1. Stability and Continuity

[194] The standard periodizations of medieval history are based on political changes (especially cataclysmic ones) which tend to stress discontinuity between periods: the emergence of the Caliphate, and its fall; the emergence of the Party Kingdoms, and their decay and conquest in al-Andalus; dynastic shifts and the transit of hegemony among various kingdoms; and the great transitional dates of the wars of conquest (1085, the fall of Toledo; 1212, Las Navas de Tolosa) in Christian Spain. Such are the biases of traditional historiography, which saw the structure of future political forms presaged in earlier events. But comparative analysis tends to blur the profile projected by political peaks and valleys, and to stress phenomena which bespeak continuity between periods, and among different societies or cultures to identify and compare their basic components. On a diachronic plane, societies are seen to undergo gains and losses in structure-processes of crystallization (or decrystallization) of social and cultural forms which play themselves out over long periods of time, frequently spanning the benchmarks of political change.

In general, the high middle ages in both al-Andalus and in Christian Spain were a period of crystallization, both social (the subject of this chapter) and cultural (Chapter 9), a period of continuity in the elaboration of social structure from rudimentary and primitive forms in the eighth century to the relatively sharply crystallized societies of the twelfth and thirteenth centuries. By "structure," however, we mean (following Thomas Smith) something more general than the specific building blocks and linkages of social networks. It refers, rather, to the "relative invariance or fixity of relationships among the elements of a system," whatever those elements may be. Stability, in this sense, implies the relative permanence of arrangements regulating the distribution of resources within the society: the greater the permanence, the more crystallized is the social or cultural network. The opposite process -- decrystallization -- refers to structural loss, to the loosening of social arrangements, or to the redistribution of resources.⁽¹⁾ We can describe, in the period under consideration, processes [195] both of crystallization and decrystallization on either side of the cultural frontier. From an analytical standpoint, however, it happens that processes of crystallization and decrystallization can occur simultaneously within the same society. What a historian sees depends (as suggested in the Introduction) on what processes are viewed as normative.⁽²⁾ Thus, Gibbon viewed the decline of the Roman Empire as a decrystallizing process, whereas modern historians have detected impulses of recrystallization in the same period.

With near universality among historians, al-Andalus has been regarded as an unstable political entity, indeed as a society one of whose characteristic features was an inability to stabilize itself. Castro's judgment may be taken as representative: "Islam was incapable of creating stable political systems based on something other than a common religion and dictatorial rule; the totally religious character of Muslim life hindered the creation of secular forms of communal existence." Equally categorical is García de Cortázar: the Muslims never established "a stable political consensus"; the fall of the Caliphate is represented as the breaking of the "fictitious equilibria" upon which the state was built. Sánchez-Albornoz accepts the divisive nature of Andalusi politics as given, attributing this characteristic to "Spanish pride."⁽³⁾ Examples could be extended indefinitely, drawing upon many recent medievalists.

Three kinds of problems arise from this characterization of the Andalusi state. First, there is a perceptual, valuative problem. Those who so characterize al-Andalus do so on the basis of an implicit model of sound political structure, against which this particular society is judged negatively, as falling short of some norm. Second is the nature of stability itself. What constitutes a stable political system? Indeed, is political stability, however defined, commensurate with or a prerequisite of social cohesion? This suggests a third order of analytical problem, that dealing with the social or social-psychological corollaries of political stability. Political instability seems to suggest that the society is flawed or imperfect. The search for the key to this imputed flaw leads to even more categorical generalizations: that the society is unable to generate sound political forms, owing to some structural, psychological, or even genetic defect.

The first problem is placed into clear focus by Guichard: there has been an excessive tendency on the part of historians to treat al-Andalus as a coherent political entity, something like a national state on the European model, when in fact it was a segmentary, tribal society, at least in its [196] early centuries.⁽⁴⁾ In such a social system, instability at the dynastic level is normal and expected because tribal political life takes the form of a constant testing of forces. This process has the result of forming continually shifting equilibria which, if "unstable" judged by a criterion that places positive value on the long-term continuity of a central government or ruling institution, are not, for that, fictitious. Their reality is the reality of segmentary organization.

Stability of social organization and continuity in its institutions were very much a part of Andalusi society, but their mainsprings did not lie in the ruling institution. In the first place, tribal organization in itself must be viewed as a stabilizing force. A number of historians have stressed the positive functions of tribal fragmentation. Guichard, in discussing the constant state of rivalry among groups, notes that the principle of opposition was essential to the equilibrium of Arab society. Such opposition was perpetually readjusted, creating equilibria between hierarchized agnatic groups.⁽⁵⁾ Burns, in commenting on the survival of Islamic society in Valencia after the Christian conquest, notes that "The paradox is that a country so rent by factionalism as to fall before the crusaders, piece by piece, was by reason of that very fragmentation tough enough to survive at its lower levels."⁽⁶⁾

In the eighth and ninth centuries, political success was tribal success, measured in tribal terms.⁽⁷⁾ Guichard documents this when he shows that all the political activities of the Magila Berbers, for example, had a tribal basis, from which all power was ultimately derived. The same was true of other groups, from the political turnovers of the early governors (the actions of the powerful Fihrî family are a case in point) to the Umayyads themselves. 'Abd al-Rahmân I built a civil and military administration staffed almost wholly by his own kinsmen and clients. Ibn Khaldûn notes that it was his dynasty's hallmark that it bestowed office, particularly judgeship, upon those who shared in the dynasty's 'asabîya or "group feeling." During the fitna of the late ninth century, 'Abd Allâh was reduced to the support of the same group which provided the regenerative nucleus for the inception of the Caliphate. The slave regimes of the early Party Kingdoms proved too unstable to survive, lacking a tribal base; these were soon eliminated by the Arab and Berber elites who founded states based on family and clientage relationships.⁽⁸⁾

The nature of Islamic law, which regulated daily as well as religious affairs, introduced a standardization or uniformity of governance and of [197] social behavior which meant that political boundaries could change kaleidoscopically without interrupting the relative stability of public order on lower levels, particularly that of the town. (In Christian Spain, a role parallel to that of Islamic law was played, on one level, by urban law, which was diffused by the process of the territorialization of the great Fueros, and, on a higher level, by the diffusion of Roman law.) The political balance in all Islamic states of the middle ages was precarious. Princes rose and fell and tribal geography shifted constantly, but the basic structure of the society was untouched. As Goitein sardonically observed, the breakdown of public order was regarded as something akin to a natural phenomenon.⁽⁹⁾

The law was dominated by middle-class values and concerns and interpreted by that bourgeois element par excellence, the faqîhs. We have previously noted the ambivalence of the fuqahâ' with regard to the ruling elites: on the one hand, the desire not to be compromised by power; on the other, the necessity of playing a public role in the interests of enforcing the dictates of religious law. In sociological terms, the legalism of medieval Islam was the principal means of social integration, at least in urban society. The characteristic social division of power between a literate urban elite and an illiterate, rural, tribal mass explains the social function of the fuqahâ'. The jurisprudential class, as interpreters of Islamic law, served to mediate between the ruling institutions and the tribal masses to interpret the law for the latter and, through the law, to provide a religious sanction or check upon the excesses of rulers.⁽¹⁰⁾

The fuqahâ' always appeared to support the status quo, providing the ruling elite with the legitimation of Islamic law, as a force to countervail political instability. The state, moreover, used the fuqahâ' and religious institutions such as the judgeship (qadâ) to ensure order and to replace, in effect, tribal relationships in an urbanizing society. Thus one gains the impression that qadis, on a day-to-day basis, were actually governing the towns, whatever the "civil" administrative structure may have been. From this web of authority in which the fuqahâ' were thoroughly enmeshed comes the prevalent historical judgment (as in Lévi-Provençal) of Andalusi Islam, embodied in the jurisprudential class, as having been rigid and conservative, indeed deliberately so, a judgment which, as Urvoy indicates, misconstrues the nature of Islamic law. In theory, Islamic law -- whether Maliki or not -- was rigid, but in practice it was fluid, malleable, and not inflexibly systematized.⁽¹¹⁾Al-Khushanî repeatedly shows judges [198] interpreting the law pragmatically, using precedents and analogies -- in sum, a customary legal system, with informal elements superimposed upon the canonical sources.

In spite of the well-known disinclination of many prominent fuqahâ' to fill public positions -- to decline the judgeship was a sign of moral probity -- members of this class were in fact deployed with amazing fluidity to staff the juridico-legal bureaucracy (as opposed to the royal household). Mones points out that the Umayyads controlled the corps of officials drawn from the faqîh class by continual juggling of administrative posts. Judgeships turned over rapidly, and fuqahâ' --frequently the same individuals -- held the posts of muhtasib and sâhib al-shurta, often in different towns. Provincial qadis were not allowed to become too powerful, and many biographies contain the phrase: "He was called to Córdoba" -- not necessarily a promotion, in Mones' view.⁽¹²⁾

Serial judgeships, whereby a faqîh would move from one town to another after a few years of service to hold the judgeship there, became characteristic of Andalusi administration, from early Umayyad times through the Almohad period. Al-Khushanî mentions several ninth-century faqîhs who held provincial judgeships before coming to Córdoba, and later the pattern became more pronounced. Of the forty-three qadis mentioned in ibn al-Zubair's biographical dictionary (twelfth and thirteenth century, but which reflects the practice of the Taifa period as well) fourteen (thirty-two percent) held the judgeship in two or more towns, indicating a measure of organizational continuity over the entire country.⁽¹³⁾

That this period typified the Taifa, as well as the Umayyad, period demonstrates that political control by a centralized state was not the only function of serial judgeships. Taifa-period judges crossed political boundaries as if they did not exist. It is this feature of the administration of Islamic law that gave stability to the social system in spite of political instability and shifting political boundaries. The contrast between the permanence of the juridical system and the impermanence of the political system demonstrates the coexistence of a highly crystallized social network involving the urban middle class on the one hand, and, on the other, a military and ruling elite which betrayed many of the signs of lack of structure inherent in the steady breakdown of tribal organization.

In later times one notes that, because of the peculiar historical situation caused by the increase of Christian military pressure, the fuqahâ' increased their social weight just when the system of Islamic law was failing to [199] fulfill its prime role as the glue which held the polity together. Ibn Khaldûn records that the Almoravids honored the fuqahâ', who "Were appointed to the council, everybody according to his influence among his people in his respective village."⁽¹⁴⁾

The height of their social and political prestige was reached, however, just as Christian pressure became intolerable. A measure of structural loss was the emergence of Sufism, which, as a religious movement, was diametrically opposed in tonality and spirit, to the rule of fiqh and its interpreters (and which also had different social roots).⁽¹⁵⁾ The failure of fiqh in the waning days of Andalusi society was, of course, the ultimate instability, a disaggregation of authority at the level which had, in the past, provided the greatest source of continuity. But this collapse was provoked from without, and was beyond the society's ability to control.

A measure of social stability is the ability of the prevailing social model to predict accurately the real division of resources. In al-Andalus, tribal affinity remained an accurate predictor of socio-economic status until the end of Islamic settlement. In the sixteenth century, when Arab names were outlawed and Muslims were forced to take Christian names, Moriscos complained that they were hampered in the pursuit of normal social and economic relations (marriage, commerce), lacking the security of knowing the lineage of the other individuals. One of the objectives of the second Alpujarras revolt was a reestablishment of tribal deference patterns through the reassumption of tribal names.⁽¹⁶⁾

But clearly there were other indicators of deference besides tribal ones. The related processes of conversion, urbanization, and detribalization created a dual structure whereby, for example, the urban oligarchy that comprised the faqîh class after the fall of the Caliphate controlled a substantial amount of resources, both power and wealth, without any meaningful participation in a tribal structure. When Mu'tamid of Seville noted in 1086 that the Andalusis formed peoples and not tribes, he was recognizing that the basis of deference had shifted and that the traditional social model, the tribal one, was no longer an accurate predictor of the real distribution of preferment. Ibn Khaldûn echoed this perception later, remarking that "group feeling" had been lost in al-Andalus, which was no longer the home of groups and tribes.⁽¹⁷⁾

In Christian Spain, feudal canons insisted upon a detailed schedule of deference. The Catalan hierarchy (see section 4, below) is a case in point: wealth increased as one ascended the scale from sotcastlà to count. [200] Another example is the way in which land was distributed in conquered Murcia in a descendant scale of value to the three classes of knights (caballeros mayores, medianos, and menores) and three classes of peasants (peones mayores, medianos, and menores).⁽¹⁸⁾ These were standard categories and were good predictors of real social rank, indicative of the permanence of these relations. The growth of a non-feudal, urban middle class created within this structure islands where titles to deference were differently based (although, as noted, burghers sought to assimilate themselves to the noble schedule). Only in the sixteenth century did the feudal structure of deference become an invalid predictor, as non-noble sectors controlled progressively more resources than the downwardly mobile hidalgos.

2. Structural Gains and Losses

At points when there is disjunction between the inherited social model and the actual configuration of society, the time is ripe for the emergence of new social forms, a time of experimentation symbolized in broad stylistic changes. In Smith's view, the lack of correspondence between the cultural model and the actual social structure is typically caused by decrystallization of social networks, by structural loss.⁽¹⁹⁾ It seems logical, though, that structural gain could also underlie such a transition. There were two periods in Andalusi history that produced stylistic revolutions of the dandy type, such as that of nineteenth-century England (Beau Brummel) that Smith takes as the typical case. Brummel's epoch was clearly one characterized by structural loss. Of the two such periods in Spanish Islamic history, the reign of 'Abd al-Rahmân II and Party Kingdom interlude, the second was a consequence of a decrystallization/crystallization cycle in which the fall of the Caliphate was a direct antecedent.

The reign of 'Abd al-Rahmân II (822-852) is more difficult to assess. Various contemporaneous phenomena, summarized in Table 4, can be placed in a gain column; others seem better characterized as loss. The epoch is remembered as one of stylistic revolution, associated with the introduction of 'Abbâsid-Persian institutions and fashions, the latter associated with the singer and courtier, Ziryâb, the kind of figure identified by Smith as an entrepreneur of style. The facts of Ziryâb's career are well-known. A black singer born in Iraq in 798 whose real name was Abû'l-Hasan 'Ali ibn Nâfi', he fell from favor at the court of Hârûn al-Rashîd [201] in Baghdad and arrived in al-Andalus just before the beginning of 'Abd al-Rahmân II's reign. Like Brummel later, he became an arbiter of style. He introduced Iraqi singing styles, created a music school, made changes in the form of the lute. He introduced toothpaste, popularized shaving among men, and decreed tonsorial styles. He taught Iraqi cooking and lent his name to a number of recipes. He influenced political decisions, such as the appointment of judges by the Emir, and his advice was sought on virtually all issues.⁽²⁰⁾ More important from our point of view than Ziryâb's innovations in fashion are the structural changes introduced in the administration of the Emirate and their significance in terms of changing social structure. It was said of 'Abd al-Rahmân II that he "organized the norms of the state."⁽²¹⁾

The initial administrative model in Islamic, as in Christian Spain, was a Roman one, in some instances modified by Visigothic practice. Livermore's characterization of early Islamic rule as a Neo-Roman system, while perhaps overlooking those aspects peculiar to segmentary societies, is nevertheless [202] suggestive. The wali (governor) was the equivalent of the late Roman dux, and there is a correspondence of the early division of Islamic kûras with Visigothic administrative districts. The Andalusi concept of the kûra seems to have been colored by Roman usage. In the East, provinces were conceived as units of agricultural land, following Persian administrative norms. In al-Andalus, administrative districts were centered in and consubstantial with cities and their hinterlands: there were very few wholly rural kûras.⁽²²⁾

The early governors had followed Syrian Umayyad practice by organizing the treasury, the army, and the administration of justice along territorial lines, which created a decentralized, tribally oriented polity wherein the vast Christian population was ruled by Muslim garrisons housed for the most part in towns. 'Abd al-Rahmân II remodeled governmental institutions along 'Abbâsid lines. He strove to concentrate power in the person of the emir, to establish a secure financial base for government operations by founding state monopolies (over coinage and textile production, for example), by bringing order and division of labor into the hierarchy of public officials (introducing new court officials based on eastern, ultimately Persian models -- the hâjib and wazîr -- and by establishing greater control over urban affairs. This last involved the introduction of a variety of 'Abbâsid officials, whose administrative authority ultimately derived from the qadâ, such as the muhtasib (a market master with widened jurisdiction and responsibility) and the urban prefect (sâhib al-madîna).

In this instance, dandyism seems tied to a crystallization process, whereby tentative arrangements based on the circumstances of conquest were yielding to more complex social needs, requiring an appropriate administrative response. One can find in the crucial process of conversion both crystallization and decrystallization, depending on whether one chooses the vantage point of tribal society or of the masses. For the former, urbanization and the conversion of masses of people who fell outside the tribal structure implied a loss of structure in the traditional social networks. The more Neo-Muslims in the body politic, the less was tribal status a good indicator of the allocation of resources. Dandyism and administrative innovation were positive responses to a new social situation. The massacre of Christians and, later in the century, the terrible tribal fitna (perhaps the last gasp of the inherited social model of a tribally based polity) were also responses to a process of structural loss. From the point [203] of view of Neo-Muslims, however, conversion and all that it entailed were part of a crystallization process. The increasing mass of Muslims called for new institutions and social and cultural responses designed to meet their needs, particularly the creation of a society that was culturally and institutionally Islamic, rather than tribal in nature. We can associate the establishment of Malikism and the beginnings of a native school of hadîth with these impulses. The fact that Malikism had full official support beginning only in the reign of 'Abd al-Rahmân II substantiates this conclusion.⁽²³⁾

The second crucial juncture, that beginning with the failure of the centralized state created by 'Abd al-Rahmân II, had similar characteristics but with the signs changed. Dandyism, in similar guise of the aping of Iraqi customs which so characterized the Taifa courts, would seem more clearly part of a decrystallizing process. Structural loss and gain seem to be paired and sequential. In the ninth century, the coalescence of the centralized state (gain) initiated a wave of conflict throughout the tribal network (loss). In the eleventh, the fall of the Caliphate (loss) initiated a reorganization of the society in decentralized polities which seemed to respond well to economic and ethnic realities (gain).

The historiography of this episode is reminiscent in many respects of the catastrophic approach to the fall of the Roman Empire. The demise of the Caliphate has been represented as just such an unmitigated disaster, with no redeeming features. One exponent of the catastrophic view, Hussain Mones, asserts that the dismemberment of the Caliphate "not only surprises the Arab reader, but saddens him and fills him with disgust through considering the so-called Kings of Taifas as criminals who destroyed the unity of the Caliphate and who, naming themselves emirs and kings, sacrificed the common interest in order to satisfy their vain, egotistic, and sterile pride."⁽²⁴⁾ Besides the valuative judgment which virtually asserts that the Caliphate was good and its successors evil, there is also present in Mones' analysis the notion that great political events come about through the conscious will of leaders. Elsewhere he states that in deciding to abolish the Caliphate, the Andalusi aristocracy had demonstrated a complete lack of political clarity.⁽²⁵⁾ James Dickie, whose conclusion as to the significance of the fall of the Caliphate is the opposite of Mones', still lays equal emphasis on the conscious motivation of the actors. Thus he states that the aristocracy "had come to terms with reality by abolishing a meaningless institution, doing what they did in sorrow and [204] sensible of the implications of their action."⁽²⁶⁾ Interestingly, al-Shaqundî, a thirteenth-century writer, asserted that the most illustrious subjects at the time of the dismemberment of the Umayyad state were unanimous in favoring the division of the Caliphate into Taifas because it led to the revvitalization of high culture.⁽²⁷⁾ Masked in al-Shaqundî's observation is his awareness that the social model no longer fit the social reality. Thus he could view the Taifas positively.

All who discuss the fall of the Caliphate -- al-Shaqundî excepted -- seem convinced that the reasons for it must lie in supposed deficiencies of the ruling Arab caste. W. M. Watt, for example, singles out "the absence of a middle class interested in maintaining a strong central government" as the prime structural deficiency leading to the collapse of the Caliphate.⁽²⁸⁾ But it is by no means clear that a middle class is a prerequisite of a stable government, just as it is not certain that shortsighted decisions by a few individuals can force the restructuring of an entire society.

The systemic nature of intergroup relations implies that the sources ot a radical social shift represented by the fall of the Caliphate must be sought, not in the structure of one group, but in the relationship among all groups. The Caliphate fell because the ethnic equilibrium which had summoned the edifice into being was disturbed. The essence of the shift was that the Berber minority had grown dramatically in size and power just at the moment when the peak of the curve of conversion was reached.

The increment in Berber immigration was the result of military reforms carried out by the emirs as they consolidated power, beginning in the ninth century under al-Hakam I and 'Abd al-Rahmân II. At each successive stage, new African contingents were inscribed in the military register. During the reign of 'Abd al-Rahmân III, the settling of new Berber groups reached even greater intensity as the Caliph sought to balance the influence of restive Arab rivals.⁽²⁹⁾ At the same time another process, described by Lévi-Provençal as the fusion of Arabs and NeoMuslims, but more accurately the swamping of the former by the latter, was nearing completion, with the peak of conversion probably reached in the first quarter of the tenth century.

The political fragmentation of al-Andalus, therefore, was not so much a fall from grace as a realignment in keeping with a shift in the ethnic balance of power. The Berbers were strong enough to disrupt the central organization of the state, but not to capture that mechanism intact. But [205] then, the elaborate civil-military apparatus needed to manage ethnic and tribal cleavages had become obsolete, once the vast majority of the population was composed of ethnically homogeneous, more or less non-tribal Neo-Muslims. A new equilibrium was sought in the political organization of regional nuclei, in accordance with realistic demographic patterns.

The dismemberment of the Caliphate, once the fitna itself had subsided, did not materially disrupt the continuity of life in al-Andalus. (We have noted recent speculations that political regionalization was conducive to economic growth.) For reasons already stated, the locus of such continuity was, in rural areas, in tribal structures, and in the towns, in the jurisprudential class whose methods of training, recruitment, and political action were unchanged by the fitna. Then too, by the twelfth century at the latest, the majority of the fuqahâ' must have been Neo-Muslims, and the society was culturally and socially more homogeneous -- even though politically fractionated -- than it had ever been under the Umayyads.

It is against this background that the eleventh-century transition must be considered. Again there is a strong stylistic component; again it is a component of Iraqi inspiration. The Taifa courts, as García Gómez indicates, strove to become "microscopic Baghdads." The problem was that, unlike the Emirate of the ninth century, the eleventh-century experience was less a creative integration of useful innovations with broad social demand than it was the imposition of superficial styles and customs that had no resonance beyond the kingly courts themselves. Thus the Party Kings adopted lofty caliphal titles typical of 'Abbâsid usage (e.g., al-Muzaffar, "The Victorious"; al-Muqtadir, "The Powerful"), together with courtly pretensions that became ludicrous when reproduced on the petty scale of the period. Smith points out that the use of empty or vestigial titles which reveal nothing of the real lives and resources of their bearers (the antiquation of biographical models, he calls the phenomenon) is a sign of disjunction between model and reality. What is peculiar in this instance is that the titles, although historically vestigial, were deliberately introduced as part of a restructuring process. The feeling that Iraq was the center of all culture seems to have been reinforced in the eleventh century by tremors of insecurity permeating a society that lived in perpetual fear of conquest. In Urvoy's view, this insecurity explains the alienation of the intelligentsia after the fall of the Caliphate, as a consequence of which heterodox religious and philosophical movements began [206] to appear on the margins of society.⁽³⁰⁾ The Taifa period seems characterized by the simultaneous play of decrystallizing forces in the cultural sphere and crystallizing ones in the political and economic spheres.

A similar phenomenon, where social restructuring was symbolized by stylistic changes, is the instauration of Neo-Gothicism in the Asturo-Leonese kingdom, a movement which alludes to the restoration by Alfonso II in ninth-century Oviedo of the Visigothic model of kingship (and which is not to be confused with any artistic movement). This movement had only tenuous links with the Germanic past and was associated with the attempts of Alfonso II and the Bishop Pelayo to reorganize the civil and ecclesiastical organization of the kingdom. Pelayo wanted to establish Oviedo as a bishopric and also to establish continuity between Visigothic and Asturian sees. In civil administration, palace officials were assigned the lofty titles of Visigothic magnates, but were in fact absurd parodies. For example, the Notary Count had been an important official in the Visigothic court, but his counterpart in Oviedo was a simple scribe. The form or style of the reform was more significant by far than its substance, since it was impossible to recreate a peninsular-sized administrative apparatus in a small mountain kingdom with few financial resources.⁽³¹⁾

According to Sánchez-Albornoz, the wholesale adoption of Germanic norms (replacing late Roman ones) was well-suited to the rude psyche of these rural mountain communities.⁽³²⁾ However, this kind of stylistic change is more likely related to socioe-conomic changes than to group psychology. Signs of structural loss are clear in the economic crisis of the mid-ninth century, when Alfonso II replaced the gold standard with silver, causing devaluation and a price rise. Since Neo-Gothicism was in part designed to create a myth of continuity, the discontinuities that brought it into being are difficult to perceive. At the end of the century, under Alfonso III, when the kingdom was much expanded beyond its initial minute radius, it became clearer that a more complex kind of administrative response was necessary to manage a society which had grown and consolidated itself.⁽³³⁾

To return now to Smith's model: it is clear that stylistic changes affecting the development of society broadly are indeed reflections of points in socio-cultural evolution at which the prevailing model of social structure does not accord with what the structure of society really is, or has become. From our point of view, however, style changes that are reflected [207] in administration, in norms of governance, are historically more significant than those of fashion only, and more difficult to explain. While structural loss (or gain) may call forth a demand for social innovation that is broadly based, administrative changes are made by a relatively small circle of elite persons. To interpret the significance of changes in administrative styles to discover what social changes called them into being, is to realize that crystallization or decrystallization may be taking place in social networks outside those which organize political life. The seeming anomaly of governments (as in ninth-century Asturias or eleventh-century alAndalus) adopting a hollow model merely demonstrates that aristocracies strive to fill voids in leadership, using whatever tools they have at hand. In some cases, the tool is cultural borrowing; in others, it may consist of the revival of an anachronistic model, in the hope that new life can be breathed into it; or an entirely new set of rules may be invented.

To turn briefly to cultural borrowing in administrative forms, and to return to the question, raised in the Introduction, of the controlling nature in cultural diffusion of the disparity in structuring between contacting societies: it appears that, at various junctures, when Christian and Islamic societies contacted, new social realities were created for which new models of organization had to be derived. In certain cases, the results were exaggerated and transitory: a case in point is the aping of Arab customs by the Christian conquerors of Zaragoza, who donned Muslim garb and played at being shaykhs for a time. But behind this passing experiment, which constituted a form of social modeling, there was a demand established for the borrowing of administrative practices by the less highly structured Christian society from the more highly structured society of the Muslims. Thus Julián Ribera's theory that the justicia of Aragón was an imitation of the Islamic appeals judge (mazâilim) has meaning if considered against this background. There was ample precedent in the Christian world. In León and Castile, the Andalusi sâhib al-sûq, later muhtasib, was adopted (as zavazoque, almotacén) at a time when the Christian states lacked the administrative sophistication of the more complex contacting society. The same was true in the thirteenth century, when Christian control of medical practice among subject Muslims and Jews was highly conditioned by the Muslim model of open access to medical practice. Another example is the Christian assimilation of the dhimma model for dealing with ethno-religious minorities, a clear example of the migration of a model of social organization from a socially more complex society to [208] one which had no ready-made solutions for such a problem.⁽³⁴⁾ Such contacts, in sum, created a situation in which the less-structured society perceived a sense of disjunction with the contacting society. This perceived lack of fit is one of the primary forces creating demand for cultural borrowing, just as the internal evolution of one society can create a demand for new social models.

3. State Systems of the Eleventh Century

The eleventh century, in both peninsular societies, was a period of political consolidation. In Christian Spain the political entities that were to direct the southward expansion, the kingdoms that would share political power for the rest of the middle ages, emerged at this time. Frontiers between kingdoms, spheres of influence or of politico-economic objectives with regard to al-Andalus, and norms of administration began to acquire coherent form.⁽³⁵⁾ In al-Andalus an analogous process unfolded, involving the development of states whose rationale was a curious combination of territoriality and tribalism. In both cultures, the emergence of these states presupposed the crystallization of subcultural entities in distinct regional foci, a process which led to the emergence of culturally distinctive political entities in Christian Spain but which was aborted in al-Andalus by attrition of territory.

The state system of the Party Kingdoms was founded on a paradox. On the one hand, as Guichard indicates, the concept of the territorial state dawned only slowly in the Islamic world. Political units, since they were ultimately defined by patterns of tribal relationships, lacked precise boundaries. Indeed, there was no interest in establishing such lines of demarcation. (The early Arabic chronicles refer to the Asturian kingdom simply as "the Banû Adfansh"-the sphere of influence of the "sons of Alfonso," conceived tribally.) On the other hand, the ongoing processes of urbanization, detribalization, and conversion meant that the Andalusi polity was steadily being deprived of its tribal ligaments, leaving as an unavoidable alternative a state encompassing some notion of territoriality.

The Arab and Berber dynasts who replaced the transitory slave rulers derived power from groups who had common ethnic interests, even if tribal ties had weakened. The division of states among Arabs and Berbers corresponded well with patterns of ethnic settlement which in some cases dated to the eighth century. Already noted (in Chapter 5, section 3) was [209] the tendency of Taifa kings to form ethnic coalitions. The political shape of such alliances is typical of segmentary politics in general. Particularly among Berbers, alliances were at a distance, to avoid direct conflict with allies over resources (pasturage, for example). This gives rise to the famous checkerboard pattern of political alliances so characteristic of North Africa, but which in modified form can be detected in eleventh-century al-Andalus -- as when, for example, Seville was opposed by Badajoz to its west and Granada to its east. The original dispersion of tribes throughout the peninsula, settling in widely scattered foci, isolated from their fellow tribesmen, yet remaining in contact through networks of familial relationships, responds to the same impulse and served to reinforce the pattern when regional entities became fully independent in the eleventh century.⁽³⁶⁾

The apparent fluidity of boundaries of the Taifa kingdoms is in part masked by family-based mergers and disaggregations, as for example the absorption of the kingdom of Denia in 1076 by al-Muqtadir of Zaragoza and of the kingdom of Lérida-Tortosa, ruled by his brother al-Muzaffar, three years later. The changing political boundaries obscure a fairly stable association between one family (the Banû Hûd, in this case) and a particular territory. If the frontier kingdoms (Toledo, Zaragoza) seemed to manifest greater cohesion, it was because of the conditions required by military preparedness.

The kings of the Christian states also did not have kingdoms with stable frontiers until the thirteenth century. Instead (in Maravall's characterization) they ruled over variable space whose boundaries changed in accordance with considerations of kinship.⁽³⁷⁾ Christian society, with its recent roots in mountain states of small area, was unaccustomed to ruling over large spaces, a historical circumstance reinforced somewhat by the steady growth of seignorial domains. The propensity for kingdoms to unite and disunite was as marked on the Christian as on the Muslim side. Sancho el Mayor divided his kingdom among three sons, who inherited Navarre, Castile, and Aragón, respectively. Later, Ferdinand I also willed a tripartite division of his realm. In view of this flux of political boundaries, Pastor suggests that a more realistic view of social structure may be gained from studying economic regions (such as the Santiago road axis of the eleventh and twelfth centuries) rather than focusing on political entities which lacked geographical stability.⁽³⁸⁾ The same is even more applicable to the Taifa kingdoms: if one focuses on economic regions [210] (an Ebro region, a Toledo region, a Levantine region, and so forth), greater continuity is apparent than would appear from the kaleidoscopic mutations of kingships.

The emergence of the Christian kingdoms and the crystallization of cultural or subcultural units associated with particular territories were achieved on a political level by interrelated groups of nobles, who were the first to perceive ethnic solidarity.⁽³⁹⁾ Rivalry between groups of nobles easily found territorial expression, as in the enmity between León and Castile, involved in competition over spheres of influence in conquered territory.⁽⁴⁰⁾ With Sancho el Mayor the center of gravity in the west shifted decisively from León to Castile, a process which gained force when Castile became a kingdom under Ferdinand I, and which culminated in the coronation of Alfonso VII of Castile as Emperor in 1135. Castilian hegemony was in all probability the result of the military power and prestige which its nobles won, as the most aggressive force in the battle against the Muslims -- a response to its forward position on the frontier. In sum, there seems to have been a certain convergence in political form, if not in content, in the two cultures. To regard the title of Taifa kings as fictive, presumably on the grounds of an imprecise or undervalued notion of territoriality, is to put a very different reading on similar geopolitical fluidity in the Christian states. The terms of the political equation were similar, although the proportions of course varied: boundaries (however conceived) were determined by considerations of kinship; they shifted for dynastic reasons; cohesion and continuity were provided by growing ethnic consciousness and by economic regionalization.

4. Feudalism

Feudalism was a term coined by seventeenth-century political theorists and historians to describe a set of relationships thought to have been typical of medieval society. These were relationships of dependence among persons of unequal status, were generally based on land tenure arrangements, and were thought to represent a privatized political structure called into being by the absence or weakness of a central or public authority. This set of relationships seems appropriate for comparative analysis, the received view being that in the peninsular states they were most sharply structured in Catalonia, imperfectly articulated in Castile and León, and scarcely developed in al-Andalus.

[211] In the present discussion, the development of "feudal" relationships, especially in Christian Spain, is seen as one element in a process of social crystallization which created patterned bonds of solidarity among individuals and social groups and therefore is viewed more as a factor of cohesion than of disaggregation.⁽⁴¹⁾ The forging of such bonds was one of a number of processes following upon the structural loss caused by the Islamic conquest, and its primary function in the Christian kingdoms was not to substitute private for public authority (a process that was aborted or at least delayed in Spain) but rather to create solidarities to complement progressively weakening kinship ties.

The discussion of feudalism in medieval Spain has been characterized by an overly facile and uncritical acceptance of an evolutionary model which assumed that all European states passed through pre-feudal, feudal, and post-feudal stages. Liberal historians of the nineteenth century proposed that countries where the implantation of feudal norms had been incomplete (Spain, England) were more advanced at the time and had escaped in some measure the servitude of man to man that was abhorrent to liberal ideals. The problems caused by an evolutionary model are not difficult to appreciate. It posits that there was a palpable classic feudalism which can be identified as having functioned in specific places and times, when in fact the concept was originally no more than a heuristic device. It follows that the model is accepted as a norm and therefore societies not meeting the norm can be typified as deviant. Such deviance can then easily be related to similar notions of a teleological nature, used to buttress Spain's supposedly unwholesome "deviation" from a "normal" historical course.⁽⁴²⁾

In its etymological sense, feudalism indicates a land-tenure structure based on the fief (Latin, feudum), a grant of land in return for military service and an oath of fealty. This is the primary criterion by which it is argued that feudalism was incompletely developed in Castile. But the term has extended meanings -- to characterize an entire society as "feudal" (Marc Bloch), indicating that "feudal" values typified an entire society whose land tenure and personal relationships were so structured or to characterize, as Marxist historians often do, the medieval economy as a "feudal" mode of reduction.⁽⁴³⁾ Other authors detect the form of feudal relations but find the substance wanting. Thus, Burns asserts that thirteenth-century Catalonia was feudal in structure but that "a most imperfectly feudal soul inhabited this body" -- in particular, that in spite of [212] feudal-style personal linkages among persons of unequal rank, in fact there was "a tendency at all levels towards independent [land] holdings."⁽⁴⁴⁾

Sánchez-Albornoz has demonstrated the existence of "pre-feudal" relationships in the Visgothic kingdom, where fideles of the king were rewarded for their service by precarious grants of land or income. Subsequent to the Islamic invasion, the circumstances of constant warfare strengthened the monarchy, whose access to wealth in the form of war booty made it possible to pay salaries for military service and therefore to avoid having to organize the state on the basis of a feudal hierarchy linked to the king by vassalatic-benefice relations. Court officials were also paid, and it was not necessary to make such offices hereditary. There were, then (as summarized by Valdeavellano) seven reasons why full-fledged feudalism never emerged in Spain: (1) Benefices were not linked to vassalage. Land grants were frequently made without requiring an oath of fealty in return. And the reverse was also true: vassalage did not inevitably require the grant of a benefice. (2) Hence, the true fief never became established as the central element of the relationship between the monarchy and the nobility. (3) Immunities were never as common as they were in France. Indeed they became comparatively more common only in the late eleventh century as an element in adaptation of French political styles during that period. (4) Royal prerogatives were rarely ceded to lords and royal justice (as in England) was still felt even in supposedly immune jurisdictions, owing to the influence of Roman law. (5) Public offices did not become hereditary. (6) With the exception of Catalonia, there was no tightly structured feudal hierarchy. (7) There was strong resistance to the subinfeudation of domains.⁽⁴⁵⁾

The feudal hierarchy in Catalonia represented the closest peninsular approximation of the French model, an apparent result of the region's Carolingian background. The upper nobility (ricshomens) consisted of counts at the top of the feudal hierarchy, then viscounts, comitores (aides of the former), and vasvessores (vassals of viscounts and comitores). The vassals of these, holding subinfeudated grants, were the knights (cavallers, the equivalent of the Castilian infanzones), who were chatelains of single castles. Their vassals were sotcastlàs, who had grants sufficient to support a horse and arms.⁽⁴⁶⁾ Even so, as Burns and others have noted, fiefs could be alienated without the lord's consent and the feudal structure was looser than that envisioned in the classical model.

[213] As a system of governance, however, it is doubtful that even Catalonia could be characterized as substantially distinct from the other Iberian kingdoms. The strong impress of French feudal norms received in the late eleventh and twelfth centuries (when virtually a wholly French feudal terminology was introduced) was counterbalanced by the ability of all Christian monarchs to create civil bureaucracies from the funds made available by parias, tribute received from the Islamic kingdoms. Castro's bizarre notion that "One obvious reason for the absence of feudalism was the fact that the royal power at times reached the people by way of officials and institutions that were not Christian" is a reflection of the variegated nature of civil administration.⁽⁴⁷⁾ Of course, the fact that the Islamic muhtasib should appear in Castilian towns as the almotacén and in Levantine towns as the mustasaf has nothing to do with feudalism or the lack of it. More accurate is Sánchez-Albornoz's inference that the model of a highly structured civil and military bureaucracy in al-Andalus emboldened Christian rulers to institute central controls rather than diffusing administration and jurisdiction in a feudal network. That the Castilian tax-collecting official should have an Arabic title, almojarife, is simply proof of the fact that a centralized fisc was an attractive model, given the obvious ability (clear to Christian rulers) of Muslim kings to control the wealth of the state.

Comparison with Islam also reveals the kinship-substitution role of feudal relations. Where elite recruitment and deference entitlement were based on kinship, within a tribal framework, there was no need for the development of feudal-type relations. There, the iqtâ' or "money fief" existed purely for the purpose of remuneration of military or government service and was not hereditary. Those areas of the Islamic world where the iqtâ' was most highly developed, as in Mamluk Egypt and Syria, were slave bureaucracies in which elite recruitment was on a non-tribal basis.⁽⁴⁸⁾ If one looks at relationships of dependence, apart from the organization of the state, however, it seems that there were ample points of convergence. As mentioned in the discussion of clientage in Chapter 4, Visigothic customs of personal dependence may have bolstered clientage relations between Mozarabs and Neo-Muslims and their Arab overlords.⁽⁴⁹⁾

In the western Christian kingdoms, we have noted that the mesnada, the group of warriors in the service of a lord, tended to have the coloration of extended kinship groups. More exactly, it was an institution intermediate between a kinship and a feudal structure. Royal vassals received land [214] grants which they were able to make hereditary, but remuneration for actual services was either by salary or by precarious grants (préstamos), which were lost to the vassal if he failed to perform the service.⁽⁵⁰⁾

The introduction of French feudal norms in the eleventh and twelfth centuries seems another example of a stylistic shift called into being by structural consolidation causing a lack of fit between the social model (in Castile, one laid down in the period of resettlement of the Duero Valley) and the actual structure of society which had become more stratified and more complex. Many have noted the formal nature of feudal relations, that they gave normative meaning to preexisting relations without affecting their structure.⁽⁵¹⁾

1. Thomas Spence Smith, "Aestheticism and Social Structure: Style and, Social Network in the Dandy Life," American Sociological Revietv, 39 (1974), 735-736. Note Reyna Pastor's use of similar terminology -- "destructuration" and "structural recomposition" (Del islam al cristianismo, p. 10). However, the scope of these terms is limited to the description of the restructuring of economic organization and land-tenure arrangements where the countryside was literally destructured through the loss of Muslim inhabitants and reconstituted with a wholly new Christian population.

2. See Smith, "Aestheticism and Social Structure," p.742 n. 10.

3. Américo Castro, The Spaniards, pp. 213-214 (emphasis mine). J. A. García de Cortázar, Epoca medieval, pp. 50 (emphasis mine), 111; cf. p. 124 for his characterization of Almohad rule as "superstructure." See Guichard's critique of Sánchez-Albornoz, Al-Andalus, p. 496.

4. Guichard, Al-Andalus, p. 272.

5. Ibid., pp. 472, 519.

6. Burns, Islam under the Crusaders, p. 413.

7. Pastor's explanation of the "political insufficiency" of Islam (Del islam al cristianismo, p. 56) is a gross misconstrual of the nature of Islamic society in which, she states, all believers are like a tribe. If anything, the umma was intended to be deliberately non-tribal.

8. Guichard, Al-Andalus, pp. 380, 549, 551; Ibn Khaldûn, The Muqaddimah, I, 47, 61.

9. On the theme of social stability in the face of political instability, see Torres Balbás, Ciudades hispanomusulmanas, p. 322; Guichard, Al-Andalus, p. 99; Turner, Weber and Islam, pp. 81, 172, 174; Goitein, Mediterranean Society, II, 348.

10. Urvoy, "Etude sociologique," pp. 233-234; Turner, Weber and Islam, p. 55. See also, Burns, Islam under the Crusaders, p. 403.

11. Turner, Weber and Islam, pp. 87, 105, 111, 172; Goitein, Mediterranean Society, II, 366; Urvoy, "Etude sociologique," p. 232.

12. Mones, "Hommes de religion," p. 78. The disinclination to serve may not have been purely religiously based. Government service was a risky enterprise, and the middle class preferred the stability of business or scholarship; see Goitein, Mediterranean Society, II, 375.

13. Al-Khushanî, Historia de los jueces de Córdoba (Aguilar ed.), pp. 561-562 (a judge in Sidonia called to Córdoba); 707 (a judge in Elvira, then Córdoba). Cf. Pp. 205-206, where.one of the descendants of 'Umar ibn Xarahil, who spawned a family of amins and judges, was judge in Jaén and Ecija. Ibn al-Zubair, Silat al-Sila, passim.

14. Ibn Khaldûn, The Muqaddimah, I, 53. On the nature of the council of elders, see Burns, Islam under the Crusaders, pp. 386-400.

15. Urvoy, "Etude sociologique," pp. 283-284.

16. K. Garrad, "The Original Memorial of Don Francisco Núñez Muley," Atlante, 2 (1954), 219.

17. Guichard, Al-Andalus, pp. 59, 21; Ibn Khaldûn, The Muqaddimah, I, 61, 316.

18. Valdeavellano, "Las instituciones feudales en España," p. 292; Torres Fontes, Repartimiento de la huerta y campo de Murcia, p. 89.

19. Smith, "Aestheticism and Social Structure," p. 736.

20. On Ziryâb: Makkî, Aportaciones orientales, pp. 181-182; Dozy, Spanish Islam, pp. 261-265; Huici Miranda, Cocina hispanomagribi, p. 179. On dandies as a social group, see Mhammed Ferid Ghazi, "Un groupe social: Les raffinés (zurafâ'),"Studia Islamica, 11 (1959), 39-71, especially pp. 48-50, 65-67.

21. Rattaba rusûm al-mamlaka (Makkî, Aportaciones orientales, p. 179). Makkî also notes (p. 190) that there was greater interest in establishing legal norms in ninth-century al-Andalus than in the study of tradition (hadith), per se.

22. Livermore, Origins of Spain and Portugal, pp. 302-303, 355; Bosch Vilá, A1barracín musulmana, p. 41; 'Abd al-Karîm, Al-Andalus en el "Mu'jam al-buldân" de Yâqût, pp. 22, 27.

23. Makkî, Aportaciones orientales, p. 94.

24. Hussain Mones, "Consideraciones sobre la época de los reyes de Taifa," Al-Andalus, 31 (1966), 323-324.

25. Ibid., p. 307.

26. James Dickie, "Ibn Suhayd: A Biographical and Critical Study," Al-Andalus, 29 (1964), 284.

27. Al-Saqundî, Elogio del islam español, p. 47.

28. W. Montgomery Watt, A History of Islamic Spain (Edinburgh, Edinburgh University Press, 1965), p. 87.

29. Cagigas, Andalucia musulmana, pp. 20-21; Lévi-Provençal, L'Espagne musulmane au Xème siècle, pp. 23-24.

30. E. García Gómez, "Bagdad y los Reinos de Taifas," Revista de Occidente, 43 (1934), 9; Pérès, Poesie andalouse, p. 49; Urvoy, "Etude sociologique," pp. 287-288.

31. Maravall, El concepto de España en la edad media, pp. 300, 314; Claudio Sánchez-Albornoz, "La restauración del orden gótico en el palacio y en la iglesia," in Orígenes de la nación española, II, 623-639.

32. 32.Sánchez-Albornoz, "Tradición y derecho visigodos en León y Castilla," Investigaciones y documentos sobre las instituciones hispanas, p. l29.

33. Sánchez-Albornoz, "El precio de la vida en el reino astur-leonés hace mil años," p. 384. For further evaluations of Neo-Gothicism, see Castro, The Spaniards, pp. 180, 183; García de Cortázar, Epoca medieval, pp. 133 (as superstructure), 291; and Barbero and Vigil, Orígenes sociales de la reconquista, p. 97.

34. On the conquerors of Zaragoza, see Verner, Los musulmanes españoles, p. 33; Ribera's theory is expounded in Orígenes del justicia de Aragón; on the diffusion of hisba, Chalmeta, El señor del zoco, and Glick, "Muhtasib and Mustasaf"; on medicine, Luis García Ballester, Historia social de la medicina en España de los siglos XIII al XVI. Vol. I. La minoria musulmana y morisca (Madrid: Akal, 1976), pp. 47, 52; on the dhimma model, Burns, Islam under the Crusaders, p. 163.

35. See García de Cortázar, Epoca medieval, p. 323.

36. Guichard, Al-Andalus, pp. 47, 338; on Berber alliances, see Seddon, "Local Politics and State Intervention," p. 125.

37. Maravall, El concepto de España en la edad media, pp. 358-359, 388.

38. Pastor, Conflictos sociales, p. 17.

39. Ibid., p. 53.

40. See Moreta, San Pedro de Cardeña, p. 52.

41. Cf. Maravall's view of feudalism as a unifying system rather than as a factor of decomposition (El concepto de España en la edad media, p. 387).

42. See Elizabeth A. R. Brown, "The Tyranny of a Construct: Feudalism and Historians," American Historical Review, 79 (1974), 1063-1088, especially p. 1076. The notion that the Islamic invasion caused a detour (desviación) in the historical route of Spanish history is that of Sánchez-Albornoz, who also argues that the circumstances of the Reconquest inhibited the normal development of European-style feudal relations. See Castro's oblique criticism of this aspect of Sánchez-Albornoz's thought in The Spaniards, p. 172.

43. See Pastor, Conflictos sociales, p. 17 n. i6, where she states that the mode of production in medieval Castile was feudal although the feudal system was not normative. Also, cf. p. 1861 where a "seignorial economy" is defined as the exploitation of servile or semi-servile labor; but "seignorial" is not "feudal" in the strict sense.

44. Burns, Islam under the Crusaders, p. 275.

45. Sánchez-Albornoz, España: Un enigma histórico, II, 56-74, and "Repoblación del reino asturleonés," pp. 437, 429; García de Cortázar, San Millán de la Cogolla, p.318 (on immunities); Valdeavellano, "Las instituciones feudales en España," p.235. On resistance to subinfeudation, see Lewis, The Development of Southern French and Catalan Society, p. 76.

46. Valdeavellano, "Las instituciones feudales en España," p. 292.

47. Castro, The Spaniards, p. 427.

48. Claude Cahen, "Iktâ," EI², III, 1088-1091; Lévi-Provençal, España musulmana:Instituciones y vida social, p. 115; Pedro Chalmeta, "Le problème de la féodalité hors de l'Europe chrétienne: Le cas de l'Espagne musulmane," II Coloquio Hispano-Tunecino (Madrid, 1973), pp. 91-115.

49. See Guichard, Al-Andalus, p. 560 n. 4, and bibliography cited.

50. Valdeavellano, "Las instituciones feudales en España," p. 243.

51. Brown, "The Tyranny of a Construct," commenting on Georges Duby.