chisels, and hammers, that would have been used to exploit timber resources. Furthermore, marine archaeologists excavating the bay of St. Tropez have discovered several Arab ships with compartments presumably for the transportation of such materials as timber and other items, notably ceramics.
Other evidence suggests that the Andalusis of Fraxinetum excelled at the development of the use of cork oak and pine resin, materials used for caulking ships. This is also supported by the fact that the Provençal word for tar, “quitran,” is derived from a similar Arabic word, “qatran,” and by the less-apparent detail that the southern Provençals are still famed for their cork-oak industry, the origins of which they attribute to the “Saracens.” Although much more work needs to be done in order to ascertain whether Fraxinetum was indeed an outpost of al-Andalus from where timber was procured, it is at least apparent that this commodity was a factor in the relationship between the Caliphate of Cordoba and Fraxinetum. In addition to the timber connection, tenth-century fragments of ceramics have been excavated off the coast of Provence bearing identical patterns to ceramics in southern Iberia, especially to patterns found in the town of Almeria-Pechina, indicating another direct link between al-Andalus and Fraxinetum.
The Arabic sources also support the notion that Fraxinetum was directly affiliated with al-Andalus. Contemporary geographers, namely Ibn Ḥawqal and al-Istakhri, refer to Fraxinetum as part of Islamic Iberia and emphasize that it was an extension of Andalusi Umayyad sovereignty into southern Francia. This is highlighted by Ibn Ḥawqal’s depiction of Fraxinetum as an island at the mouth of the Rhone River near al-Andalus, similar to how he portrays the Balearics, in his map of the western Mediterranean. This cartographic depiction has been taken literally by some scholars, who have misinterpreted it to mean that Ibn Ḥawqal and other Muslim geographers visualized Fraxinetum as an island. However, as is the case with many pre-modern maps, it should not be viewed as a strict attempt to depict geographic territory. Rather, the map should be understood as a spatial representation of the reality that Fraxinetum was highly (geographically) isolated from al-Andalus, but simultaneously (politically) affiliated and connected with it.
(10th century Muslim map of the Mediterranean from al-Istakhri, “Kitab al-Masalik wal Mamalik,” Liber Climatum…codicis Gothani..curavit J.H. Moeller. Gotha: libraria Beckeriana, 1839. Fraxinetum [“Jabal al-Qilal’] is represented as a large triangular island near the top of the map) http://www.christies.com/lotfinder/lot/al-istakhri-abu-ishak-al-faresi-1801664-details.aspx?pos=18&intObjectID=1801664&sid=)
Another indication of the relationship that existed between al-Andalus and Fraxinetum is a reference by Ibn Hayyan al-Qurtubi to how in 940/941 Abdurrahman III had copies of a peace treaty with the “Franks,” including Hugh of Italy, made and sent to the governors of Valencia, the Balearics, and to the qa’id of Fraxinetum (mentioned by name as Nasr ibn Ahmad), who were likely the main parties concerned with such a treaty. The use of the word “qa’id” implies the existence of a unique relationship between al-Andalus and Fraxinetum, one different than that which existed between the Balearics and Cordova, since it does not simply denote a “governor” or a civil servant administrating on behalf of the centralized authority. Unlike inland towns, which were governed by a wali (“governor”), port cities were usually under the control of the qa’id, who typically had more powers than the regular governor and was responsible for military as well as civil affairs. That the qa’id of Fraxinetum is mentioned by name in such an important chronicle also suggests that the Muslim presence in Provence was of some importance to al-Andalus, and that the events in Fraxinetum were of some interest to Cordova. “Qa’id,” an Arabic geographical and administrative term, usually designated the commander of a frontier zone, or thaghr. The Arabic word thaghr is understood to mean a virtual no-man’s land studded with fortresses (husun), located on the frontiers of Muslim territory. and used for staging raids into non-Muslim territory.  This form of boundary was typical of al-Andalus, and formed the first line of defense between Muslim-ruled Iberia and its northern Christian neighbors. The thaghr was quite different than other border regions, generically known as hudud, because it typically contained ribats, and its role was primarily military, although several ribats also developed agriculture and industry.
These frontier fortresses were a unique political entity, distinct from the core parts of the kingdom, and were given a special administrative status. They were under the control of a military commander, the qa’id, and their link to the centre of power tended to be much more tenuous than that of other regions of the state. The qa’id fulfilled many of the functions of a regular ruler, and at times could act in defiance of the wishes of the central government. In light of these facts, it is possible, based on the testimonies of authors like Ibn Ḥayyan and Ibn Ḥawqal, to envision Fraxinetum as a thaghr of al-Andalus, inhabited by frontier warriors, or ghazis, rather than as a fully independent and self-sustaining entity. The description of Fraxinetum in the Arabic sources therefore corresponds very closely to the conception of the Muslim frontier-fortress of the ninth and tenth centuries. Although none of the contemporary Arabic sources explicitly claims that it was a frontier-province of al-Andalus, their descriptions of the fortress, its inhabitants, and their activities all point to the fact that Fraxinetum can at least partially be understood within the broader framework of frontier warfare in the early medieval Mediterranean, as opposed to a “pirate base.”
The surviving description by Arabic chroniclers and geographers of the Andalusis engaging in agriculture, utilizing and exporting timber, and other activities in Provence suggests that, due to their relative isolation, they needed to sustain themselves through means other than relying on al-Andalus or raiding. Nevertheless, even if it is accepted that Fraxinetum was a thaghr of the Caliphate of Cordoba, its autonomous nature is implied by the distance between Provence and Iberia (a five day sea-journey from the Valencian coast). This reality was no doubt underscored by the destruction of Fraxinetum’s harbor by Byzantine forces, and the elimination of its entire fleet in 941, ending any connection that it had with Iberia and perhaps influencing the decision by the Muslims there to intensify their raids beyond the Alpine passes.
Due to the lack of evidence, it is highly problematic to label Fraxinetum as an extension of al-Andalus or overemphasize the connections between the Muslims in Provence and the caliph in Cordoba. Indeed, there is every reason to believe that the relationship between al-Andalus and Fraxinetum never extended beyond informal ties due to the weakness of the Umayyad ruling authorities in the late ninth century (as mentioned above, Fraxinetum was probably established by sailors/raiders from the independent port of Pechina-Almeria) and the concern of the Umayyads with internal rebellions and then the Fatimid threat throughout the early-mid tenth century. The establishment of the Shi’i Fatimid dynasty in North Africa around 909, coinciding as it did with major anti-Umayyad rebellions across Iberia, posed a major strategic and ideological challenge to Umayyad-ruled Iberia. This threat from the Fatimids also partially explains why the Umayyads probably had no interest in aiding their co-religionists in Fraxinetum, which had become a liability and a major source of tension, at a time when they were hoping to win allies from Christian powers such as Otto I and the Byzantines against their formidable North African enemies. It should also be recalled that already in 953 Otto I had sent his emissary to Cordoba to protest the raids originating from Fraxinetum. It would certainly not have been in the interests of Abd al-Rahman III to provoke further tensions with Latin Christendom by supporting the Muslims at Fraxinetum.
The declining fortunes of the Umayyads in Iberia, especially following the disastrous psychological and military defeat of forces of ‘Abd al-Rahman III by Ramiro II of León and his allies at Simancas in 939, also provides the immediate context for assessing any possible relationship between Cordoba and Fraxinetum. This battle, which was predicted to end in a swift and decisive victory for the Umayyads, resulted in their crushing defeat and the capture of the personal Qur’an of Abdurrahman III. As a result, the Caliph was compelled to agree to a humiliating peace treaty with León, surrendering large amounts of territory to the Spanish kingdoms, and was forced to reassess the strategic situation in the Iberian Peninsula. He never took the field of battle again and maintained a defensive posture against the Christian kingdoms for the remainder of his career. The loss at Simancas, therefore, likely reduced any Umayyad tendency to support the Muslims of Fraxinetum.
In light of conflicting historical accounts and the paucity of solid evidence, how then should Fraxinetum be viewed? Based on the available sources, it is certainly plausible to think about Fraxinetum as a Muslim frontier state in the early medieval Western Mediterranean. It is clear that the activities of the Muslims in Provence encompassed both violent raids and more peaceful activities, such as agriculture and trade, thus resembling other regions across the Muslim-Christian frontier in the ninth and tenth century such as Crete and Bari. As for the frequency of the concept of jihad that is constantly employed with regard to Fraxinetum in the Arabic sources, it is only analytically useful if it is understood as providing a legitimating framework for the activities of the Muslims in Provence, rather than as an explanation for their activities. The importance of booty, timber and the slave-trade for the polity suggests that Fraxinetum cannot and should not be thought of solely in terms of either ideology or “piracy,” but a complex combination of both categories. The frontier attracted a variety of personalities, many of whom had mixed motivations for their activities, and it is probably due to its nature as a complex frontier entity that Franxinetum is so difficult to characterize. Moreover, while there are pieces of evidence that suggest a close relationship with the Umayyads in al-Andalus, it is likely that such ties never extended beyond informal relations or trade.
It is nevertheless clear that the military and political presence of Muslims in Provence and the Alpine passes presented a serious challenge to Latin Christendom, as is evident from the various Latin chronicles examined above. As with Andalusi Crete in the Byzantine context and Monte Garigliano in Italy, Fraxinetum provided the necessary impetus for the lords and nobles of Provence to put aside their petty struggles and establish a common front in the face of a common enemy. As the Muslims in the eleventh century were to realize, it was this consciousness of feeling vulnerable to Muslim raids and attacks that would prompt the Franks, Castilians, Catalans, Italians, and Normans to exert all their efforts to prevent the Muslims from ever gaining a foothold in strategic areas in the Mediterranean, as they had at Monte Garigliano, Bari, Tarentum, Crete, Cyprus, and Fraxinetum during the tenth century. Indeed, by the late eleventh century Latin Christians would go on to eventually endanger and conquer significant parts of the Islamic world itself: Barbastro in 1064, Toledo in 1085, and Jerusalem itself in 1099. It is quite significant that it was the fall of Fraxinetum in 972 to a coalition of Italian, Frankish, and German knights fighting under the banner of the cross that ushered in this new era.