Jump to ContentJump to Main Navigation
The Articulation of Power in Medieval Iberia and the Maghrib$

Amira K. Bennison

Print publication date: 2014

Print ISBN-13: 9780197265697

Published to British Academy Scholarship Online: January 2015

DOI: 10.5871/bacad/9780197265697.001.0001

Show Summary Details
Page of

PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use.date: 06 June 2020

The Ransom Industry and the Expectation of Refuge on the Western Mediterranean Muslim–Christian Frontier, 1085–1350

The Ransom Industry and the Expectation of Refuge on the Western Mediterranean Muslim–Christian Frontier, 1085–1350

(p.217) 11 The Ransom Industry and the Expectation of Refuge on the Western Mediterranean Muslim–Christian Frontier, 1085–1350
The Articulation of Power in Medieval Iberia and the Maghrib

Camilo Gómez-Rivas

British Academy

Abstract and Keywords

This chapter argues for establishing a connection between the ransoming of captives and the hosting of refugees as a politically legitimising practice. It considers twelfthcentury military and demographic changes that led to an increase in capture and ransom, the legal framework and social response to the ransoming industry, and leaders’ involvement in the release of captives as a high concern of state. An example of large-scale conquest, enslavement, and ransom in the thirteenth century illustrates how ransom and refuge were causally related and predicated upon the reciprocal social expectations of frontier societies.

Keywords:   captives, refugees, ransom, jihād, frontier, Iberia, the Maghrib


DURING THE LONG TWELFTH CENTURY of the territorial decline of al-Andalus, the Iberian peninsula witnessed a sharp escalation of large-scale military conflict between Christian and Muslim armies.1 This escalation was part of a process that solidified the frontier, as a reality and an idea, in the minds of the inhabitants of the peninsula, through an awareness of its symbolic import and violent effect on neighbouring individuals and communities. This frontier or border — even if difficult to pinpoint geographically — became a regional focal point for social mobilization and religiopolitical legitimation.2 The (p.218) dramatic territorial advances of one politico-religious community against the other between 1085 and 1248, and the articulation of the associated military conflict as one between religious communities, amplified the importance of the frontier in the minds of peninsular inhabitants and North Africans (as well as others further afield) and raised awareness of the violence suffered by communities as they came into contact with or were overrun by this frontier. This awareness is amply documented in the textual production of the era, in historiographical, literary and legal forms, to name but some of the most prominent.

While much of the violence of the frontier region and its vicinity did not fit into neat definitions of large-scale holy war or jihād, often involving the pacification of coreligionists (i.e. Muslims versus Muslims and Christians versus Christians), and much of it occurred on a small scale and on a regular basis, one of the characteristic forms of frontier violence was predicated on the intensification of religious identities and involved the capture, enslavement and ransoming of individuals by non-coreligionists (i.e. Christians captured Muslims and Muslims captured Christians by virtue of their religious identity). This activity of capturing and ransoming, while occurring prior to the twelfth century, exploded into an industry in this period and persisted unabated into the early modern era. Typically this meant that individuals whose residence or occupation put them into contact with those on the other side of the ideological frontier (e.g. frontier villagers, soldiers, merchants, sailors) were liable to be captured and sold into slavery or ransomed to their community of coreligionists. Many thousands ended up living in bondage in foreign lands, serving as domestic servants, performing heavy labour, and sometimes engaging in more skilled crafts. Many others, however, were ransomed at hefty sums. The expectation of being saved through ransom often kept captives alive and, along with the substantial profits made by captors and middlemen, formed the economic basis of a ransoming industry that, in many ways, defined and deepened the boundary between the warring ideological communities.

The social responses to captivity and ransom on the Christian side of the frontier have been amply documented and analysed,3 those on the Muslim side much less so. One of the aims of this chapter is to deepen our knowledge (p.219) of the social responses of Muslim communities to the capture and ransom of coreligionists across the Muslim–Christian frontier. The argument I wish to put forward here, however, concerns not just captivity and ransom but also refuge, that is, the act by a host society of receiving a refugee and how this act is socially construed. I argue here that the social expectations and responses that developed as a result of the militarization of the frontier, its solidification in the imagination of peninsular society and the development of an industry of ransoming captives paved the way for the reception of refugees by host societies. In other words, the notion of an obligation to host the displaced was linked to the notion of an obligation to ransom coreligionists (be they family members, community members or subjects). And while one did not directly lead to the other (i.e. ransoming did not necessarily lead causally to the development of the notion of the refugee), I argue that they do come out of the same set of social expectations that developed in a marked way in this period, on both sides of the frontier, and in mirroring ways, for reciprocity between Muslim and Christian societies was a crucial element allowing this violent but calculated interaction to achieve its expected effects. Expectations of being ransomed and expectations of being given refuge were shared by both societies, and the social responses, while different and diverse, resembled and paralleled each other in important ways. The link between ransom and refuge, moreover, can be appreciated in instances in which the ransoming of a large number of individuals led directly to displacement and refugee status.4

That ransoming, displacement and refuge developed as interlinked realities and ideas, at an accelerated pace in the twelfth and thirteenth centuries, can be gleaned from several textual sources, a select number of which I will discuss below. These include a case from twelfthcentury al- Andalus, in which we see how Andalusīs viewed their situation as exceptional and how this exceptionalism affected the law of ransoming prisoners; a legal consultation from the Far Maghrib on questions concerning the treatment of minorities in response to their forced displacement from al-Andalus; a question on endowments (aḥbās or awqāf; sing. ḥubūs, waqf) illustrating the development of social responses to the capture and ransom of Muslims; and diplomatic correspondence between Muslims leaders and the rulers of the Crown of Aragon showing how ransoming became a major foreign relations issue in the thirteenth and fourteenth centuries (and thus closely connected with political legitimation).

This chapter thus explores a dimension of the relationship between the Muslim–Christian frontier and political legitimacy in the Islamic west between the twelfth and fourteenth centuries, by considering a few discrete (p.220) texts and episodes: a fatwā on ransoming Christians, one on building churches and one on creating endowments for ransoming Muslims, the twelfthcentury deportation of Mozarabs to the western Maghrib, and the nature of the diplomatic correspondence between Muslim rulers and the Crown of Aragon. Consideration of these developments and discrete questions together allows one to argue that granting refuge, as a social response and legitimating policy, is a development associated with captivity and ransom, especially when occurring on large scale. I argue that the increase in the ransoming of coreligionists formed part of the development of a framework of social expectations that included a discourse on the virtue and obligation to host the displaced. The ransoming industry itself partook in precipitating the displacement and reception of displaced populations in the western Mediterranean. Such obligations and duties to ransom coreligionists and host the displaced (while not always legally defined) squarely overlapped with notions that informed the bases of political legitimacy of Muslim and Christian rulers. These notions were complicated, moreover, by the fact that these rulers presided over ethnically and religiously diverse communities of subjects whose loyalty was contested, even as ransoming and population exchange was institutionalized on a larger scale than ever before in the region.

The right to release captives for ransom

The end of the eleventh century and beginning of the twelfth constitutes a key moment in the development of the interaction between religious identity, violence and sovereignty in the western Mediterranean, as Christian and Muslim forces became ideologically galvanized against each other along a more clearly delineated frontier. In the Iberian peninsula, the years between 1085 and 1118 — years that witnessed the fall of the first two great Muslim urban centres to Castile and Aragon, beginning the long twelfth century of the territorial decline of al-Andalus (1085–1248) — marked a turning point in the history of the frontier and of its presence in the social imagination of its neighbouring societies. The fall of the Ṭāʾifa kingdoms of Toledo (to Alfonso VI of Castile in 1085) and of Saragossa (to Alfonso I of Aragon in 1118), with the resulting doubling in size of the two kingdoms and the absorption of large populations of Muslims and Arabized Jews and Christians, affords an appropriate starting point for the history of population displacement, and the concomitant rise of a greater awareness of the violence and changing nature of the frontier, its demarcation, militarization and movement (as villages, fortifications, towns and cities fell), which affected growing numbers whose loyalties were fought over by a mosaic of warring parties.

(p.221) The rising importance of the frontier in the minds of contemporaries was expressed in several significant ways. One was through the further articulation and development of a theory and ideology of jihād (Abū Bakr b. al-'Arabī's Risālat al-jihād ma'a al-kuffār (Epistle on jihād against the unbelievers) and Ibn Rushd al-Jadd's fatwā s for the Almoravid amīr, 'Alī b. Yūsuf, on the primacy of jihād, stand out as prime examples).5 Another way was through juridical or legal activity, visible in the responsa literature. Fatwā s and nawāzil (compiled consultations and cases) from the period, dealing with people affected by the movement of the frontier, show a growing awareness of how its movement and violence affected the lives of individuals inside and outside the political community (Muslims, Jews and Christians alike). This awareness, attested to by Muslim jurists of both al-Andalus and the Maghrib, gave rise to a discussion on the interaction of what could be termed religious violence with political power and the rights and obligations of individuals and religiopolitical communities. The cases I describe here form part of this discussion.

The first example was formulated in response to a question directed to the well-known jurist Abū 'Abd Allāh b. alḤājj (d. 529/1134), chief judge of Seville.6 The question concerned the fate of a group of Christian captives who had been seized, we are told, in the general vicinity of the frontier (bi qurb arḍ al-rūm). Ibn alḤājj's opinion in response to this question is brief and straightforward. It was incorporated four centuries later, however, into the larger Maghribī discussion of the law of captives by al-Wansharīsī in his great compilation (which is one of the reasons we have it today). It was preserved, most likely, because Ibn alḤājj, as muftī (or jurisconsult) in this fatwā, stands apart from what had been the majority opinion of his school in one significant way.

(p.222) It is not my intention to claim here that the jurists of twelfthcentury al-Andalus were responsible for formulating the Mālikī law of ransom. Ransoming and captivity go back to the earliest experience of the Muslim community as it fought its first battles in the contest with the pagans of Mecca, and Islamic scripture (especially the reports of the sayings and actions of the Prophet) is replete with references defining lawful and unlawful comportment regarding captives and prisoners of war. What I do believe one can argue, however, is that the first half of the twelfth century marks the point when the jurists of al-Andalus began to put forth a claim that their community lived under exceptional circumstances; and this was an exceptionalism with clear legal ramifications. This idea of Andalusī exceptionalism, moreover, would inform the legal discourse of the Islamic west in later centuries and would be invoked in key subjects such as the obligation of pilgrimage (ḥajj) and the virtue and nature of jihād.7

The Mālikī law of captives and ransom is embedded in the chapters on the law of jihād, because taking prisoners is included among the types of harm permitted to those engaged in lawful war against enemies.8 These enemies are identified on the basis of their religious status; in this particular case (about which Ibn alḤājj was consulted), the enemies susceptible to capture are described as nontributary mushrikūn.9 Because of the historical experience of the early Muslim community, which was characterized by political and military success, the majority of the legal discussions and questions regarding captivity involved determining who could be made captive, who had the right to release that captive and under what circumstances. Issues debated thus included the admissibility of enslaving women, children, the old, the mentally impaired and the pious (e.g. monks) and the admissibility of their being ransomed. One issue of particular contention (according to the comparative legal work of Ibn Rushd alḤafīd) was the permissibility of ransoming ablebodied men, the possible preferability of execution over ransom, and the (p.223) admissibility of receiving money (instead of performing a prisoner swap).10 The whole issue is complicated by conflicting interpretations of the Qurʾānic text and the Prophetic Sunna. Ibn alḤājj, however, argued his position based on the exceptional circumstances of the Andalusī community and on serving the public interest of this ‘exceptional’ community.11

The scarce details of the fatwā continue as follows: these nontributary Christians, who eagerly wanted to return to their country, we are told, were captured ‘close to the lands of the Christians’, and their captors wanted to sell them. The captives produced documentation, however, that, they claimed, secured their right to purchase their freedom or stipulated that they were to be redeemed or manumitted.12 As mentioned above, Muslim jurists (including representatives of legal schools or madhhab s other than Mālikism) were historically reticent in the matter of the permissibility of ransoming ablebodied male prisoners of war for money (often arguing in favour of either execution or the exchange of persons).13 Ibn alḤājj, by contrast, as a response to the question of whether this group of Christians had a legal right to avail themselves of ‘safe conduct’ or release documents, argues that it is permissible to free enemy captives, in exchange for either Muslim captives or money — in this particular case, to be sure, once the proffered documents had been authenticated. The compiler al-Wansharīsī includes this opinion in his larger discussion on ransom and the particularities of different times and places, which may allow for considerations of ‘public interest’ (or maṣlaḥa).14 The exceptional conditions of the Muslim communities of al-Andalus justified such considerations, as Ibn alḤājj and other Andalusī ʿulamāʾ argued (or at least this is how al-Wansharīsī understood their position), the idea being that Andalusīs were in special need of securing the release, safety and ‘humane’ or good treatment of Muslims enslaved or captured by Castilians and Aragonese (there was, therefore, a conscious attempt to influence bilateral relations in hope of reciprocity). They argued that the community was in need of this money (paid in ransom) for its own (p.224) benefit and to offset some of the losses of property and wealth incurred in the wars with the Christian kingdoms.15

Ibn alḤājj's opinion became accepted practice regionally, as attested to by al-Wansharīsī's inclusion of the fatwā and his own comments on the issue in the Mi ʿyār. Al-Wansharīsī, by way of commentary and returning to the broader discussion in which this opinion is couched, emphasizes a point that is not immediately apparent from Ibn alḤājj's opinion, which is that not anyone could give permission for captives to be freed or ransomed. Allowing this, according to al-Wansharīsī, goes explicitly against majority opinion and is based on faulty legal reasoning and support.16 The correct opinion, according to him, is that only the leader (or religio-legal authority, the imam) could oversee and effect such action. This prerogative, moreover, was granted to the imam explicitly based on the notion of serving the public interest, which the imam was considered to be in the best position to determine.17

Contested loyalties

Ibn alḤājj's opinion and its compilation by al-Wansharīsī point to the influence of the Iberian frontier on the awareness among Muslim leaders and ʿulamāʾ of the misfortunes suffered by the victims of the violence on the frontier. I am arguing here that this awareness informed or was related to the developing idea of an obligation or responsibility to provide refuge to the displaced. While this needs further elaboration, the second fatwā I analyse here, also by Ibn alḤājj, supports this argument, albeit somewhat ironically. This fatwā also illustrates how powers involved in the interreligious wars of the Reconquista, while staking their legitimacy on their ability to crusade or wage jihād, also struggled with securing the loyalty of their minority religious communities. In the Muslim case, and in this particular fatwā, this led to the invocation of an obligation to host a displaced population, who, because of its religious identity, presented difficult or controversial issues for the hosts.

This second fatwā thus deals with the rights of religious minorities (Christians in this case) to build new places of worship. The events that (p.225) elicited the formulation of Ibn alḤājj's opinion (which is again rather straightforward) constitute a complex but key episode in the history of the population displacements of the long twelfth century. In September 1125, Alfonso I of Aragon, known as el Batallador, led a spectacular raid (or a devastating one, depending on your point of view) deep into the heart of Almoravid al-Andalus. Alfonso set out from Saragossa with four thousand knights. His principal aim was to attack and, if all went as planned, take Granada, a city that had emerged as a major cultural and economic centre over the past century. Alfonso's army, after advancing through Daroca, Monreal, Teruel and Segorbe, made its way into the province of Valencia, where it sacked the countryside.18

The attack has been interpreted as a sign of the growing weakness of the Almoravid presence in al-Andalus. It is also understood as symptomatic of the emboldening of the Christian kingdoms in the north of the peninsula. While characteristic of the territorial decline of al-Andalus, what appears to have alarmed many of the Muslims of al-Andalus most at this particular juncture was not just the force and extent of the raid, as both sides had long practised such tactics, but that so many Arabic-speaking Christians or Mozarabs had joined the ranks of Alfonso's army, indeed had even invited the attack. By 8 January 1126, when Alfonso I laid siege to Granada, some fifty thousand Mozarabs (a likely exaggerated number) had joined, and a substantial number appear to have done so as instigators (i.e. in premeditated actions). It also appears that some among their number promised to deliver Granada, the Almoravid administrative capital of al-Andalus, to Alfonso. After centuries of coexistence, which had led to the Mozarab community's adoption of Arabic over Latin in almost all aspects of its members’ lives, a general change of mind in this community appears to have been a causal factor in Alfonso's military intervention. It was thus that, contemplating the gravity of the situation, the once chief judge (qāḍī al-jamā ʿa) of Cordoba, Ibn Rushd al-Jadd, set off on 30 March 1126 to Marrakesh, where, as one of the region's most respected legal religious authorities, he would deliberate with and advise the Almoravid leader, ʿAlī b. Yūsuf b. Tāshfīn, on the course to follow.

It appears from the chronicles that recorded the episode (Ibn ʿIdhārī, Ibn Simāk and Ibn al-Khaṭīb) that Ibn Rushd al-Jadd recommended to ʿAlī b. Yūsuf that the Mozarabs found to have been implicated in co-operating with Alfonso should be removed from al-Andalus. A substantial number appear to have thus been displaced to the region of Meknes and elsewhere in the (p.226) Far Maghrib. It should also be noted that before Ibn Rushd's consultation with the Almoravid amīr and, in fact, in the process of Alfonso's retreat to Aragon, as many as ten thousand Mozarabs followed, fearing, most likely, retribution from the Almoravid authorities. What we have in this episode (one of three similar ones in the twelfth century)19 is the beginning of the disappearance of the Mozarab communities of al-Andalus, through forced displacement and migration. It would be fair to argue that, in the long-term contest for the loyalty of the Mozarab community, the leadership of Castile and Aragon won over the Almoravids and later Almohads and Naṣrids, as this episode intimates. More significantly, the dynamics by which that loyalty was obtained, asserting religious identity over other cultural and political bonds, was a powerful factor in the reorganization and cultural and demographic transformation of Iberia in this period.

Ibn Rushd al-Jadd died shortly after his trip to Marrakesh. Sometime afterwards, a substantial number of Mozarabs were forcibly relocated to the Far Maghrib, and it is then that Ibn alḤājj's second fatwā enters the picture. As a prominent jurist (also as part of the loosely institutionalized hierarchy of legal consultation fostered by the Almoravid state), Ibn alḤājj was asked what was to be done about the ‘Christians who have entered the “country” from the other side of the strait’ (al- ʿudwa). They had requested to build monasteries and churches (biya ʿ wa-kanāʾis) in their new places of residence.20 Ibn alḤājj answered that these Christians were to be described as under treaty (wūsifū bi’l-mu ʿāhadīn), and that they were thus entitled to the rights inherent in their protected status. Faithfulness to them is obligatory. Each sect, Ibn alḤājj continues, is allowed to build a temple (or church) to follow their rite or law (sharī ʿatihim).21 This is because the amīr of the Almoravids ordered they be transferred from the Iberian peninsula on account of the fear of them and for the protection of the Muslims.

Ransom funds

The thirteenth and fourteenth centuries gave rise to the full institutionalization of ransom, prisoner swapping and population exchange between the increasingly homogeneous populations of Muslim Granada, the Maghribī (p.227) principalities of the Marīnids, ʿAbd al-Wādids and Ḥafṣids, and the kingdoms of Castile and Aragon. A fourteenthcentury opinion by the Granadan muftī, Abū Saʿīd Faragh b. Lubb (d. 782/1381), illustrates this normalization and institutionalization, as well as an important dimension of the social response among Muslim communities to the ransom industry. The question concerned a captive (asīra) who secured her release from the arḍ alḥarb (land of war) by promising to pay her ransom. A group of mu ʾadhdhins (prayer callers) appealed and collected charity for her and deposited this charity money, which was specifically designated as ransom money, with an amīn (or guardian) until a time at which she would be able to come up with the rest of the money or the full amount was collected. Meanwhile, the captive appears to have been able to raise her ransom on her own, travelling to several locations and appealing to various individuals, eventually managing to gather the full sum. After achieving this, she asked the amīn for the charity funds that had been deposited with him (presumably to cover whatever other personal need she may have had). These funds, however, had not been intended for her personal use in the deed (ʿaqd) when they were deposited with the amīn. The question thus became one about what was to be done with unused or left-over funds, originally collected for the purpose of redemption or ransom. ‘Should these be given to her in spite of the fact that she secured her own ransom without them?’ the mustaftī asks.22

Abū Saʿīd Faragh b. Lubb answered that the question hinged on the issue of whom the money was intended for. In his opinion, the funds became her property when they were given in charity for the purpose of her ransom. The fact that she secured her ransom with funds gathered elsewhere, he suggests, was a matter that could be looked into, but the decision of who the recipient should be lay, in the end, with those who donated the money. He concludes by stating that two possibilities are legally acceptable: she may receive the money or the money may be put aside or asked back from her, if it was stipulated that it be used for ransoming. In this case the money should be used for ransoming someone else. Al-Wansharīsī contextualizes the issue by placing it within the discussion of funds destined for specific purposes and what to do with a surplus remaining from such funds. Citing the famous Mālikī scholar Saḥnūn, he states that if the money was given specifically for ransoming (fikāk), then it cannot be given in charity but must be returned to the donors, who are free to give the money as charity but have to say so formally.23

The issue of the intended uses of such funds is relevant to this investigation, furthermore, because it formed the legal context for many of the questions (p.228) concerning the ransoming of Muslims (unlike that of Christians). Whether ransoming Muslims was ‘permissible’ was not something that was asked often in the legal literature. On the contrary, we find unequivocal statements that ransoming Muslim captives is obligatory (based on a report from the Prophet).24 There is in this regard, however, a more extended discussion involving the use of funds established by pious endowments towards the end of ransoming Muslims. We find in the legal literature a tradition of bestowing money, endowed in the form of aḥbās (or awqāf: pious endowments) destined specifically for ransoming captives. In the Mi ʿyār of al-Wansharīsī, we find a series of questions, spanning from the tenth to the fifteenth century, dealing with the permissibility of specifying or commuting ‘ransom-ḥubūs funds’ or profits; there are more questions dealing with these ḥubūs funds for ransoming Muslims than there are questions over allowing non-Muslims to be ransomed. Of the twenty-one fatwā s in the Mi ʿyār that deal with ransoming captives, seven concern the use of ḥubūs funds that were specified or could be used for liberating captives. All of these deal with Muslim, or potentially Muslim, captives. Another five concern commercial transactions or financial questions, also mostly involving Muslim captives (e.g. what becomes of the property of Muslims held captive abroad?). Of the total of twenty-one fatwā s on ransoming captives that appear in the Mi ʿyār, seventeen appear to deal with questions concerning Muslim captives. In questions involving a ḥubūs or property, however, there were often fewer details about the process of captivity and ransoming itself, since the questions and disputes were over the use of funds, the constitution and functioning of a ḥubūs, or the status of a property (not over the captives themselves).

Diplomatic correspondence

The direct involvement of political and military rulers in the ransoming of captives does not figure prominently in the legal literature, but is in ample evidence in the diplomatic correspondence between Muslim and Christian rulers. Of the 176 Arabic documents held at the Archive of the Crown of Aragon, 28 explicitly concern captives and pleas and negotiations for their ransom (these are mostly treaties and letters sent by the Muslim rulers of Granada, North Africa, and Egypt to the rulers of Aragon).25 Most of these (p.229) letters were written during the fourteenth century. Many of them concern Muslim captives, although there is mention of Christian captives as well. A typical example of this correspondence took place between Ismāʿīl I of Granada and James II of Aragon. A letter written on 3 August 1323 begins with the appropriate salutations, makes a statement of loyalty or friendship (confirming the political relationship), and moves to entreat the king to release captives:

[To] the eminent and sublime sultan, the most faithful, most renowned, the blessed and lauded and most noble Don Jaume, King of Aragon, Valencia, Sardinia and Corsica, and Count of Barcelona. God protect him for his piety and make him happy for obeying and accepting His will. He who extols his power, who thanks and honours his rank, the amīr ʿAbd Allāh Ismāʿīl b. Faraj b. Naṣr. We write to you the present letter from the Alhambra of Granada — God govern its future. And nothing but universal good and the most beautiful divine favours occur [but] through the will of God, may He be exalted. All praise to Him. Your faithfulness is appreciated and your rank among the great kings is celebrated. The object of the present letter is [to request] that you release the three Muslim individuals of our country, captured during peace time, about whom we wrote to you and our messenger, Abū ʿAlī al- Gharrāf, informed you, and whom you [in effect] have ordered to be released, as befits your well-known faithfulness, news which we received with utmost appreciation and commendation for you. When your order was executed, however, [only] two of the individuals were freed and the third remained captive. We therefore write to you again, to thank you, and with regard to this third individual, Abkar b. Ḥasan al-Azraq, of Almería, captured in Alicante, as per the document you are receiving, we ask, Oh King (ayyuhā al-sulṭān), that you complete your favour in a way worthy of your promise. This is what we ask of you. God protect you for your piety and help you accomplish what pleases Him. Many noble salutations correspond to your salutation. Written 29 Rajab al-Fard 723 [3 August 1323]. This is valid — End.26

Other letters from Ismāʿīl I emphasize the reciprocal relationship, especially where the freeing of captives is concerned. His loyalty and the preservation of peace treaties depended on it. In a letter to the lieutenant of the Crown of Aragon, Jaume Andreu, he puts it starkly:

We have always respected the peace treaty with the great Don Jaume. We have complied with the treaty and ordered that his country and people be respected… Taking this into account, we hope you satisfy our demand, sending our men with their mounts and all they have with them.27

Otherwise, he threatens, ‘we will not tolerate it and will be obliged to send our troops to that land so they bring [the captives]’.28The exchange of persons (p.230) was inextricably bound with the ruler's legitimacy as sovereign and protector of his people. A letter from the Marīnid ʿAbd al-Raḥmān b. Abī alḤasan (Abū ʿInān Fāris) promises to return twenty prisoners to Pedro IV of Aragon and encourages him to send ‘the Christian merchants’ (al-tujjār al-naṣārā) where they will be afforded protection.29 And Mūsā b. Abī Yaʿqūb (Abū Ḥammū II, amīr of the ʿAbd al-Wādids)30 on 23 Ṣafar 764/12 December 1362 wrote to Pedro IV of Aragon stating that he would retain thirty prisoners who wanted to return to Aragon and whom Pedro wanted freed (the others, argued Mūsā, chose to stay) until a number of his Muslim subjects from Hunayn,31 Oran and Mostaganem were freed and their captors brought to justice (‘the perpetrators sanctioned and the [stolen?] properties confiscated’).32 Similar correspondence — describing peace treaties, promises of safe passage for merchants and pilgrims, and the exchange of prisoners — was sent from Mamlūk Egypt.

Constituting more than 16 per cent of the extant Arabic language correspondence between Muslim leaders and the Crown of Aragon, it would be fair to say that the exchange and ransoming of captives had become a major concern of state (or kingdom) by the fourteenth century, and that the prestige and legitimacy of these rulers were at stake in their ability to demand, negotiate and secure the release of coreligionists (understood as subjects),33 victims of the religious violence and the ransom economy of the Christian– Muslim frontier.


The foregoing exposition contains a somewhat disparate set of ideas and elements: two muftī s, three fatwā s, the Islamic law of captivity and ransom, the deportation of Mozarabs from al-Andalus, the right of minorities to build new places of worship, the management of endowments for ransoming Muslims, and the Arabic diplomatic correspondence to the Crown of Aragon. I would like to postulate that, however disparate, these elements constitute or represent (p.231) some of the key ingredients for writing a history of refuge and refugees between al-Andalus and the Maghrib. The militarization and movement of the frontier and the changing balance in power, along with the mobilization of public sentiment in the direction of that frontier through the cultivation of ideologies of jihād and crusade (which could be described as a kind of war effort at the centre of which was the religious identity of the warring parties), created a set of new conditions on the peninsula to which the political and legal authorities on both sides of the frontier had to answer with new formulations.

That the ransoming of captives forms a crucial backdrop to the idea of the obligation of communities to receive or host victims of frontier violence may best be illustrated by the episode that took place in 1287, when Alfonso II of Aragon (r. 1285–91) invaded the island of Minorca and, having met resistance, enslaved several thousands of its inhabitants. He held public auctions across his kingdom of between twenty thousand and forty thousand Minorcans captured as a result of the island's capitulation.34 Many, if not most, were ransomed. The rest were sold into slavery.35 While the exact numbers and details are not entirely clear, the basic elements of this episode draw our attention to some key social expectations: just as the Aragonese who settled in the new territories of the crown (such as was likely in our first case) could expect to be ransomed in the event of capture, so could Alfonso II expect that Muslim leaders and their communities in al-Andalus and the Maghrib would pay the high price of ransoming as much of the captive population as possible. From this he could expect to make a substantial profit. The captives, on the other hand, could expect, or at least hope, that they would be rescued and have a place to make a new home, once the price was paid and the negotiations completed. And indeed it is clear that both societies had institutionalized efforts to gather the resources to fund such rescue operations.36

(p.232) On the Muslim side, of course, such ransom, negotiation and rescue led to migration to a place where the refugees could expect, or again, at least hope for, assistance and beginning lives anew. By the fourteenth century we see these social expectations and population exchanges institutionalized on a large scale. The case of Minorca thus dramatically brings together the most important of these disparate elements (large-scale captivity, ransom, displacement and an implicit hope or expectation for refuge). The question remains about what precisely the captors were counting on: love of family, the prestige and legitimacy of redeeming rulers, or the social obligation of the community of believers. Most likely it was a combination of these, but one which the warring parties knew how to manipulate effectively.


(1) Development of this chapter began at a National Endowment for the Humanities summer seminar in Barcelona on ‘Cultural Hybridities: Christians, Muslims, and Jews in the Medieval Mediterranean’ led by Sharon Kinoshita and Brian Catlos. Thanks are due to my colleagues at the American University in Cairo who read and responded to some of this material at the MENA Colloquium organized by Amy Motlag. Amira K. Bennison and James Brown, and the participants at the seminar on political legitimacy in the Islamic west at Magdalene College provided valuable feedback and a sharper conceptual context from which to consider the material.

(2) By solidification of the frontier in the social imagination of Iberian (and Maghribī) societies, I mean the intensification of the ‘border’ between Muslims and Christians, between what came to be thought of as, in effect, ethnic groups, the crucial differentiating characteristic of which was religion, even when these faith communities were diverse in other terms. This anthropological definition of boundary/frontier involves the amplification of difference through increased contact, boundary crossing and exchange, and an emphasis on ascription (i.e. being born into the group) as well as the selective attribution of the group's defining characteristics (which allows for cultural development and variety among members within the group through constant dichotomization with the ‘other’). There is a growing literature on boundary studies in anthropology and modern historical studies. For a classic formulation, see Frederik Barth, ‘Introduction’, in Ethnic Groups and Boundaries: The Social Organization of Culture Difference (Bergen and Oslo: Universitetsforlaget, 1969). For a discussion of the Iberian frontier which evokes a sense of its symbolic dimension and socially mobilizing force, see Amira K. Bennison, ‘Liminal States: Morocco and the Iberian Frontier between the Twelfth and Nineteenth Centuries’, in North Africa, Islam and the Mediterranean World: From the Almoravids to the Algerian War, ed. Julia Clancy-Smith (New York: Frank Cass, 2001), pp. 11–28. Thanks to John Schaefer for discussion and guidance regarding the anthropological concept.

(3) For full discussions of Christian captives in medieval Aragon and the early modern Mediterranean, see Jarbel Rodriguez, Captives and their Saviors in the Medieval Crown of Aragon (Washington, DC: Catholic University of America Press, 2007); Robert Davis, Christian Slaves, Muslim Masters: White Slavery in the Mediterranean, the Barbary Coast, and Italy 1500–1800 (New York: Palgrave Macmillan, 2003).

(4) Here the term ‘status’ is intended not legally but socially.

(5) Abū Bakr Muḥammad b. 'Abd Allāh b. al-'Arabī, Risāla tata' allaq bi-umūr al-kuffār min al-jihād wa-ghayrih (Epistle concerned with the affairs of the unbelievers in jihād and other matters), MS suppl. 50, Beinecke Rare Books Library, New Haven; Ibn Rushd, Fatāwā Ibn Rushd: li-Abī’l-Walīd Muḥammad b. Aḥmad b. Aḥmad b. Rushd al-Qurṭubī al-Mālikī (The legal opinions of Ibn Rushd), 3 vols. (Beirut: Dār al-Gharb al-Islāmī, 1987), vol. 2, pp. 1021–7. The first is a full treatise on war with the unbelievers and is known by either one of the two titles (Brockelmann, Supplement II, p. 647, 25 a). Abū Bakr b. al-'Arabī (d. 543/1148) was chief judge of Seville under the Almoravids and is credited with having introduced al-Ghazālī to the west and with composing works (on legal Qur'ānic interpretation, for example) that form part of the higher learning curriculum of Islamic studies to this day. Ibn Rushd al-Jadd (d. 520/1126), grandfather of Averroës, was one of the most important Mālikī jurists of the medieval period. His opinions are widely cited by later collections; on more than one occasion he was consulted on major religiopolitical issues by the Almoravid amīr, 'Alī b. Yūsuf (r. 500–37/1106–42). See Delfina Serrano, ‘Ibn Rushd al-Jadd,’ in Biblioteca de al-Andalus (2006), vol. 4.

(6) Aḥmad al-Wansharīsī, al-Mi' yār al-mu'rib wa’l-jāmi'al-mughrib 'an fatāwā 'ulamā' Ifrīqiya wa’l- Andalus wa’l-Maghrib (The manifest measure and the amazing collection of the opinions of the scholars of Ifrīqiya, al-Andalus and the Maghrib), 13 vols. (Rabat: Ministry of Culture and Religious Affairs, 1981), vol. II, pp. 179–80.

(7) Ibn Rushd al-Jadd's fatwā to 'Alī b. Yūsuf on the virtue of jihād over ḥajj in al-Andalus (because of its exceptional circumstances, including the lack of security to travel to the Ḥijāz) is a salient example by a contemporary of Ibn alḤājj. Ibn Rushd, Fatāwā Ibn Rushd, p. 1022. For an analysis of this opinion, see Camilo Gómez-Rivas, ‘The Fatwas of Ibn Rushd al-Jadd to the Far Maghrib: Urban Transformation and the Development of Islamic Legal Institutions under the Almoravids’, PhD thesis, Yale University, 2009, pp. 117–23.

(8) On the Mālikī law of ransom, see Milouda Hasnaoui, ‘La ley islámica y el rescate de los cautivos según las fetwas de al-Wanšarīsī e Ibn Tarkāt,’ in La liberazione dei ‘captivi’ tra cristianità e Islam: Oltre la crociata e il ǧihād: Toleranza e servizio umanitario (Vatican City: Archivio Segreto Vaticano, 2000); Russell Hopley, ‘The Ransoming of Prisoners in Medieval North Africa and Andalusia: An Analysis of the Legal Framework’, Medieval Encounters, 15: 2 (2009): 337–54.

(9) Sing. mushrik: polytheist, or someone who associates something with God, here meaning nontributary Christian.

(10) Ibn Rushd alḤafīd, Bidāyat al-Mujtahid (The distinguished jurist's primer), trans. Imran Ahsan Khan Nyazee, 2 vols. (Reading, UK: Garnet, 1994), vol. I, pp. 456–61.

(11) What interests me here is not whether al-Andalus was, in fact, exceptional, but that jurists would consider it such and use this as a basis of argument.

(12) The precise nature of this documentation or ‘contracts’ (ʿuqūd) is unclear. Presumably each individual had a different kind of document (it is likely they were merchants who carried such documents for precisely this kind of eventuality). Some of the documentation was produced in Malaga and some in Marrakesh (i.e. it was in Arabic and produced by Muslim authorities). ‘Redeemed’ (iftikāk) here is equivalent to ransom; what these Christians most feared, it appears likely, was being enslaved and/or sold.

(13) Ibn Rushd alḤafīd, Bidāyat al-Mujtahid, vol. I, p. 456.

(14) A Mālikī term for a criterion on which one ruling (ḥukm) can be preferred over another.

(15) Al-Wansharīsī, al-Mi ʿyār al-muʿrib, vol. II, pp. 179–80.

(16) The issue of whether the imam and only the imam was entitled to release captives was controversial, although according to al-Wansharīsī it was only the opinion of Ashhab (d. 204/819), an Egyptian Mālikī jurist. See Jonathan Brockopp, ‘Ashhab’, EI3.

(17) Al-Wansharīsī supports this opinion on the authority of Ibn Rush al-Jadd, whose al-Bayān wa’ltaḥṣīl is quoted several times in the exposition. Ibn Rushd, al-Bayān wa’l-taḥṣīl wa’l-sharḥ wa’ltawjīh wa’l-ta ʿlīl fī’l-masāʾil al-mustakhraja (The exposition, study, explanation, instruction and explication of the subjects excerpted), second edition, 20 vols. in 19 (Beirut: Dār al-Gharb al-Islamī, 1988), vol. II, pp. 517–614.

(18) For an analysis of this event, see Vincent Lagardère, ‘Communautés mozarabes et pouvoir almoravide en 519H/1125 en Andalus’, Studia Islamica 67 (1988): 99–119; Delfina Serrano, ‘Dos fetuas sobre la expulsión de mozárabes al Maghreb en 1126’, Anaquel de Estudios Árabes 2 (1991): 163–82.

(19) Lagardère, ‘Communautés mozarabes’, p. 99.

(20) Al-Wansharīsī, al-Mi ʿyār al-muʿrib, vol. II, p. 215.

(21) This opinion is echoed in al-Burzulī's citation of Ibn alḤājj on the matter of whether Christians and Jews could build religious structures: each ‘sect’ (ṭā ʾifa) can construct one building so that they can practise their rites separately. Abū’l-Qasim b. Aḥmad al-Balawī al-Tūnisī al-Burzulī, Fatāwā al- Burzulī: Jāmi ʿ masāʾil alaḥkām li-mā nazala min al-qadāyā bi’l-muftīn wa’lḥukkām (The fatwā s of al-Burzulī), 7 vols. (Beirut: Dār al-Gharb al-Islāmī, 2002), vol. II, p. 19.

(22) Al-Wansharīsī, al-Mi ʿyār al-muʿrib, vol. II, p. 211.

(23) Ibid.

(24) Apparently without disagreement, across all schools. See, for example, al-Burzulī: ‘And that is obligatory after the transmission of Ashhab, Muṭarrif and Ibn al-Mājishūn from Mālik: That the ransom of Muslim captives is obligatory (wājib). Saving them is requisite, and if they are still in the lands of Islam they may not be enslaved.’ Al-Burzulī, Fatāwā al-Burzulī, vol. II, p. 40.

(25) Alarcón y Santón describes 162 documents of which 24 deal with captives; today the archive counts 176 Arabic documents, which have been digitized and are available to the public from the Spanish Archives Portal ( http://pares.mcu.es/).M. A. Alarcón y Santón, Los documentos árabes diplomáticos del Archivo de la Corona de Aragón (Madrid: Imp. de E Maestre, 1940).

(26) Ibid., pp. 40–1.

(27) Ibid., pp. 49–51.

(28) Ibid., p. 50.

(29) Alarcón y Santón dated this letter 2 Dhū al-Hijja 755/18 December 1354, the year being difficult to read. Abū ʿInān Fāris died in 1351, so this year seems unlikely. Alarcón y Santón, Los documentos árabes diplomáticos, p. 204.

(30) Regnal dates, 760–91/1359–89.

(31) Medieval seaport in western Algeria, 45 kilometres northwest of Tlemcen. EI2, s.v. Hunayn.

(32) Alarcón y Santón, Los documentos árabes diplomáticos, p. 233.

(33) Interestingly, the letters tend to underscore the territoriality of origin of victims and transgressors involved in the correspondence, such as, ‘Those pirates who captured your subjects are not from our “country”’ and ‘Individuals from our country have been captured, please return them.’ Territorial origin and subjecthood are underscored more than religious identity.

(34) The chronicler Muntaner provided the higher number. Verlinden revised it down. Rodriguez cites the episode as exceptional for its scale. Rodriguez, Captives and their Saviours, p. 9; Ricard Soto Company, ‘Una oferta sin demanda? La esclavitud rural en Mallorca antes de la peste negra (ss. XIII–XIV)?’, Historia Agraria 21 (2000): 11–31; Charles Verlinden, ‘La esclavitud en la economía medieval de las Balearicas, principalmente en Mallorca’, in Cuadernos de Historia de España 47–48 (Buenos Aires: Instrucción de Historia de España, 1982); Charles Verlinden, L’esclavage dans l’Europe médiévale, 2 vols. (Bruges: DeTemple, 1955).

(35) Which brings up the important point that the ransom industry went hand in hand with the slavery industry, in which the experience of the Balearics after the conquest is a significant forerunner of the larger operation that was to come in later centuries.

(36) As noted above, the creation of endowments appears to have been one of the important social responses on the Muslim side. Military orders, such as the Trinitarians and Mercedarians, constitute a key response on the Christian side. Rodriguez, Captives and their Saviours, pp. 140–7. On Muslim slaves in the military orders’ hospitals, see Ana Echevarría Arsuaga, ‘Esclavos musulmanes en los hospitales de cautivos de la Orden Militar de Santiago (siglos XII y XIII)’, al-Qanṭara 28: 2 (2007): 465–88.