1. Systems of Stratification

[135] From a comparative perspective, social structure offers problems of analysis which are difficult to resolve. Social historians have described Islamic and Christian societies of the middle ages according to different norms and using different terminologies. Christian Spanish society, like those of Western Europe, is conceived as having had a typical class structure, the definition of whose components is based mainly upon economic criteria: an aristocratic upper class that controlled the means of production; an emergent middle class which sought to wrest such control from the noble elite; and a dominated mass of free and servile peasants. But for Islamists the class model has had considerably less explanatory value, because the bias of the sources is to play down the significance of class distinctions; because social classes lacked the corporate distinctiveness they had in western law; because a number of other cleavages (ethnic and religious) have been seen as more significant than class ones; and because status (as in the cases of military and bureaucratic elites) was not necessarily a function of class in the western sense.

The discussion here of medieval Islamic society will be in accord with the general judgment of Islamists that satratification was by groups in which membership was ascribed on some basis other than economic (e.g., ethnicity, religion, kinship), although frequently ethnic, religious, or other cleavages tended to be coterminous with economic ones, or at least to have strongly patterned economic ramifications. The class structure was not as sharply articulated as in the Christian West and, as a result, economic mobility was more fluid.⁽¹⁾

The success of a comparative treatment of medieval societies would seem to rest upon the selection of structural features general enough to permit comparison without foundering on overly specific or ethnocentric constructs which are not easily transferred to different cultural settings. Thus, as Archibald Lewis wisely concluded in a comparative study of "feudalism" in medieval Europe, Islam, and Japan, the very real points of similarity tend to be obscured by the use of a term loaded with Western [136] connotations.⁽²⁾ If terminological problems can be overcome, then valid comparisons and contrasts may be found: for example, the religious ideologies of both cultures encompassed social philosophies which legitimated unequal distribution of wealth among various groups.

In the "polemic of Spanish history" the disagreement between Castro and Sánchez-Albornoz as to the nature of the differing social structure of medieval Islamic and Christian societies rests upon terminological and conceptual imprecisions. For Castro, ethno-religious cleavages outweigh the significance of class distinctions and therefore the interaction of "castes" (Muslims, Christians, and Jews) leant tonality to, and provided the basic dynamics of, medieval Christian social life. To Sánchez-Albornoz, Castro's "horizontal" structuring of Castilian society seems patently absurd. For the former author, Castilian society was articulated vertically, by classes, not castes, like other societies of the Christian West.

This aspect of the polemic is confused by semantic problems. Specifically, Castro does not adequately define what he means by caste. To the present author, who first read Castro in English, the term has a specific meaning, amply and precisely defined in anthropological literature, of a closed, endogamous group, the lateral mobility of whose members (interaction with other castes) is severely circumscribed. In Spanish, however, the term has more restrictive connotations of purity of lineage (i.e., the usage castizo, pure of lineage) upon which Castro centers much of his argument without adequately defining the social, as opposed to ideological or cultural, structuring of caste interactions. Castro appears not to have realized that the castes were themselves internally stratified by class and that, as a result, interaction among members of different castes was colored by a common social structure in which all participated.⁽³⁾

Equally incomplete is Sánchez-Albornoz's discussion, couched as a resounding denial of the relevance of caste analysis. In his zeal to refute Castro, he in effect denies that ethnic groups have roles which are not susceptible to definition in purely class terms. Indeed, he even asserts that although Jews constituted a true caste in other European societies, they did not so do in Spain, where they were simply part of the middle class.⁽⁴⁾ This kind of argument is simply playing with terms and results in oversimplification of a complex social system.

Indeed, a social situation as complex as that of medieval Iberia demands analytical strategies which do not deny that complexity. In the discussion which follows, class and "caste" are not seen as mutually exclusive modes [137] of social organization. Rather, I stress the co-existence of both class and status structures in both societies (though in different mixtures) and attempt to determine the rules of social interaction prevailing in each. First to be discussed are the systems of kinship that underlay the structuring of more complex social solidarities; then the class structure; and, in Chapter 5, the dynamics of ethnic relations. It is true that in separating for purposes of analysis the different features of kinship, class, and ethnicity, the resultant picture of social structure may seem disjointed. I do so with this risk in mind.

2. Kinship

Until recently, the nature of kinship and its shaping effect upon social and political institutions in medieval Spain was not a topic accorded much importance by historians. The results of this oversight have been disastrous, leading above all to a complete and general misinterpretation of the nature of Islamic society in Spain. This imbalance has been rectified by the work of Pierre Guichard, who has demonstrated the tribal organization of Andalusi society of the Emirate and, in the Christian orbit, Ruíz Domenec, García de Cortázar, and others have identified the dissolution of the extended family as a significant and central social process of the high middle ages.

The Arabs and Berbers who conquered the peninsula did so not as isolated warriors, but as members of organized tribal groups. The Arabs and most of their early Berber allies were members of agnatic, patrilineal groups forming a segmentary social system, whereby individuals belonged to a hierarchy of increasingly inclusive segments, from the clan up to the tribal confederation. The basic tribal unit, the qawm (variously translated fraction or clan), is a unit of several hundred tents or families, linked agnatically. That is, the kinship system ascribes importance only to relationship through males. In such a system, endogamous marriages are viewed as the ideal because through endogamy power, prestige, and wealth are retained within the agnatic group rather than shared with a competing group into which a daughter might marry, with parallel-cousin marriages (the wedding of one's son with the daughter of the paternal uncle) preferred. A cross-cousin marriage (with the daughter of the maternal uncle or paternal aunt) is considered exogamous because the offspring gain a different lineage. The more powerful a tribal group is, the [138] more women it will attract from outside, the fewer it will lose, and the more endogamous it will become.⁽⁵⁾

Guichard demonstrates that the early Muslim residents of the peninsula settled in tribal or sub-tribal groups and that, indeed, it was the policy of important figures to travel with tribal entourages and to reconstitute their clans once the decision to settle in al-Andalus had been reached. This was the policy, to cite the most obvious example, of 'Abd al-Rahmân I, who gloried in having reunited his clan in Spain. A corollary to this strategy is that leaders build power by controlling their agnates.⁽⁶⁾ Prestige and power in segmentary societies is predicated upon the strength and cohesiveness of agnatic groups, and the political strategy of such groups is to attain the necessary cohesion through the practice of endogamy and through constant testing of the group's strength in competition with rival groups. Thus, as many students of Arab and Berber society have noted, a permanent state of war ("organized anarchy," in Robert Montagne's characterization) is the direct result of segmentary social organization.⁽⁷⁾

Segmentary organization gives rise to typical political forms. The basic unit is the clan -- the Arab qawm, the Berber canton -- which lives and fights together. The segmentary tribal structure makes it possible for such groups to subsist in relative isolation and, at the same time, because they are embedded in larger solidarities, to join in political or military federations with related groups. This gives rise to the kaleidoscopic pattern of atomization and amalgamation which is so characteristic of western Islamic, particularly Berber, society.⁽⁸⁾

The dispersion of clans yields one of the most distinctive visible patterns of the cultural geography of al-Andalus: the profusion of Beni- toponyms, particularly in the east of the peninsula. (These place names are generally identifiable as descriptive of a clan, named for its founder -- e.g., Benimamet, sons of Muhammad -- or of a clan as a tribal segment, as in Benigomar, from the Gûmara Berber tribe.) Such names, which are mainly Berber and represent settlements which seem to have postdated the emirate period, have been used as a historical source by Guichard in order to establish the tribal organization of the countryside. Most of these names are of insignificant places, mainly rural (and, when urban, indicative of the settlement of quarters or suburbs by agnatic groups) and typically mountainous (small valleys with twenty or so hamlets of which half might have Beni- names).⁽⁹⁾ There is other data complementing Guichard's picture of the dispersion of related groups throughout the countryside.[139] Oliver Asín has noted traces of Berbers (probably Christian) in the heartland of Old Castile. He identifies the toponym Quintana there with Berber settlement, based upon the khimâsa ("one-fifth") sharecropping system typical of many regions of North Africa.⁽¹⁰⁾ I also believe that these Quintana, Quintanilla settlements may be of Berber origin, but that they relate not to the agrarian, but to the kinship, structure. Among many Moroccan Berber groups, tribes are divided into five primary segments or clans (khams khmas =: "five fifths"). In such a way, clans reproduce themselves spatially, maintaining the same name and enabling the creation of linked but dispersed networks of tribal relationships. It is this sense of "fifth" which may underlie the Quintanas of Old Castile, as well as such a place-name as Algemesí, a town in Valencia which was a region of heavy Berber settlement.⁽¹¹⁾

Arab and Berber tribal structure found political expression in the organization of confederations or alliances, which were formed according to the underlying logic of segmentary societies. The essence of this kind of political organization is that politics is viewed as a zero-sum game. The wealth, power, and prestige of one's own group are increased only by decreasing those of a rival group, leading to a more or less permanent state of conflict between neighboring groups as well as to characteristic patterns of alliances. The aggressive tendencies of this society, which has its roots in the socialization of children (through the encouragement of sibling rivalry and competitive behavior which is later transferred to those outside the agnatic group), expressed itself in typical socio-political formations, paramount among which was the division of the entire ethnic group into moieties. Such a dual division, based on fictive genealogical traditions, characterized both Arabs -- divided into Qaysîs (Mudarîs) and Kalbîs (Yemenîs) -- and the Berbers -- divided into Branes and Botr. The result is a socio-political system which requires that there be two, and only two, parties to any conflict, which in turn requires the formation of dichotomous coalitions.⁽¹²⁾

Much of the political history of al-Andalus, therefore, is occupied with accounts of tribal in-fighting, generally along lines of moiety division. Yemenis had originally concentrated in the southwest of the peninsula and in the Upper March; the Mudaris settled in the east but were displaced towards the center by the arrival of Yemeni contingents under Balj in the 840's. The governorships, as noted earlier, passed back and forth between Mudari and Yemeni candidates (the jund controlled the [140] governorship but was itself tribally organized and therefore participated in this typically tribal game). The Fihrîs -- the Qaysi family to which the last governor Yûsuf belonged -- based its claim to power on Mudari strength in the central regions, in alliance with Berbers to counter the superior force of the Yemenis. These latter supported the Umayyads in their successful attempt to establish an independent emirate. That the Mudari/Yemeni split was a very real element in the socio-political dynamics of al-Andalus, especially through the ninth century, cannot be doubted. Abu'l-Fidâ's much-quoted anecdote about seven years of tribal warfare in Murcia resulting from the picking by a Mudari of a grape-leaf from a Yemeni's vine is symptomatic.

Throughout the ninth century, the neo-Muslim Banû Qasî allied themselves with Mudaris to oppose the more numerous Yemenis in the Ebro Vallev, and, in various places around the peninsula, the fitna (unrest) of the latter part of that century was initiated by Mudari/Yemeni conflict. Strife between the moities tended not to be reflected in the sources thereafter, as the population became predominantly neo-Muslim, but the fall of the Caliphate was interpreted by some medieval Arab commentators as the result of Mudari/Yemeni antagonism. During the eleventh century, if inter-Arab hostility played less of an overt political role, differentiation at the sub-cultural level and group distinctiveness continued to survive: in the religious debates of that period Mudaris, according to Dominique Urvoy, were portrayed as religious zealots, Yemenis as tolerant of non-Arabs.⁽¹³⁾

The tendency in the social evolution of Islamic society was for the tribe to yield to the clan as the main unit of social organization, to be replaced in turn by relationships of a client-patron type. This progression is not necessarily linear, however, and seems to be a function of the dual forces of sedentarization and urbanization. According to Guichard, tribes ceased to be dynamic organisms in Islamic Spain because they ceased segmenting, which is the basic historical process of tribal organization. The permanence and stability of ethnic names from the eighth through the thirteenth centuries demonstrates that new groups, bearing new names, were not being created, a situation which reflected the separation of Andalusi society from the great, and still living, tribal organisms of the Islamic East. On the other hand, consanguinity remained a powerful social force (and so remained even among the Moriscos of the sixteenth century, who resisted taking Spanish names because such an act made it impossible to keep [141] track of agnatic lineages), as did ethnicity.⁽¹⁴⁾ Guichard notes that tribal feeling declined consistently, but that the mechanisms causing that decline are not understood. It seems to me that tribalism ceased to be a controlling factor in Andalusi social life at that point along the curve of conversion where, in mid-tenth-century, non-tribal neo-Muslims began to swamp the smaller Arab and Berber populations numerically. After the conversions were completed, the cultural center of gravity shifted to neo-Muslims, who were not involved with, and consequently cared little about, tribal issues. The enlargement of the non-tribal sector of society did not mean, however, that tribal values and forms of organization disappeared among previously tribalized elements, or among newly arrived Berber groups. Guichard's data on the incidence of Arabic tribal names (nisbas) among Andalusi jurisprudents seems to confirm both Bulliet's general hypothesis and his methodology (discussed in Chapter 1). In the caliphal era, the number of persons with Arabic tribal nisbas was still small (about one-third of ibn al-Faradî's biographies); but two centuries later about two-thirds of ibn al-Abbâr's sample have tribal names.⁽¹⁵⁾ This is about what one would expect, given the acculturation of converts to Arab norms, and is more a function of the modality of conversion than of tribal organization per se.

The kinship systems of Christian Spain present a less clear historical physiognomy: the model of blood relationships seems never to have been as sharply defined as was the patrilineal, agnatic structure among Arabs and Berbers. On the one hand, one finds a general drift away from the typical bilateral, cognate structures of primitive Germanic society, towards an agnatic, patrilineal structure. But this movement was an oscillating one which left a good deal of room for cognate relationships and in which the degree of agnaticism varied from epoch to epoch and from class to class. Indeed, the class differentiation of kinship relations is possibly the single point which most distinguishes Christian from Muslim kinship systems. On the other hand, there was a more clear, more deliberate evolution which saw the extended family (whether of cognatic or agnatic orientation) yielding in the face of a socio-economic context that favored the stem or nuclear family. The stem family was powerfully supported not only by certain social and economic advantages accruing from the special conditions of frontier life, but also by ideological support from the Church. Here is the second distinct element of differentiation: Islam provided a framework which legitimated tribal values and gave them [142] religious significance; Christianity tended to work in the opposite direction, toward the development of inter-personal, rather than inter-group bonds.

The gens, or family, of the Hispano-Romans, one assumes, was patrilineal and agnatic. This is in contrast to the matrilineal kinship systems of the Cantabri and Basques, the least Romanized peoples of the peninsula, who retained segmentary organization in tribes and clans into the high middle ages, although under Roman influence the modal form of kinship seems to have become patrilineal. In a transitional state, succession among the Cantabri passed from male to male, but in the maternal line.⁽¹⁶⁾

There are numerous points of comparison between the kinship structures of the Basques-Cantabri and the Arabs and Berbers, against whom the former were the first to offer active opposition in the initial phases of the "reconquest." Such is the burden of Barbero and Vigil's analysis, which makes perfect sense in the light of Guichard's portrayal of Andalusi social organization: the segmentary Basque-Cantabrian society was, by the very nature of its tribal structure, able to meet the challenge offered by the alien tribal groups, through the process of continual warfare characteristic of such societies. Moreover, the evolution of tribal society among the indigenous peoples of the Cantabrian coast underwent a development similar to that which the Muslims were undergoing in the early decades of their domination of the peninsula: the devolution in organization from the tribal level to that of the clan.⁽¹⁷⁾

The kinship systems of the indigenous peoples of Iberia absorbed not only the influence of the Roman gens but of the Germanic family, which, unlike that of both the Romans and the Arabs, was bilateral. Elements of this bilaterality survived well into the middle ages, particularly in Germanizing customary law. In Castile, for example, the family's joint liability in penal matters was recognized. The Fuero of Sepúlveda provided that vengeance for the murder of a relative was not a crime if the avenger had the same great-great-grandfather as the original victim. Other stipulations, such as those regarding the necessity of the wife's permission in the sale of property (Fuero of Jaca), seem related to this tradition.⁽¹⁸⁾

Nevertheless, bilaterality seems throughout this period, and all over the north of the peninsula, to have been yielding to a patrilocal, patrilineal, agnatic kinship system. The extended family was still in evidence in the tenth century and can be identified as the basic land-tenure unit. Thus, Ruiz Domenec likens the tenth-century Catalan mansus or mas to the Anglo-Saxon hide -- the terra unius familiae -- a continuous space exploited [143] by a single, extended family.⁽¹⁹⁾

On the other hand, there are ample signs that the extended family, in whatever form, was in decline throughout the period, although among the nobility it held its form longer. The structure of the Cid's family in the eleventh century is a case in point: his mesnada or entourage included four of his nephews as well as the husband of his sister-in-law, a group that was clearly enough bilateral. But the Cid's formal name, Rodrigo Diaz de Vivar (the latter element a place name, the locus of the familial landholding), indicates that the emphasis of social organization had shifted from the community of blood relationships (as revealed in late Hispano-Roman naming patterns which included tribal ascriptions) to the common hold of the extended family.⁽²⁰⁾

Throughout the high middle ages the scope of the extended family was consistently diminished, particularly as the state, or public authority, was able to provide enough personal security to diminish the need for blood vengeance; as the Church promoted a similar policy through the promulgation of the "Peace of God"; as new institutions such as guilds and military orders replaced the older solidarities, based on kinship; and as urbanization and colonization, reflecting a more mobile society, hastened the fragmentation of old kinship groups. Feudal relations, involving the dependency of unrelated persons, were, in effect, substitutions for bonds formerly maintained within the kindred circle.

On the Catalan frontiers of the late tenth and early eleventh centuries, the process of disaggregation of the extended family or (in Ruiz Domenec's characterization) the scaling down of patrimonies to unicellular units of production can be observed in full force. A major impetus to the cultivation of new lands there was owing not to population pressure per se, but to the accelerating movement of the fragmentation of the extended family, as a result of which conjugal units left behind a joint patrimony in order to cultivate a parcel of their own on the frontier. Consequently it is among the frontier peasantry that Bonnassie detects the greatest fragility in family structure, with the insecurity of frontier life favoring a conjugal family. The move towards the frontier was also abetted by the practice of partible inheritance, whereby those with small shares in the family patrimony were eager to migrate. At the same time, inheritance rules were shifting towards a situation favoring indivisibilty of patrimonies and inheritance by primogeniture: through the practice of inmeloratio, marriages were restricted and the better portion of the patrimony was bequeathed to one son only.⁽²¹⁾ Thus, in a struggle between the [144] rights of the kindred group (partible inheritance) and the need to keep the patrimony in substantially one piece, the economic imperative won out over an already weakened family structure.

As in Castile, elements of extended kinship survived among the nobility, as Catalan counts tended to marry endogamously (although the counts of Barcelona, for political reasons, were an exception, preferring to marry non-Catalan nobility).⁽²²⁾

The differential structure of kinship in Islamic and Christian societies offers a fundamental point of departure for comparative analysis. Historians of each society have asserted that the way in which kinship relations were articulated lent flexibility and adaptability to the society in question -- flexibility which proved crucial at certain critical junctures of competition between the two blocs. Thus Guichard asserts that the flexibility of agnatic systems allowed such societies to retain vitality while inflexible state systems declined, with the result that the former were easily able to conquer the latter. Similarly, Barbero and Vigil assert that, among the Basque and Cantabri, agnatic flexibility allowed these tribally organized peoples to resist the centralizing grip of the Roman state and to oppose successfully Arab and Berber military pressures through the similarity of their social structures.⁽²³⁾ On the other hand, Ruiz Domenec claims that bilaterality lent to Catalan society of the tenth century an extraordinary flexibility in the manipulation or economic resources and in the making of political alliances. Indeed, Guichard demonstrates that in the case of strictly agnatic groups endogamy acts to promote a rather brittle political structure, conducive to continual tribal in-fighting and to the forming of unstable, if predictable, alliances. When bilaterality survived, as in the case of the Neo-Muslim Banû Qasî whose daughters married Christian Basque nobility, the result was to produce a territorial domain with relatively stable governance lasting several generations: a situation more typical of Christian than of Islamic political forms.⁽²⁴⁾

Thus, in al-Andalus, agnatic flexibility proved an advantage at a time when Christian states were disorganized, a disadvantage when it discouraged the creation of a strong, stable institutional base to defend the society against cohesive Christian powers. Bilateral flexibility allowed, among the Christian powers, the creation of a natural system of family-based alliances among princely houses, while at the same time allowing the military class to build, through similar processes, a stable and substantial economic base.

Looking ahead to a consideration of the comparative dimensions of [145] "feudalism," we can agree with Marc Bloch's observation that strong kinship ties and those of a feudal nature are, to a degree, antithetical (although feudal institutions did have the secondary result of reinforcing kinship bonds already on the wane). The weakness of kinship ties explains why people seek, or accept, other kinds of relations that offer security and protection. Bloch shows that wherever powerful agnatic groups survived (e.g., in Celtic areas), there was no feudalism.⁽²⁵⁾ Extending the argument, one might say that feudal relationships come into existence only at the expense of kinship arrangements which they were designed to replace or to compensate for. One would expect that wherever and whenever agnaticism survived in al-Andalus there would have been no feudal arrangements whose end was protection or group cohesion, since those features would still be supplied by the kinship group.

Comparative kinship study also makes clear that certain values, such as honor and aggressivity, generally (although erroneously) thought to inhere in a culture, are in fact related to specific aspects of the kinship and family structure, and to the way in which children are socialized. To give but one example: Guichard provides an elegant analysis of the different social bases of the feeling of honor in Islamic and Christian societies. In the tribal society of Arabs and Berbers, honor is directly tied to agnatic feeling and is reified: when one's agnates are attacked or vilified, one suffers a direct loss of honor. Among Christians, honor as a value varied with social status and was defined not in terms of one's embeddedness in a particular group, but in more external terms, such as wealth. In Christian society, for example, a poor man's honor is increased by marriage to a wealthy woman with higher social status. In agnatic society, a wife taken from outside the agnatic group cannot enhance the prestige of her husband, no matter how much wealth she might bring him.⁽²⁶⁾

Values, in sum, do not float freely in a culture and cannot be handed down from distant ancestors as part of a general cultural heritage. Thus, to say that medieval Spaniards exhibited certain Iberian, Roman, or Gothic personality traits is to say nothing, unless it can be shown that there was also continuity in the social structure that produced those traits.

One further aspect of a comparative nature bears mention. Christians, upon conquering and occupying al-Andalus, had frequently to replace institutions which had been tribally organized with non-tribal organizational forms. (To be sure, the progressive weakening of the extended family had adequately prepared them for this task.) A case in point are [146] the communally organized irrigation systems of southern and eastern Spain. We know that previously irrigation canals, especially those in areas of Berber settlement, were organized along tribal lines. Evidence of this is found, for example, in the huerta of Valencia, where the Favara Canal was at one time an entity of the Hawwâra Berbers (saqit al-Hawwâra), and in Gandia and Murcia, where secondary canals retain the telltale Beni names indicative both of clan settlement and of the division of water by lineage groups. In tribal societies, as in certain irrigated districts of Morocco today, such canals are administered through the normal workings of tribal political organization: the community of elders decides distribution arrangements, settles disputes, determines the custom, and provides executive authority. When the Christians took over these canals, the customary arrangements regarding distribution procedures, allocation of water rights, and so forth, were learned from the Muslims but were subsequently administered either by town councils (stressing the authority of the elders) or by autonomous communities organized along guild lines (stressing the egalitarian aspects of Berber tribal organization).⁽²⁷⁾

3. The Lower Classes: Freedom and Dependence

In Christian Spain one can detect a general pattern of the early disappearance of servile tenure and the concomitant predominance of small, free proprietors. Although Roman emphyteutic contracts did survive in such socially conservative areas as Galicia, in fact the mass of cultivators who participated in the settlement of areas like the Plain of Vich and the Duero Valley of the ninth and tenth centuries were holders of allodial parcels and possessed other properties (such as shares in mills) consistent with their free and independent status. At the same time, perhaps the most striking aspect of this frontier liberty was its relative shortlivedness, its tendency to be eroded by the progressive extension of seignorial domains and prerogatives. By the time the Duero line was breached, the peasant proprietor was severely disadvantaged.

In describing the social structure of the Asturo-Leonese kingdom, Sánchez-Albornoz distinguishes between two socially distinct regions: the mountainous northwest, where late Roman social structure survived in an evolved form (notably in Galicia which had plenty of serfs with which to populate the north of Portugal), and the zone to the south of the mountains in León and Castile, settled by Mozarabs and individuals [147] from points north.⁽²⁸⁾ Those who participated in the repopulation of the Duero Valley tended to be free men, more so in León than in Galicia, and even more so in Castile, where the military needs of the counts served to enhance the status of freemen even more than elsewhere on the frontier. The settlers tended to be free migrants -- advenientes et escotos -- although another large group of settlers held a more qualified status. Lands granted ad populandum frequently went to juniores de hereditate who held emphyteutic contracts but who were legally free and could alienate their holdings.⁽²⁹⁾

The emergence of hamlets and villages (villae) of freemen is one of the most singular aspects of Iberian society of the ninth and tenth centuries. A plethora of documents reveal small free proprietors acting in collectivities, as in the donation of a collectively owned meadow to an abbot in 972,⁽³⁰⁾ or singly, in the sale or donation of small parcels or other appurtenances.

The economic freedom of action of these small proprietors can be appreciated in their ownership, either individual or collective, of mills or rights to irrigation water. Unlike the typical western European situation, whereby the ownership of mills was vested in the king or territorial lords (a monopoly which underscored the economic dependence of the peasantry, forced to grind grain at the lord's mill), in Castile and León groups of freeholders owned and evidently built mills for their own use or benefit. Indeed, such mills vastly outnumbered, during this period, those owned by kings or by ecclesiastical institutions. Even when a nobleman owned a mill, such possession may not have had the seignorial connotations one might expect. The Cid, for example, was scorned by the Beni Gómez as a working miller, and therefore was deemed unfit as a prospective in-law:

Who hath news of My Cid of Vivar?
To the Ubierna [River] bid him go, to tend his mills
And exact his toll as is his wont!
Shall a miller's daughters wed with scions of Carrion? (l.- 3378-3381)

Once a mill was constructed, it would either be divided into shares by [148] the builder, or else usufructory shares (called vices, typically the equivalent of one hour or one day's use per week) were distributed. These shares could be alienated and formed an important portion of the freeholder's patrimony. There are frequent documents from the tenth century detailing transactions among smallholders involving the ownership of such shares in mills. That such persons owned as little as one hour's use per week, or a one-eighth share in the mill itself, indicates that men of modest means were involved. In 1012, no less than twenty-one such proprietors collectively sold their rights (nostras vices) in a mill on the River Vena to the abbot of Cardeña.

The same was true of water rights and the appurtenances of irrigation systems. One finds individual proprietors owning an hour of water, or a share in a diversion dam or canal. Such holdings were subjected to the same pressures as were felt upon land. Thus, alongside substantial small ownership, one finds as well monasteries buying up mill shares, in an effort, perhaps, to acquire in a piecemeal way the monopolies characteristic of their European counterparts.⁽³²⁾

The emergence of whole settlements of freemen raises the issue of the similarities and differences between the social structure of the frontier and that of the hinterland. In form, certainly, free and seignorial villae must not have differed much. For tenth-century Catalonia, Bonnassie describes a minimal aggregation of seven or eight families of allodial landholders, plus a priest and a smith -- the basic village community. The same would appear true of León and Castile, with perhaps a larger population per settlement. Such hamlets were the basis of agrarian exploitation and could evolve in two directions: they could become the nuclei of free towns (concejos) or they could become seignorial, subsequent to their acquisition (for example) by a monastery.⁽³³⁾

An intermediate state, that of collective commendation to a lord, with no loss of freedom, is mentioned below.
 

The social structure of the seignorial village (the mansus) seems not to have differed much from the general organization of manors in pre-feudal Europe. There was a group of servile peasants, called casatos, originally small in number but increasing in the eleventh century. These were adscripted to the land in the typical manner of serfdom. Then, there was a large class of collazos or solariegos (the successors to holders of emphyteutic contracts), not adscripted to their land but who owed substantial dues (taxes and small amounts of grain, livestock or wine) and [149] work services to the lord. This class resembles the Anglo-Saxon ceorl. In Catalonia, there seems to have been a similar distinction between the adscripted homines de redemptione and the villani de parato, who paid rent.⁽³⁴⁾ But two distinctions must be made concerning the internal stratification of the peasant class. First, the distinction between free and servile was not a critical class division, in spite of the interest of medieval jurisprudents to make it so. Second, by the end of the eleventh century, in Castile, casatos and collazos had effectively merged and the terms describing them became interchangeable, as the status of all seignorial peasants was leveled out.⁽³⁵⁾

The role of the frontier as a creator of freedom has been much stressed in the literature, especially by Sánchez-Albornoz (as in his famous characterization of the Duero Valley as an island of freeholders in feudal Europe).⁽³⁶⁾

Certain of the settlers, such as Mozarabs and Basques, seem to have arrived on the frontier with no previous or relevant bonds of dependence. On the other hand, flight to the frontier was not entirely unrestricted. In fact, since kings and lords did not wish to weaken previously established seignorial domains, men called homini excusi . were especially favored as migrants. The notion of "excused men" appears, for example, in a charter of Fernán González of 941, in which he grants the monastery of Cardeña a place to be settled "not with men from my villages, but with excused men, and from other villages." "Excused" in this sense meant free from tribute and seems to have represented an attempt to prevent the migration of any who were not of originally free status.⁽³⁷⁾ Thus the frontier was increasingly settled by men who were totally free of obligations but whose freedom to move varied with circumstances and custom. As Carlé implies, peasant families were seen by their lords as dues-paying units: so long as the dues continued to be paid, mobility was possible. Those who owed no dues had no limits placed on their movement -- and such status was most typical of the collateral relatives of the head of household and his heir. Sometimes movement was contingent upon forfeiture of the original holding, or upon the sale of such a holding to someone who would continue to pay the due. The simplest solution was for a lord to shift his own men around, from older settlements to less densely populated ones.⁽³⁸⁾

The eleventh through thirteenth centuries were characterized by an accelerating movement of seignorialization (particularly ecclesiastic domains, which grew rapidly through pious donations, whereas lay domains, [150] which had to be enlarged through purchase, lagged behind); this meant a concomitant loss of freedom of action for the small proprietor. In the plain of Barcelona, we note the allode yielding to the precarium in the eleventh century, as well as the adscription of the peasant to his land; in Galicia, the spread of the incommuniatio contract, whereby the peasant yielded half his property to a lord and received it back in tenancy; in León and Castile, the amalgamation of free and servile peasant sectors to form a more homogenous peasant class which, if not entirely tied to the land, was definitely not free.⁽³⁹⁾

Most typical of the retrenchment of freedom was the spread of the benefactorium or behetria contract in Castile and León from the eleventh century on. This contract, one of partial dependence, was at first entered into freely by (for example) old persons who wanted protection in their old age, women without families, or indebted persons. These peasants continued to farm their own parcels but now paid rent to a lord to whom they owed, in return for protection, loyal service, and whom they called dominus or senior. Thus their status came to approach the servile norm; yet they remained legally free, could break the contract and change domicile at will (vadit ubi voluerit, in the words of the Fuero of León). With the passage of time, the contracts tended to become hereditary and, by the thirteenth century, behetrias had become collective, involving whole villages which had formerly been completely free.⁽⁴⁰⁾

It should be noted that this process of the progressive loss of freedom by the peasant class, whether actual (as when expanding seignorial domains acquired peasant lands) or fictive (as when jurisprudents, under the impetus of Roman law, associated various classes of peasant tenure with Roman emphyteutic contracts), was quite common in western Europe. In Norman England, the old ceorl class lost its freedom as feudalizing jurists assumed its equivalence with unfree villein tenure. Given the frequent political and social turmoil of the period, peasant groups who were protected may have enjoyed an actual advantage over those who retained full legal freedom but who had to struggle against powerful noble interests on their own. A result of the partial loss of freedom implicit in the behetria contract was a gain in group cohesion which may have been to the peasants' advantage in a climate of political uncertainty and declining economic returns.⁽⁴¹⁾ The intermediate situation of the hombres de behetria may be compared favorably to that of seignorial tenants, such as those of Sahagún, who formed a hermandad (brotherhood) in [151] 1111 to oppose the excesses of the Abbot -- a revolt which failed owing to lack of group cohesion.⁽⁴²⁾ Or, one could compare their relatively stable position with that of their counterparts south of the Duero who were severely disadvantaged because of the virtual impossibility of the small proprietor's participating in the dominant sector of the agrarian economy, the sheep-raising industry.⁽⁴³⁾

By comparison with accounts of the Christian lower classes, data concerning the masses in Islamic Spain are scant indeed. The urban lower class -- the 'amma -- appears in historical sources (such as those describing the participants in the Revolt of the Suburb in ninth-century Córdoba) as a socially undifferentiated group: it included not only workers, but also artisans and small merchants.⁽⁴⁴⁾

The occupational structure of this class becomes visible in ibn 'Arabî's description of Andalusi sufis of the early thirteenth century. The mystical movement was characteristic of lower-class urban artisans, as revealed in their occupational sobriquets (nisbas), summarized in Table 3. Cobblers, smiths, tailors, potters, as well as workers in trades of obviously low status (henna siever, tanner) found solace in Sufism in the waning years of Andalusi civilization, faced with tremendous pressure from the Christian states.

There were slaves in al-Andalus, to be sure, but these tended to enjoy a domestic, military, or administrative role, rather than an economic or agrarian one.⁽⁴⁵⁾

Considerably more material survives concerning the rural lower classes, but many of these sources are of Christian provenance, describing the Muslim minorities of the twelfth and thirteenth centuries. Studies of the Ebro Valley and of the Valencian region make clear that the sharîk (meaning partner, party to a sharecropping tenancy contract), or in the Latin documentation, exaricus, was the predominant social type in the Islamic countryside. Exaricos were seen by an earlier generation of historians as servile tenants, probably reflecting the tendency of medieval jurists to assimilate their status to that of servile Christian peasants, adscripted to the land and holding emphyteutic contracts. In fact, the Muslim exarico continued to hold the same status under the Christians as under Islamic rule: he was a contract-tenant with real, butlimited, ownership rights. The basis of the contract was a perpetual lease at a fixed rent, but the exarico was a free man and could alienate a parcel held under such a contract at will.⁽⁴⁶⁾

Under such terms, Muslims living under royal jurisdiction (as in the Ebro Valley during the thirteenth century) tended to [153] be better off than their Christian counterparts, who inhabited seignorial domains. Christian documentation, therefore, tends to support the view that sharecropping was the universal form of agrarian exploitation in Islamic Spain.⁽⁴⁷⁾

Castilian, Leonese, and Portuguese documents of the eleventh century and later refer to persons called maulados (Latin, maullatus). The term seems always to have been used in the sense of encomendado, and its abstract substantive form, maladía, was used interchangeably with comenda, always referring to land arrangements. These maulados were Mozarabs who used the term, derived from Arabic mawla (client) to describe their social status which, according to Sánchez-Albornoz, was at times that of freedman (liberto) and, at others, that of an hombre debehetria.⁽⁴⁸⁾ We can accept, therefore, Sánchez-Albornoz's conclusion that maulado was a simple synonym for hombre de behetria, and also that these maulados or their progenitors had not in al-Andalus been mawlas in the strict sense of the word. Clientage was, strictly speaking, a relationship of dependence with an Arab tribe entered into by converts to Islam. But then, what does this usage imply in terms of the perception by Mozarabs of social life in al-Andalus? It would seem a reflection of the common status of all rural peasants, whether Neo-Muslim or Mozarab, of dependence upon a patron, with or without formal adherence to a tribe (which in practical terms may not have made much difference anyway). The use of the term by Mozarabs doubtless can be taken as an indirect proof of the generality of clientage relations in al-Andalus.

Lévi-Provençal carried this discussion a step further by noting the semantic equivalence between the Arabic mustana' and the Latin benefactus, concluding that there was another type of clientage relationship prevalent in al-Andalus, the istina', which he supposed to be a simple continuation, in Muslim dress, of the late Roman patrocinium. ⁽⁴⁹⁾

Such a common origin would explain the convergence of Islamic clientage relations among Neo-Muslims and Arabs on the one hand and between Mozarabs and their Muslim patrons on the other.

4. The Middle Classes: Two Failed Revolutions?

Middle classes were comprised of those who, through the practice of trade and of artisan crafts, gained control of a portion of the means of production, formerly monopolized by an aristocratic elite. Because this [154] control was partial, the middle class had to form cross-cutting alliances of one or another kind with the elite. The existence of such alliances in all medieval societies inevitably leads to confusion as to where the boundary between the middle and upper classes lay. Since the objectives of the middle class included the desire to join, supplant, or countervail the aristocracy, the boundary between the groups was, in part at least, a function of the relative success which the bourgeoisie attained in reaching such goals. Moreover, since the dominant mode of analysis rests upon an evolutionary hypothesis (that at a certain stage of economic development a "middle class" will emerge, grow, and eventually seize power), the problem of defining class boundaries becomes confused with issues that are, in effect, teleological. Thus one finds, particularly in Castilian historiography, concepts such as "true bourgeoisie" and "the normal development" of a middle class, indicating the measurement of that group's evolution against a reified and abstract standard. Thus Valdeavellano asserts the "normal" development of a bourgeoisie in towns where French influence was felt, as opposed to a truncated development elsewhere.⁽⁵⁰⁾ Reyna Pastor clearly demonstrates the cross-cutting alliances made by Castilian burghers of the twelfth century with the nobility, yet then asserts that these burghers did not constitute a "true" middle class because of their parasitic relationship with the feudal structure.⁽⁵¹⁾ Elsewhere, the same author concludes that there was likewise no "true bourgeoisie" in al-Andalus, even though the middle class there did manage to escape subordination to a feudal nobilility.⁽⁵²⁾

Therefore the question of class definition is related to, and partiallyobscured by, the issue of the sharing of power between aristocratic and bourgeois sectors. In the Islamic world, the bourgeois revolution is reckoned a failure because the attainment of aristocratic status was a matter either of lineage or of entrance into a state military bureaucracy. Yet a distinctive element of the middle class -- the fuqahâ', or jurisprudential elite -- while it disdained a certain kind of political power (the direct seizure of or participation in state control), still exerted enormous influence over certain policy matters (those directly touched by the concerns of Islamic law) and over the tonality of public life in general. In Castile, the bourgeois revolution foundered on its own success: noble status was too accessible and the middle class was progressively deprived of its leadership from the mid-thirteenth century on, as caballeros villanos were ingested into the lower nobility.⁽⁵³⁾ The very notion of a medieval bourgeois [155] revolution seems an anachronism, conceptualized in terms of nineteenth-century processes which can only obscure the variegated socio-economic patterns of medieval societies which allowed for considerable variation in the structuring of class boundaries, in spite of attempts to define them with legalistic exactitude.

The expansion of Islam created conditions favorable to a commercial revolution of gigantic proportions by linking sources of raw materials to centers of production, and these to centers of consumption, and by joining together areas of specie surplus with those of deficit. The agents of this phenomenon were a commercial class of Muslim, Jewish, and Christian traders, based in the great Mediterranean entrepôts, who ranged over the entire length of the network, from Spain to India. The commercial and industrial sector was the social base of an important middle-class subgroup, the fuqahâ', who provided the personnel which staffed the judicial and executive offices of urban jurisdictions (e.g., judgeships and prefectures), which, in the absence of a differentiated political structure, provided the low-level institutional continuity which gave cohesion to the society even in the face of political instability at its upper levels. The Islamic middle class of the middle ages was characterized, therefore, by an intertwining of commercial and scholarly interests: the reverence of the merchant for learning, the interconnectedness of commercial and scholarly networks of communication (revealed in the complex patterns of travel with the frequently combined motives of scholarship and commerce), and the fact that Islamic law, the metier of the fuqahâ', was codified during the epoch when a commercial mentality dominated Islamic urban society.⁽⁵⁴⁾

The fuqahâ' are the only sector of the Islamic middle class about whom there is enough information to provide at least the starting point for a social analysis. Fuqahâ' are consistently characterized as "people of means" (ahl al-khair), and it is clear that many came from mercantile families, when not engaged in commerce themselves. 'Abd al-Haqq b. al-Hasan, a judge in Almohad Almuñécar was known as al-Mâlî ("the wealthy"), noted ibn al-Zubair, and the prosperity of his forefathers explains why. Another faqîh, 'Abd al-Haqq b. Muhammad al-Zuhrî, whose biography was recorded by the same author, was described as a "professional businessman (muhtarif b'il-tîjara). ⁽⁵⁵⁾ Similar examples could be extracted from any of the biographical collections.

There are indications, too, that the fuqahâ' were not a homogeneous [156] class but were internally stratified. According to Hussain Mones, there was a serious cleavage among Andalusi fuqahâ' of the ninth century. One group, the shuyûkh al-fiqh, were narrowly legalistic and interested in exercising power. Later in the century a new group, primarily interested in the emerging religious sciences (especially that of tradition -- hadîth), came to the fore. These shuyûkh al-hadîth traveled to the East in search of learning and gained, upon their return, a moral prestige superior to that of their rivals.⁽⁵⁶⁾

The cleavage in scholarly focus must have been underlain by social distinctions, although one can only surmise what these might have been. Certainly those who traveled to the East were socially, as well as geographically, mobile. Yet there are no apparent distinctions in wealth between the two groups; a typical figure of the fiqh group, Muhammad b. Mu'awiya, was a businessman, but those who learned hadîth must have had the capital to finance their journeys. More likely, this cleavage is related to the curve of conversion. The shuyûkh al-hadîth who returned from Eastern studies to attain positions of scholarly leadership in the first half of the tenth century were learned enough to create an Andalusi school of jurisprudence. This creation of a group of scholars no longer dependent upon Eastern tutelage seems a characteristic feature of the explosive period when local institutions are needed to serve a rapidly expanding mass of Muslims. If Bulliet is correct, one might guess that the shuyûkh al-hadîth included numbers of recent converts, proceeding from a higher stratum than that represented by the shuyûkh al-fiqh. The hypothesis merits testing, although the adoption of Arab names and false genealogies makes this all but impossible. Certainly by the thirteenth century it is apparent that most fuqahû' were muwallads.⁽⁵⁷⁾

As a result of the massive conversions, the fuqahâ' achieved tremendous authority as moral and religious preceptors of Islamic society, particularly urban society. The political expression of this authority is discussed below (Chapter 6, section 2), but here it is necessary to point out that the legitimacy of the fuqahâ's prestige among the Muslim masses rested in part upon their learning and their role as religious mediators and in part upon their ability to disassociate themselves from the more unsavory aspects of political power. Inevitably, therefore, the relationship of the middle class, whose values were those of fuqahâ' (i.e., religious and commercial), with the ruling elite, whose values were tribal, aristocratic, and tended towards laicism, was bound to be ambivalent. Refusing to accept gifts or political appointments from princes was a sign of moral prestige. Civil power [157] was of no interest to them. In the late tenth century, a faqîh dreamed that he had said to ibn Zarb, chief judge of Córdoba: "I have never found anything more bothersome than the frequenting of kings." On the other hand, for the good of the social order, kings had to be frequented, especially when they set bad examples for the rest of the Muslims. The sixty days of fasting imposed by Yahya b. Yahya upon 'Abd al-Rahmân II for fornication is a well-known example.⁽⁵⁸⁾

Once their group cohesion was established, the fuqahâ' were able to maintain it through generations by virtue of the structure of learning itself, which provided a kind of mechanism of elite recruitment modeled after tribal lines of authority. The prestige of a given faqîh was in part the reflection of that of his masters, and the masters of his masters, and of the entire chain of teachers upon whose authority he transmitted tradition. Thus scholarly genealogies can be established for scholars (and were, in fact, in biographical dictionaries) which paralleled those of family.⁽⁵⁹⁾

The necessity of establishing a scholarly genealogy accorded well with both the tribal and occupational structure of the society. Members of the same family tended to follow similar pursuits. Of the four hundred thirty-four jurisprudents of the Almohad period registered by ibn al-Zubair, forty-nine (11.3 percent) transmitted on the authority of, or studied with, their fathers. An additional twenty-one (4.8 percent) studied with a grandfather, cousin, uncle, or another relative.⁽⁶⁰⁾

If the internal structure of this class is not quite clearly delineated, the boundaries between it and the tribal or military aristocracy seem to have been sharply defined, by tradition and by the differing value systems that each group embraced.

In Christian Spain the problems of analysis are reversed: the internal structure of the middle class is easily defined, but the boundaries between it and the nobility are often hazy. In the Barcelona region, Bonnassie sees a "pre-bourgeois" class, the sons of villagers who had become rich by selling agricultural surpluses, emerging in the late tenth century. By the early eleventh, the class is consolidating itself as an urban patriciate, taking advantage of an inflationary economy which permitted it to reinvest profits in rural property.⁽⁶¹⁾ The tight ligatures between town and countryside seem to be a mark of the emergent Catalan middle class.

In the west of the peninsula, although professional merchants are documented in the town of León by 910 and references to merchants increase throughout the eleventh century, it is only in the twelfth that such [158] individuals really concentrate in cohesive groups in towns. According to Valdeavellano, the middle class of León and Castile was, in its origins, dependent on the flow of commerce along the pilgrimage route and seems to have arisen only in those areas colonized by francos. In towns not located along the pilgrimage route, the middle class was but thinly established and in some of these places it disappeared completely as a result of the civil wars of the twelfth century.⁽⁶²⁾

The consolidation of the bourgeoisie along the pilgrimage route can be appreciated from the privileges accorded to the town of Jaca by Sancho Ramírez in the last quarter of the eleventh century: burghers were vouchsafed the inviolability of their domiciles, exemption from onerous military duty, and a modicum of immunity from royal justice (the king's merino could not collect fines within the city).⁽⁶³⁾

In both the east and west of the peninsula the steady consolidation of the middle class was accompanied by the progressive stratification of the middle groups into blocs distinguished by differential economic and political power. Thus, the population of the burgo of eleventh-century Zamora was stratified into mayores (those who had a horse) and menores. The same was true in Santiago and Sahagún a century later, where there was clear social differentiation among an incipient patriciate of merchants and master craftsmen and those apprentices and artisans who worked for them, and, in the Catalan country, in Lérida, where a commercially based patriciate of prohombres dominated a more numerous pueblo menudo of artisans.⁽⁶⁴⁾

Such cleavages were normative throughout the towns of medieval Europe, as political power was subjected to the exigencies of competing guild interests and as the commercial patriciate began to make its way into the ranks of the lower nobility.

This latter phenomenon is particularly characteristic of Castilian and Leonese towns, where conflictive relations between burghers and nobles in the twelfth century gave way to a certain amalgamation of interests in the thirteenth. Rebellions of burghers against seignorial privileges in Sahagún (1087, 1110-1115, 1136, 1152), Santiago de Compostela (1116 against the powerful bishop Diego Gelmírez; and 1136), and Lugo (1110, 1159, 1184) are the clearest indications of the social maturity of the new class. In Pastor's particularly lucid analysis, the francos and other immigrants who populated the towns along the pilgrimage route were social non-conformists who quickly rebelled against the status quo and who were able to instill their values of non-conformity in their children; this [159] fact explains the generational periodicity of the revolts. These rebellions began over specific issues, such as objections to seignorial monopolies, and soon broadened to include a variety of general objectives, such as urban political and fiscal autonomy. The burghers intended to control the market and to break the power of the bishops to set prices, coin money, and otherwise dominate the economy. To achieve these ends, the burghers organized brotherhoods (hermandades) to oppose the lords, in armed struggle when necessary.⁽⁶⁵⁾

Toward the end of the twelfth century, recurrent monetary crises in the western kingdoms provided an opportunity for burghers to increase their economic and political leverage, particularly since the kings were able to use them as a countervailing force against the nobility. When royal income failed to meet the fiscal needs of the kingdom, it was customary for the king to call in silver and to reissue debased coins with the same face value. Middle-class resistance to this practice led the towns to intervene in meetings of royal councils, which led to their expansion to include burghers along with the noble and ecclesiastical estates in the Cortes of León (1188), Castile (ca. 1200), and in the rest of the peninsular kingdoms during the course of the thirteenth century.⁽⁶⁶⁾

But Castilian burghers were not only opposing noble interests. In certain areas of governance they were in fact superseding nobles, while at the same time forging matrimonial alliances with the lower nobility as the first step in acquiring such title themselves. Teofilo Ruiz's studies of bourgeois families in twelfth- and thirteenth-century Burgos make this pattern clear. There, we note such families as the Sarracin and Bonifaz using personal fortunes gained in the importation of cloth from Flanders, Brabant, and England to gain positions of authority in the municipal government, in the royal administration (e.g., appointment as moneyers), and in the church. Ruiz concludes that the upper strata of Castilian bourgeoisie, far from failing to establish the power of their class, were, in effeet, "too successful" in that, by the mid-fourteenth century, they had entered the lower nobility. In this process, the reign of Alfonso X was a watershed. Alfonso took measures to limit the power of the nobles and at the same time to enhance that of the caballeros villanos by granting them tax exemptions, as a result of which the latter were able to acquire municipal offices. At the same time, the petty bourgeoisie (the master craftsmen, homes buenos) were effectively disenfranchised, leaving municipal power in the hands of this new nobility.⁽⁶⁷⁾ The rise of the [160] caballero villano (discussed in section 5, below) is perhaps the most distinct aspect of medieval Castilian social structure.

The incipient middle class of the Christian kingdoms was, in its early stages, ethnically heterogeneous, including not only local families, but francos, Mozarabs and Jews as well. The latter appear to have emerged early, on a near-equal basis with other bourgeois sectors. Around A.D. 1000 there was a prosperous Jewish colony in Barcelona whose members seem to have invested their commercial property in dispersed landholdings. The same is true of the early Jewish community of León, where residents of the castrum judeorum were engaged in commerce (some, for example, were in the employ of the commercial house of Menéndo González, engaged in the silk trade) in the eleventh century and appear in the twelfth as owners of agricultural parcels dispersed over a wide area in the vicinity of the town and held mills on the Bernesga and Torío rivers.⁽⁶⁸⁾

A symbol of the economic value of these communities was the high wergild placed upon Jewish individuals, the equivalent in many cases of a noble's. This is not an example of tolerance, however, so much as the recognition by the nobility who profited from Jewish commerce that Jews had to be protected from harm in order for them to pursue their trade in a hostile social environment. From a position of juridical near-equality with Christians in Fueros of the late tenth and early eleventh centuries, the social status of Jews progressively deteriorated as anti-Jewish strictures were introduced, beginning in the mid-eleventh century.⁽⁶⁹⁾ It seems clear that the worsening situation of Jews is a reflection of the growth of a Christian middle class which came increasingly to resent Jewish economic competition and legal prerogatives. Jews then began to be squeezed into their familiar role as money-lenders, a development which increased tensions until relief from Jewish debts became a common cry in the Cortes.

5. The Nobility and Social Mobility

To a considerable extent, the structure of medieval aristocracies, whether Christian or Muslim, was inherent in the nature of agnatic kinship. In both societies noble status was adscripted, although the force of adscription may have varied. In al-Andalus, those tribes that had been prominent in Arabia (Arabs) or in North Africa (Berbers) tended to remain prominent, although, with the passage of time, the creation of urban, non-tribal power centers, and the swamping of tribesmen by masses of converts, [161] created a more fluid situation, whereby individuals whose tribal roots were weaker or entirely fictive could achieve high status. The pattern in Christian Spain was the opposite. There the social fluidity of the early middle ages appears to have impeded the formation of strong blood solidarities until the thirteenth century, after which time noble status became strongly agnatic in character.⁽⁷⁰⁾

There are two basic characteristics of the Christian nobility which bear special comment. The first is their ability to control resources; over a comparatively short span of time this allowed them to convert vast areas formerly the province of smallholding freemen into seignorial preserves. The second is the nature of the internal stratification of the class, whose lower strata were surprisingly permeable to the inclusion of new stock.

Aside from the agnatic character of noble families, what primarily characterized them was their ability to control resources, both human and natural, through the exercise of immunities, taxation, and monopolies.⁽⁷¹⁾ In the ninth through eleventh centuries one notes the steady growth of seignorial domains, the ecclesiastical estates growing quickly in wealth and power owing to pious donations, with the lay aristocracy gaining at a consistent, but markedly slower, rate. That gains in lay seignorialization came only slowly is best symbolized by the fact that the balance of power did not shift decisively from the monarchy to the nobility until the great conquests of the thirteenth century, as a result of which seignorial forms of political control were institutionalized in the newly conquered territories.⁽⁷²⁾ Nevertheless, the growing marshalling of economic resources by the nobility is much in evidence by the late tenth century: direct or indirect (through the ability to levy taxes in kind) control over iron mines, and the typical medieval pattern of monopolies over mills and ovens (the latter a point of conflict between nobles and burghers).⁽⁷³⁾

The second point of distinctiveness is the permeability of the lower echelons of nobility by upwardly mobile burghers. The noble estate was internally stratified into a higher class of magnates (maiores) and a lower one (minores) -- the viri nobiles of tenth-century Catalonia, the infanzones of Castile and León. This lower class was distinguished by its lack of wealth and by its close vassalatic dependency, on the counts in Catalonia, the king in the Asturleonese and successor kingdoms. In Catalonia of the tenth century, a noble of this class was typically chatelain of one castle; in León, his patrimony was based on personal presuras of his family, and Sánchez-Albornoz has demonstrated the modest dimensions of such [162] holdings. The power of infanzones derived not from their personal patrimony but from feudal exemptions and from salaries paid them as soldiers. In a legal sense, the rights of high and low nobles were essentially the same.⁽⁷⁴⁾ The exigencies of constant warfare made it possible for wealthier townsmen to fight as knights and quickly gain a quasi-noble status. Thus one notes the rise of the caballero villano or ciudadano in tenth-century León, forming a distinctive sub-class that grew more numerous with time. Once burghers acquired enough money to maintain horses and to fight as knights, they received the perquisites of knighthood -- notably exemption from dues -- which enhanced their financial position and enabled them to begin to control municipal councils by the end of the twelfth century. In the northern towns, as already noted, caballeros villanos retained their commercial interests, whereas closer to the frontier, in the towns of New Castile, they appeared as cattlemen.⁽⁷⁵⁾

Not only burghers, but also rural freeholders, were able to benefit from opportunities offered by the frontier and by the spoils of war in order to better their social positions. Sánchez-Albornoz notes the rise of Castilian free proprietors to the status of infanzón in the Duero Valley of the late tenth century and, a century later, the wholesale creation of lower nobility from the ranks of foot soldiers in order to bolster mounted forces. This phenomenon is apparent in the Poema de mio Çid, where the hero gives land in Valencia to his men and makes knights of those who had fought on foot.⁽⁷⁶⁾

This kind of nobility was not unknown in other parts of Europe. In Anglo-Saxon England, for example, it was recognized that a man of non-noble birth could "thrive to thegnhood." What distinguishes this phenomenon in the Castilian orbit is its apparent massiveness and also the multiplicity of stimuli encouraging upward mobility: the abundance of land on a constantly expanding frontier, a chronic shortage of labor, urbanization, the military needs of states fighting the Muslims, the spoils of war. In this way, the freedom of the small rural proprietor and that of the town-dweller were ultimately exercised. Either one used that freedom to attain noble status through one of the available avenues, or freedom was lost.

By the thirteenth century the lower nobility of caballeros villanos and infanzones comprised the bulk of the mounted units fighting in southern Spain, and the military unit was easily converted into a settlement unit once the fighting was over. It was at this point that the more evolved seignorial structures which had grown slowly in the north were [163] instituted in the south, not only in the formation of great estates granted to magnates and military orders, but, even more significantly, in the creation of small seignorial domains in urbanized areas and in the irrigated zones surrounding them. This was accomplished as a result of royal policy, as in Murcia, where James I wanted to create a cohesive ruling and military class for the defense of the territory and where an hombre de valor was granted an estate comprised of a minimum of one hundred tahullas (the irrigated land measure).⁽⁷⁷⁾ Further discussion of the Christian nobility is reserved for the discussion of feudalism, in Chapter 6 below.

Analysis of nobility in Muslim Spain has been subsumed in our discussion of Arab and Berber kinship systems. Two major differences between Islamic and Christian society must be mentioned here, however. First, as Guichard makes clear, power in Islamic societies resided in lineage and not in land and, as a result, the Islamic states that faced the Christian onslaught had difficulty in forming a cohesive military class.⁽⁷⁸⁾ The genius of the feudal system of relations was that it wedded lineage, wealth, and the obligation to fight, forming a close-knit system whose basic link had become, by the thirteenth century, the grant of land. The Christian frontier advanced as a structured unit (more sharply structured with the passage of time) against which a society whose main ligaments were still tribal, and weakly tribal at that, could offer little cohesive resistance. This situation was further exacerbated by the replacement, beginning in late Umayyad times, of a blood nobility by a service nobility (the so-called "Slavs"), which further weakened the military capability of the society. Indeed, strong resistance was offered only when the peak of society was effectively tribalized, as a result of the Almoravid and Almohad invasions, which featured the implantation of Berber armies, fighting in tribal units. The devolution of fighting capability was a concomitant of urbanization, detribalization, and the domination of society by converts who had never formed part of a tribal structure. This was a common pattern in the states of the Islamic East, none of which, however, had to face the peril of permanent conquest by a non-Muslim enemy.

Finally, there were enough points of contact between Christian and Muslim nobility to suggest that both the values and the social structure of the two groups were similar enough to make each at least recognizable to the other. Throughout the high middle ages Christian desnaturados and exiles sought refuge and were warmly received in al-Andalus by their noble counterparts in whose houses they resided and in whose armies they [164] did battle. The stays of Alfonso VI in Toledo, his brother García in Seville, and the Cid in Zaragoza were typical. "To live among Moors was the ineluctable destiny of every exile," Menendez Pidal has observed.⁽⁷⁹⁾

Such alliances seemed normal to Christians, but scandalized the Muslim masses, especially after 1085, when Christian pressure upon Islamic territory seemed relentless.⁽⁸⁰⁾ When, as a result of conquest, Muslim nobles ended up under Christian rule these aristocrats seem to have had fewer problems of assimilation than did their correlegionaries of lower status. Thus Christian knighthood was conferred upon loyal Muslim allies, according to chivalric ideals shared by both groups.⁽⁸¹⁾

1. On class versus status stratification in medieval Islam, see Guichard, Al-Andalus: Estructura antropológica de una sociedad islámica en occidente, p. 276, where he states that the most important divisions in society were ethnic; Bryan S. Turner, Weber and Islam: A Critical Study (London: Routledge and Kegan Paul, 1974), p. 93; Chalmeta, Señor del zoco, p. 14; Goitein, Mediterranean Society, I: 76, 80; and, as an example of a succesful class analysis by an Islamist, Ira M. Lapidus, Muslim Cities in the Later Middle Ages (Cambridge, Mass.: Harvard University Press, 1967).

2. Archibald R. Lewis, "The Midi, Buwayhid Iraq, and Japan: Some Aspects of Comparative Feudalisms, A.D. 946-1055," Comparative Studies in Society and History, 11 (1969), 47-53.

3. See my discussion of the deficiencies of Castro's analysis, Chapter 9, section 4, below.

4. Sánchez-Albornoz, España: Un enigma histórico, II: 8, 44.

5. Guichard, Al-Andalus, pp. 60-65, 81. 217.

6. Ibid., pp. 204-205.

7. Ibid., p. 262; Robert Montagne, The Berbers: Their Social and Political Organisation, David Seddon, trans. (London: Frank Cass, 1973), p. 46.

8. On the clan, see Guichard, Al-Andalus, pp. 256, 463; Montagne, Berbers, pp. 30-31; David M. Hart, "Segmentary Systems and the Rule of 'Five Fifths' in Tribal Morocco," Revue de I'Occident Musulman et de la Méditerranée, 3 (1967), 69-95.

9. Guichard, Al-Andalus, pp. 423-424, 430-440.

10. Oliver Asín, "En torno a los orígenes de Castilla," pp. 352-353, 378.

11. Hart, "Segmentary Systems," passim.

12. J. David Seddon, "Local Politics and State Intervention: Northeast Morocco from 1870 to 1970," in Arabs and Berbers, Ernest Gellner and Charles Micaud, eds. (London:Duckworth, 1972), pp. 122-123; Raphael Patai, The Arab Mind (New York: Scribner's, 1973), pp. 215-218, 226; Ernest Gellner, "Introduction," in Arabs and Berbers, p. 19.

13. Guichard, Al-Andalus, pp. 328, 354, 363, 531-532, 544-547, 275, 282, 20. On the grape-leaf incident, see R. Dozy, Spanish Islam: A History of the Moslems in Spain, Francis Griffin Stokes, trans., new impression (London: Frank Cass, 1972), p. 66, and comments by Guichard, p 325, and Patai, The Arab Mind, p. 217. On moieties in eleventh-century al-Andalus, see Urvoy, "Etude sociologique," p. 259. Guichard (p. 489) believes that this particular aspect of group feeling (as in the writing of pro-Mudari poems in the late tenth century) was an "aristocratic pretense," devoid of real social meaning.

14. Turner, Weber and Islam, p. 30; Guichard, Al-Andalus, pp. 487-488.

15. Guichard, Al-Andalus, pp. 18-19, 490.

16. Barbero and Vigil, Orígenes sociales de la reconquista, pp. 30, 171-172, 187.

17. Ibid., pp. 185, 194; cf. Guichard, Al-Andalus, p. 262.

18. Ramón Menéndez Pidal, La España del Cid, 5th ed., 2 vols. (Madrid: Espasa-Calpe, 1956), I: 94; Marc Bloch, Feudal Society, L. A. Manyon, trans. (London: Routledge and Kegan Paul, 1961), p. 138; El Fuero de Jaca, Mauricio Molho ed. (Zaragoza, C.S.I.C., 1964), p. 72.

19. Ruiz Domenec, "Una etapa en la ocupación del suelo," pp. 505, 507. On bilaterality, see Guichard, Al-Andalus, pp. 127, 141, and idem, "Les arabes ont bien envahi l'Espagne: Les structures sociales de l'Espagne musulmane," Annales, 29 (1974), 1496-1497.

20. Menéndez Pidal, España del Cid, I: 21; Barbero and Vigil, Orígenes sociales de la reconquista, p. 187.

21. Ruiz Domenec, "Una etapa en la ocupación del suelo," pp. 492, 507; Bonnassie, La Catalogne, I: 280-281. Bonnassie ("A Family of the Barcelona Countryside," p. 117) notes that large families (such as those which felt the economic disadvantages of partible inheritance) were not necessarily extended.

22. Bonnassie, La Catalogne, I: 288, and "A Family of the Barcelona Countryside," pp. 117-118. Guichard, "Les arabes ont bien envahi l'Espagne," p. 1,499, notes that endogamy was present in Christian Spain, but was never systemic.

23. Guichard, Al-Andalus, pp. 257, 474; Barbero and Vigil, Orígenes sociales de la reconquista, p. 194.

24. On marriages of the Banû Qâsi, see Guichard, Al-Andalus, pp. 231-234. In agnatic society, strong groups typically receive from, but do not give women to, alien groups.

25. Bloch, Feudal Society, p. 142.

26. Guichard, Al-Andalus, pp. 134-138.

27. Pedro Díaz Cassou was the first to notice the social implications of the Beni-names of secondary canals: Ordenanzas y costumbres de la huerta de Murcia (Madrid, 1889), pp. 54-55. See Glick, Irrigation and Society, p. 163. More recently, Julio Caro Baroja notes the same phenomenon as an enduring factor in the social organization of irrigation in Murcia and the Ebro Valley: "Cortes de Navarra. El Ebro como eje," Revista de Dialectología y Tradiciones Populares, 25 (1969), 76-79, and "Mundos circundantes y contornos histórico-culturales," ibid., 29 (1973), 45-46. See comment by Guichard, Al-Andalus, p. 420.

28. 28.Claudio Sánchez-Albornoz, "Pequeños propietarios libres en el reino asturleonés. Su realidad histórica," in Investigaciones y documentos sobre las instituciones hispanas, p. 178.

29. On free settlers, ibid., p. 179; and the same author, Despoblación y repoblación del valle del Duero, p. 283, and "Repoblación del reino asturleonés," pp. 304, 317.

30. Sánchez-Albornoz, "Pequeños propietarios libres," p. 185. Cf. Bonnassie, La Catalogne, I: 307, on collective ownership of pastures.

31. Poema de mio Cid, Colin Smith, ed. (Oxford: Clarendon Press, 1972), p. 102:

Quien nos darie nuevas de mio Qid el de Vivar
Fuesse a Rio d'Oyirna los molinos picar
e prender maquilas commo lo suele far
Quil darie con los de Carrion a casar?

Translation is that of Harold Sutherland in Menéndez Pidal, The Cid and his Spain, new impression (London: Frank Cass, 1971), p. 66. Maquila, an Arabism (from makîla, a dry measure for grain) was a due collected in kind; see Neuvonen, Arabismos del español, p. 41. Picar has the technical sense of dressing mill millstones in order to restore their coarseness; see below, Chapter 7, section 4(a).

32. Sánchez-Albornoz, Despoblación y repoblación del valle del Duero, pp. 286, 336-338 n. (mills), 285 (water); idem, "Repoblación del reino asturleonés," pp. 412 n. 41 (mills), 279-280 (water); idem, "Pequeños propietarios libres," pp. 183-184, 194; Gautier-Dalché, "Moulin à eau," pp. 339-340, 344, 346; Bonnassie, La Catalogne, I:307 (water).

33. Bonnassie, "A Family of the Barcelona Countryside," p. 112; García de Cortázar, San Millán de la Cogolla, p. 85. Cf. Carlé, "Migraciones de corto radio," p. 134, on size of hamlets in Toledo and Soria. A villa of forty men was considered unusually large; Sánchez-Albornoz, Despoblación y repoblación del valle del Duero, p. 341 n. 139.

34. García de Cortázar, San Millán de la Cogolla, pp. 228-229, 234; Luis G. de Valdeavellano, "Las instituciones feudales en España," in F. L. Ganshof, El feudalismo (Barcelona: Ariel, 1963), p. 299; Sánchez-Albornoz, Despoblación y repoblación del valle del Duero, p. 320. On duties of peasants with emphyteutic contracts, see idem, "Repoblación del reino asturleonés," p. 321.

35. García de Cortázar, San Millán de la Cogolla, p. 230 n.111.

36. Sánchez-Albornoz, "Pequeños propietarios libres," p. 178.

37. On homini excusi, see Sánchez-Albornoz, Despoblación y repoblación del valle del Duero, p. 340; Moreta, San Pedro de Cardeña, p. 162; and Carlé, "Migraciones de corto radio," pp. 131-132.

38. Carlé, "Migraciones de corto radio," pp. 125-129.

39. Ruiz Domenec, "Urban Origins of Barcelona," pp. 284-285. Sánchez-Albornoz, "Repoblación del reino asturleonés," p. 426; García de Cortázar, Epoca medieval, pp. 224, 273, 276, and idem, San Millán de la Cogolla, pp. 229, 236. See also, Pallares and Portela, Bajo valle del Miño, p. 31, and Reyna Pastor de Togneri, Del Islam al cristianismo: En las fronteras de dos formaciones económico-sociales: Toledo, siglos XI-XIII (Barcelona: Peninsula, 1975), p. 108.

40. Sánchez-Albornoz, "Las behetrias: La encomendación en Asturias, León y Castilla," in Estudios sobre las instituciones medievales españolas, pp. 74-83, 87, 92.

41. See George C. Homans, English Villagers of the Thirteenth Century, reprint ed. (New York: Russell and Russell, 1960), pp. 235-239.

42. Pastor, Conflictos sociales, p. 64.

43. Ibid., p. 191.

44. Hussain Mones, "Le role des hommes de religion dans l'histoire de l'Espagne musulmane jusqu'à la fin du Califat," Studia Islamica, 20 (1964), 56.

45. See Guichard, Al-Andalus, p. 114.

46. J. M. Lacarra, "La reconquista y repoblación del valle del Ebro," in La reconquista española y la repoblación del pais, pp. 71, 78; Burns, Islam under theCrusaders, pp. 102-104; González Palencia, "Riegos de Veruela," p. 83; Pastor, Del Islam al cristianismo, pp. 49, 75. Pastor asserts that the Neo-Muslim peasant as sharîk was better off than his predecessor under the Goths and that this explains the impetus to conversion. But this kind of mechanism was only marginally involved in the conversion process (see Chapter 5, section 4, below).

47. Chalmeta, Señor del zoco, p. 107; Oliver Asín, "En torno a los orígenes de Castilla," p. 353.

48. Sánchez-Albornoz, "Las behetrías," pp. 69-73, 85-86. In some cases, the maulado had to pay his own wergeld; in others, the lord paid, giving the impression that some maulados were technically more free than others.

49. Lévi-Provençal, España musulmana. Instituciones y vida social, pp. 117-118.

50. Valdeavellano, Sobre los burgos, pp. 74, 83, 117, 129, 131.

51. Pastor, Conflictos sociales, pp. 106, 115, 179.

52. Pastor, Del Islam al cristianismo, p. 46.

53. This is the conclusion reached by Teofilo Ruiz in a series of important articles on the bourgeoisie of Burgos (see n. 67, below). The notion that Castilian cities lacked a middle class is also attacked by José Ignacio Gutiérrez Nieto, "La estructura castizo-estamental de la sociedad castellana del siglo XVI," Hispania, 33 (1973), 532-533, who notes the development of secondary and tertiary sectors in even small towns by the late middle ages.

54. Goitein, in particular, is insistent on the role of merchants as bearers of Islamic culture: Studies in Islamic History, pp. 215, 219. On the impact of the commercial milieu on law, see ibid., p. 243.

55. Ibn al-Zubair, Silat al-Sila, E. Lévi-Provençal, ed. (Cairo, 1938), nos. 11, 13; ibn al-Abbâr, Kitâb al-Takmila li-kitab al-sila, Francisco Codera, ed. (Madrid, 1887), no. 1810. The function of religious education became that of confirming high socio-economic status, or of achieving it; see Bulliet, Conversion to Islam, chapter 11.

56. Mones, "Hommes de religion," pp. 69-74.

57. Urvoy, "Etude sociologique," p. 238.

58. Abdel Magid Turki, "La vénération pour Mâlik et la physionomie du Malikisme andalou," Studia Islamica, 33 (1971), 58-62.

59. The tracing of scholarly genealogies is Urvoy's method of analysis in the study cited in n. 57. Cf. also Figure 3 in Chapter 7, below.

60. A. I. Rozi, work in progress.

61. Bonnassie, La Catalogne, I: 495; Ruiz Domenec, "Urban Origins of Barcelona."

62. Valdeavellano, Sobre los burgos, pp. 59, 64, 68, 131; Ubieto, Ciclos económicos, p. 128.

63. Lacarra, "Repoblación de las ciudades en el camino de Santiago," p. 469.

64. Represa, "Genesis y evolución urbana de la Zamora medieval," p. 535; Pastor, Conflictos sociales, pp. 75-76; Lladonosa Pujol, "Proyección urbana de Lérida," p. 197.

65. Pastor, Conflictos sociales, pp. 34-35, 58, 86.

66. See Sánchez-Albornoz, "Primitiva organización monetaria," p. 474, and Espafia: Un enigma histórico, II: 90.

67. Teofilo Ruiz, "Prosopografía burgalesa: Sarracín y Bonifaz," Boletín de la Institución Fernán González, 185 (1975), 467-499, and "The Impact of the Conquest of Seville on Castilian Society, 1248-1350," typescript.

68. Bonnassie, La Catalogne, I: 493; Justiniano Rodríguez Fernández, La judería de la ciudad de León (León: C.S.I.C., 1969), pp. 29, 57.

69. Nilda Guglielmi, "Cambio y movilidad social en el cantar de Mio Cid," Anales de Historia Antigua y Medieval, 12 (1963-65), 47; cf. Sánchez-Albornoz, Espana: Un enigma histórico, I: 290. See also García de Cortázar, Epoca medieval, p. 272.

70. This assertion, somewhat hyperbolic, is that of Pastor, Conflictos sociates, p. 49.

71. Guglielmi, "Cambio y movilidad social," p. 44, is correct in insisting that taxes and exemptions were the key to distinctions (both real and legal) between the classes.

72. See Sánchez-Albornoz, "Repoblación del reino asturleonés," p. 426, and España: Un enigma histórico, II: 88.

73. Gual Camarena, "El hierro en el medioevo hispano," p. 279; García de Cortázar, San Millán de la Cogolla, p. 253; Pallares and Portela, Bajo valle del Miño, pp. 28-29; Ubieto, "Los 'Votos de San Millin,'" p. 324; Pastor, Conflictos sociales, p. 55.

74. Bonnassie, La Catalogne, I: 283ff., 318; Sánchez-Albornoz, "Repoblación del reinoasturleonés," pp. 402, 404-405, 407; García de Cortázar, Epoca medieval, p. 277.

75. Sánchez-Albornoz, Estampas de vida en León, pp. 78-79, and España: Un enigma histórico, II: 52; Pastor, Conflictos sociales, pp. 164, 189-90. Cf. Valdeavellano, Sobre los burgos, p. 156, on the concentration of caballeros in frontier towns.

76. Sánchez-Albornoz, Despoblación y repoblación del valle del Duero, p. 320: Damus foro bonos ad illos caballeros, ut sint infanzones (974); idem, España: Un enigma histórico, II:49, citing Poema de mio Cid, lines 1,212-1,215:

quando mio Çid gaño a Valencia                     y entro en la çibdad;
los que fueron de pie cavalleros se fazen,
el oro e la plata quien vos lo podric contar?
Todos eran ricos quantos que alli ha.
Poema, Colin Smith, ed., pp. 39-40.

77. José M. Font y Rius, "La reconquista y repoblación de Levante y Murcia," in La reconquista española y la repoblación del pais, P. 104; Torres Fontes, Repartimiento de la huerta y campo de Murcia, pp. 86, 97; idem, "Jaime I y Alfonso X," p. 333.

78. Pierre Guichard, "Un seigneur musulman dans l'Espagne chretienne: Le 'Ra'is' de Crevillente (1243-1318)," Mélanges de la Casa de Velazquez, 9 (1973), 323.

79. Menéndez Pidal, Espaga del Cid, I: 31, 170, 176-178, 279-285; on the point generally, see Valdeavellano, "Instituciones feudales en España," pp. 253-254.

80. Dominique Urvoy, "Sur 1'évolution de la notion de gihâd dans l'Espagne musulmane," Mélanges de la Casa de Velazquez, 9 (1973), 339.

81. Burns, Islam under the Crusaders, p. 310. Burns also alludes to the well-known practice of Andalusi knights outfitting themselves in the Christian style; earlier, the Christian captors of Zaragoza had arrayed themselves in Muslim style. In both cases, there may have been motives for emulation related as much to perceptions of superior military technology as to style and social prestige. For a Muslim noble family in the Christian feudal structure, see Guichard, "Un seigneur musulman," passim; such arrangements, though socially congruent, were transitory.