1. The Linguistic Model and Cultural Continuity

[277] To take linguistic change as a model for cultural change in general is wholly appropriate. The contact of two different languages provides a microcosm of the contact of cultures. One notes a range of effects, from profound to superficial, depending upon factors intervening in specific phenomena (syntax, phonetics, vocabulary), as well as mechanisms accounting for the contours of selectivity in each area. Linguistics and philology provide a methodology for tracing the history of such changes and of assessing the impact of culture contact in an accurate way, sometimes even quantitatively.

Virtually all the models generated for explaining cultural change in medieval Spain have come from the field of comparative philology. Particularly influential has been the concept of the linguistic substrate, used as a heuristic device to account for the survival in the language of archaic elements of diverse provenance. The concept has been extended to provide a general mechanism for analyzing contents of Spanish culture as the product of the historical succession of cultures. The concept of the substrate makes clear that this is a cumulative, selective process-which is undeniable-but the geological metaphor has had unfortunate consequences because the implicit geological processes have no cultural analogues.

The notion of the substrate is something akin to the onion-skin view of geology: that if one peels layers off the surface, one finds a series of distinctive strata beneath. If a layer is peeled off partially, selectively, as if by erosion, portions of underlying layers will be visible from the surface. The problem is that there is no process of cultural change analogous to erosion. Thus the concept has not only fitted in well with arguments (such as Sánchez-Albornoz's "temperamental inheritance") of a genetic nature, but can also be used by those who reject the racist bias of the genetic argument but whose theory of culture is more or less the same. The fallacy of the geological model is that there is no cultural bedrock, no steady-state culture. Culture is always changing and transmuting itself [278] by dropping old elements, elaborating new ones, or by conjoining traits in different ways. The geological metaphor implying that a people can dig down, as it were, through the surface of its own culture and find the bedrock, or that there will be outcroppings of original, bedrock culture sticking up among superficial upper layers, is an inexact model of cultural change.

The model was originally designed, and most accurately used, for the purpose of explaining relationships between linguistic phenomena bearing the imprint of cultural succession. Perhaps the best-known example is the inhibiting effect exerted by Cantabro-Basque linguistic structure upon the Latin initial f-. Its disappearance in Castilian is attributed to the substratic action of the Basque language. The substrate is thus seen as inhibiting one element in the process of Romanization and as encouraging the future evolution of Castilian Romance in a particular direction. When molded into a theory of general cultural action, the result is a very conservative view of cultural change. For example, J. A. Maravall (building upon Amado Alonso's definition of substratic action) notes that "when the language of an invading people dominates that of the conquered, it happens that in the evolution of the latter there will appear laws of change (phonetic, lexicographic, syntactic) which respond to the characteristics of the language which has disappeared. Thus when Mozarab population decays, eliminated by the Hispano-Muslims or absorbed by the Christians of the north, in the history of the latter tendencies shoot up which are rooted in that primitive stratum (capa), whose civilization, whose historical concepts, had been stifled." He goes on to explain that the stratum in question was basically Hispano-Roman upon which had fallen a relatively weak German superstrate. Indeed, the survival of Latin proves the great influence of this substrate, which would exercise "a later and decisive action" upon the Romanic-European civilization of the Christian principalities of northern Spain. Such action explains not only the distribution of Germanic and Roman toponyms, for example, but also the preservation of social rules later codified in the Fueros. This, Maravall associates with Menéndez Pidal's concept of cultural latency.⁽¹⁾

The notion that cultural bedrock can intrude at any moment through a superficial superstrate is similar to Sánchez-Albornoz's likening of culture to an incompletely expressed phenotype, an analogy which fails because, just as there is no bedrock specific to a given culture, there is no cultural genotype. Yet this construction has had enormous influence upon [279] historians and persists as a kind of reductionist argument, to explain the unexplainable by allusion to a more or less remote substrate. To this is added the valuative problem of determining just which substrate is exercising control. To Maravall, it is the Hispano-Roman. For Sánchez-Albornoz, it is essentially Visigothic, at least insofar as law and institutions are concerned, although -- as noted -- he holds psychological traits as constant, virtually from prehistoric times.⁽²⁾ For Pierre Bonnassie, culture of the mountain dwellers of the eastern Pyrenees prior to A.D. 1000 was explicable in terms of pre-Roman and even pre-Celtic foundations, which crop up "in the archaism of a language still influenced by the Basquoid substrate, the maintenance of pre-Christian beliefs," and in elements of Roman and Visigothic law.⁽³⁾ For Lucie Bolens, dealing with Andalusi culture, the significant substrate is a Punic one, an argument which has become almost traditional in Spanish Arabism, and which recurs in Arnold Toynbee's notion of the Arab-Berber retrieval of Spain for Syriac society, a retrieval because in this view Iberia had once been Punic -- an argument which is so hyperbolic and indemonstrable that it needs no refutation, the more so since its partisans offer no proof. Thus Bolens argues that Berber influence attenuated the orientalization of al-Andalus and reinforced the Celtiberian base of its culture.⁽⁴⁾ This is tantamount to saying that if the United States were conquered by China the oriental traits of American Indians (whose ancestors crossed the Bering Strait in some distant millennium, but no more distant than Celtiberian times perhaps) would thereby be supported and Asiatic culture strengthened.

This kind of argument is easy to make because, first, the structural impossibility of a remote substrate manifesting dominance over a series of diachronic cultural boundaries means that no proof of these contentions will ever be forthcoming, and hence no refutations can be offered. Iberian allusions are particularly safe inasmuch as the time is so remote and the state of knowledge so fragmentary that it can explain either everything or nothing. Second, as Guichard wisely notes, many elements identified by historians as "Hispanic" customs dating to one substrate or another are in fact Mediterranean traits widely diffused throughout the entire basin.⁽⁵⁾

Once a theory of substratic action is accepted, other forms of cultural explanation are virtually excluded. A few examples of how the substrate defeats other kinds of explanation will suffice. Sánchez-Albornoz and his school claim that in view of the subsistence of Hispano-Roman culture [280] among Neo-Muslims the latter could not have transmitted elements of Eastern culture to the Christian north. We have already noted the insubstantiality of this line of reasoning. It is also used to explain how the Mozarabs were unable to transmit to northern Christians styles of life or thought foreign to themselves. Proof of this, according to Hilda Grassotti, is the "non-conformity" of Mozarabs, an attitude proved by their migration. Here, non-conformity is a social characteristic from which a cultural deduction is made: the more Christian, the less Arabized. The supposed lack of acculturation is the obverse, and therefore a proof of, the persistence of a Hispano-Gothic substrate.⁽⁶⁾ But, of course, partially or even minimally acculturated groups can well serve as agents of diffusion. Were it not so, the histories of art, science, or technology in Christian Spain would be unintelligible.

The substrate, I would suggest, had best be limited to explaining individual traits, without attempting to cross too many diachronic boundaries. It is a valuable tool of philological explanation and can fruitfully be extended to other areas of culture, particularly those where the elements in question are palpable, such as technology. Martínez Ruíz's study of the Berber plow, associating some elements with a Roman substrate, others with an Arab adstrate, is an excellent example.⁽⁷⁾ But as a model for explaining modal personality or even the broad characteristics of Spanish culture at any given period, it clearly is insufficient and inexact. Américo Castro was correct when he said that the provenance of individual traits is basically irrelevant to their integration within a living culture.⁽⁸⁾ Indeed, trait analysis yields an atomized view of culture. That is why Castro invented the morada vital ("dwelling place of life") to supply a context for understanding the linkages between traits. In reality, these linkages are provided by the dominant value systems (e.g., Islam, Christianity, "feudalism," etc.) which provide contexts for integrating elements of disparate origin.

The notion of the substrate also implies that cultures defined in so lapidary a form ceased to exist at a precise moment in time, the moment when the next stratum was laid down (by invasion). Obviously, cultural change is a more complex process. Cultural continuity occurs, but within bounds that must be defined. Linguistic changes are significant as general benchmarks and form the most obvious boundaries between a succession of cultures in a given historical sequence: e.g., Celtiberian, Roman, Visigothic, Spanish. But the periodization and parameters of [281] continuity differ for each family of traits. Thus technological continuity was broken, not by the process of Romanization, but by the economic and social upheaval of late Roman times.

2. Cultural Crystallization

Another philological concept which has had extraordinary resonance in recent historiography (primarily through Américo Castro's work) is the notion of competition between variant norms. This is the basis of Castro's explanatory system, the germ for which is found in the thought of Ramón Menéndez Pidal. Menéndez Pidal noted that "the coexistence (convivencia) of many norms which struggle among themselves with very comparable forces" is characteristic of pre-literary epochs, those periods in which cultures are in a stage of crystallization.⁽⁹⁾ That Castro makes convivencia the primary mechanism of cultural interchange is, I think, prefigured in Menéndez Pidal's very similar, but more restricted, use of the same word.

Again, the notion of the coexistence and competition between norms has much higher explanatory value when restricted to linguistic phenomena than when extended to encompass the whole of culture. In Castro's usage, convivencia becomes a catch-all mechanism used to explain all phenomena of cultural change contingent upon the contact of cultures, an inclusivity which obscures what are in reality a number of different mechanisms.

In the historiography of Christian Spain, emphasis has been placed upon related processes of ethnic, linguistic, and political differentiation between the various groups of northern Christians. These processes were first described by philologists. Menéndez Pidal pointed out that Castile "was defining itself from the tenth century on."⁽¹⁰⁾ The central feature of this process, to the linguist, seemed to be the apparent decisiveness of Castilians in resolving linguistic ambiguities characteristic of the speech of their neighbors. Psychological inferences were soon read into this Castilian "decisiveness," but the exaggerations of philologists ought not to detract from what this process represented: the emergence of an ethnic community, with group awareness, which soon found political expression (in the emergence of the county and kingdom; in border skirmishes with Leonese and Navarrese, etc.). Analogous processes characterized all of the Christian kingdoms: ethnic reference crystallized in regional foci, [282] although each shared in common with the others the superethnic adscription of "Christian" in contradistinction to Muslims.

The crystallization of a distinctively Andalusi Islamic culture was ultimately dependent upon the process of conversion. As a progressively higher proportion of the population became Muslim, institution's and activities of Islamic orientation came to dominate the culture. Since such activities were predicated upon a certain density of Muslims, these are also benchmarks in the process of crystallization and serve as an indirect check on the progress of the conversion movement. In this light one can point to such phenomena as the establishment of an Andalusi school of hadîth and fiqh, the gradual slowing down of the current of scholarly travel eastward as religious networks at home became thick enough to supply high-quality teachers; the severing of the Jewish community's ties to Babylonian (Iraqi) Jewry in the mid-tenth century; the emergence of Andalusi scientific schools in the eleventh; the compilations (dîwâns) of works of Andalusi poets in the eleventh and twelfth centuries; the demand in the East for Andalusi scholars to teach there -- all of these are indicative of growing cultural cohesion and of the elaboration of self-perpetuating cultural norms.⁽¹¹⁾ Similar processes characterized the other emergent nations of the Islamic world. According to Bulliet's data, al-Andalus underwent a conversion process identical to that in other areas, offset by about one hundred years owing to the lateness of the conquest (see Figure 7).

The history of the Arabic language in al-Andalus also bears out the general trend described above. Here one must distinguish between two distinct, although related, phenomena: the progressive Arabization of the population, a process not completed until the thirteenth century, and the emergence of a standard, colloquial Spanish Arabic as a distinctive dialect within the Arabic-speaking world. This last process was completed by the tenth century. Although Andalusi Arabic had regional nuances, it had a common core which betrayed the historical circumstances in which it emerged. The contact between Arabic and Romance phonemic systems produced a number of distinctive elements. Because of Romance influence, Spanish Arabic was one of a number of Arabic regional dialects where the p (a sound generally transcribed in Arabic with the symbol either for f or b) attained full phonemic status. Phonemic stress was substituted for vowel length; the particle li was used to mark accusatives, parallel to the Romance a; and the diglossia between [282] classical and colloquial Arabic together with Arabic/Romance bilingualism combined to produce sentence structure and word order which were much freer than those of classical Arabic. The selectivity of this process is apparent. Syntactical interference from Romance was pronounced, but morphological interference was minimal, due to the extreme structural dissimilarity (anisomorphism) between Romance and Semitic language groups.⁽¹²⁾

The process of cultural differentiation that was taking place was noted at the time. According to Federico Corriente, speakers of Spanish Arabic "were aware of the personality of their dialect and not a bit ashamed of it" and preferred it to classical Arabic for certain literary genres.⁽¹³⁾ To foreigners, Andalusi Arabic seemed impure. A Bedouln in the service of Ibrâhim ibn al-Hajjâj of Seville in the ninth century complained that his speech had been corrupted by Andalusi vulgarisms. There are further complaints, from the eleventh century, about Andalusis who spoke Arabic imperfectly, but these ought to be viewed not so much as a comment on the quality of Andalusi culture as the realization that specific forms [284] of linguistic expression had evolved there. There was no correlation between good Arabic and Arabic ethnicity, incidentally. As Makkî notes, Neo-Muslims were interested in improving their Arabic. Such a man as 'Abd Allâh ibn 'Umar (d. 889) of Seville spoke Arabic beautifully, while detesting Arabs.⁽¹⁴⁾

As in the Romance languages, there was a process of the sorting out of competing norms. In this light, al-Tîfâshî, a Magribi writer of the twelfth century, noted that "of old the singing style of the people of al-Andalus was either according to the style of the Christians, or else according to that of Arabic camel-drivers, without their having norms to guide them until the establishment of the Umayyad dynasty." Standardization occurred precisely during the reign of 'Abd al-Rahmân II when Ziryâb and other singers imposed the Iraqi style.⁽¹⁵⁾ A further sign of crystallization was the emergence of distinct regional foci, areas of cultural subdifferentiation within a broader Andalusi Islamic culture. Thus Urvoy detects regional nuances which, if they appear hazy from the perspective of the present, give evidence of cultural differentiation at the regional level: western al-Andalus was less intellectually and religiously vital than the peninsular east; the Middle March, an advanced strategic zone, was the focus of an accentuated juridical orientation; the Upper March (Ebro) became progressively more Arabized as it passed from Neo-Muslim to Berber to Arab rule, and in virtue, too, of the strong links the region maintained with the Islamic East. Córdoba was the focus of a region of religious and cultural traditionalism.⁽¹⁶⁾

3. Modalities of Cultural Change

The passage of cultural elements along a spatial trajectory may come bout through a variety of means and agents. In some few cases, the mechanism of cultural change may be described as "formal" or institutionalized -- cultural change directed by some institution.⁽¹⁷⁾ Such direction may be applied positively or negatively. An example of the former would be the introduction of Persian-style 'Abbâsid court customs and institutions by the Umayyads under 'Abd al-Rahmân II; of the latter, pressure by government on a religious or ethnic minority to abandon any element of its culture (a phenomenon that became salient in Christian Spain in the later middle ages), incentives for conversion, and so forth. But the preponderance of cultural changes come about through non-formal [285] mechanisms: the migration of peoples, the conductivity of merchants, travellers, books, or whatever channels of communication of information that may have existed. There was always a selective factor: not all innovations offered by the "donor" culture were accepted by the "recipient" culture. The factors influencing selectivity were varied: economic demand, the market for stylistic innovation, a desire for new knowledge all contributed to climates favorable to innovation; hostility, warfare, and religious difference contributed to a climate that may have been inimical to borrowing, but only selectively so. Sánchez-Albornoz's "antibiosis" is in fact a common context of cultural borrowing: competition with an enemy can be a powerful stimulus to cultural innovation. If the barrier is strong enough, such imitation may not be consciously acknowledged; rather, a process of reinvention may take place within the recipient culture. This is what A. L. Kroeber called "stimulus diffusion," and a good example is the borrowing by Christians of the Islamic ribât, or frontier fighting "monastery," and its reinvention as the crusading military order.⁽¹⁸⁾

Next to migration, travel was perhaps the best-defined agency of cultural diffusion. The introduction of eastern customs was the result of a steady movement of travellers, typically from al-Andalus eastward in the first two centuries of Islamic rule, from east to west in later centuries and, throughout, a dense pattern of interaction between al-Andalus and the Magrib, particularly within the sphere of influence of Fez.

Travel to the East was very common among Andalusi fuqahâ', especially during the first two centuries of Islamic rule. It was typical to read in a biographical account that a man had travelled in the lands of the East (tâfa bilâd al-mashriq) and that he had done so "in search of knowledge" (fîtalab al-'ilm). To a certain extent, the pilgrimage destination of Mecca determined the places visited (e.g., Qayrawân, Alexandria, Cairo, all places with scholarly communities), but the search for specialized knowledge deflected scholars to, for example, Basra and other Iraqi centers to study grammar. (A similar pattern held for Jewish pilgrims, whose destination was Jerusalem, however.) The number of people making such trips was relatively high: of 246 scholars from the Ebro Valley studied by Juan Vernet, one-quarter had studied in the East.⁽¹⁹⁾

Travel to the East, to be sure, was a necessity in the years before Andalusi educational institutions had reached a density sufficient to support scholarship of high caliber over successive generations. Given this [286] situation, to have an eastern teacher was a way to increase prestige at home. The career of Baqî ibn Makhlad, a Neo-Muslim traditionist of the ninth century (817-889) is revealing in this regard. He was said to have spent thirty-four years studying with no less than 284 eastern masters (a fact that both reflected his prestige and contributed to it). But it was also said that if a country had such a man as Baqî, then there was no reason for its scholars to travel abroad in search of tradition. A scholar of the next generation, Qâsim ibn Asbagh al-Bayyanî (858-951), was noted as a pole of attraction for indigenous scholars: "towards him were directed the travels of students in al-Andalus" (wa kânat al-rihla fi'l-Andalus ilayh). In later times, logically, the pace of travels to the East slowed.⁽²⁰⁾ It never ceased; the pilgrimage ensured its continuance. By the twelfth century, Fez had replaced Qayrawân as a pole of attraction, and the cultural unity of Morocco and al-Andalus was apparent. The numerically most important contacts between Andalusi and Magribi towns during this period are represented in Map 1, compiled on the basis of ibn al-Zubair's mentions of scholars resident in one town reported to have studied in or travelled to another.

We have already noted, particularly with regard to the transmission of science, the pivotal role of Jews in the process of cultural exchange. Jews were frequently bicultural or tricultural and moved between Arabic and Romance cultures with facility. Shifting cultural referents, either by conversion or by migration, Jews were likely both to be receptive to cultural stimuli of different origins and to be able to assess the market for cultural innovation in the host society. Interestingly, Andalusi Jews seemed not to have travelled much to the East for purposes of study -- the result, no doubt, of scholarly traditions already highly developed before the growth of the Jewish community in al-Andalus. Books and scholarly (rabbinical) communications were regularly received from the East, thanks to the coming and going of merchants.⁽²¹⁾

Even highly Arabized Andalusi Jews were able to function in the Romance language. Jewish poets like Yehuda ha-Levi wrote strophic poems in Hebrew with final couplets in Romance, following an Arabic literary convention. Beyond this, the cultural mobility of Jews made them particularly well suited to play the role of cultural broker. An apt example is ibn Buklârish, a Jew from Islamic Zaragoza, who composed tables of Arabic and Romance synonyms of materia medica.⁽²²⁾

The ability to judge the market for cultural products was another [288] Jewish hallmark. Jewish instrument-makers residing in Christian Spain, for example, made no scientific instruments bearing Hebrew letters. Rather, they produced Arab-style instruments (like astrolabes) for a Christian market which required Latin instrumentation.⁽²³⁾ Converted Jews were in an even more sensitive position when it came to assessing the cultural needs of their new reference group. Pedro Alfonso, who converted to Christianity in 1106 and was familiar with the Arabic scientific tradition, became a kind of propagandist for Arab science, particularly astronomy; he wished "to raise once more to life the knowledge of that science which is in such a deplorable state among those educated in the Latin manner."⁽²⁴⁾ A converted Jew (who had been a physician trained according to Arabic norms), he was in a position to compare the competing "manners" of education, and to find the Latin one wanting. Jews, moreover, remained receptive to cross-cultural information in times when the host culture was closing down. Luis García Ballester points out that the winding down of the translation movement severely retarded the transmission of new ideas from Muslims to Christians. Thus the important medical treatise of ibn Zuhr, the Taysîr, was not known to Spanish Christans until 1281, more than a century after it was written. But it was known to Jews in Barcelona in 1165, four years after ibn Zuhr's death.⁽²⁵⁾

Mozarabs played an analogous role, although less is known of the dimensions of their cultural conductivity. Levi della Vida presumed Andalusi Christians to have been in contact with Nestorian and other eastern communities whose influence was felt in the adoptionist and other heterodox movements. We know there were contacts between Córdoba and Jerusalem, a center of Christian religious opposition to Islam, notably in the person of the monk George, a Jerusalemite executed in Córdoba during the martyrdoms of 852. In contrast to the Jews, who moved in tandem with Arabic culture, Mozarabs appear to have acted almost wholly within a Christian context and to have made few contacts with the secular knowledge of the Islamic world. An exception is the dedication of a treatise on urine by a ninth-century Coptic physician to Khâlid ibn Rûmân, a Mozarab physician from Córdoba.⁽²⁶⁾ Coptic numerals appear in Mozarab manuscripts, and the Calendar of Córdoba also represents some Mozarab participation, via Arabic, in Eastern culture.

Mozarab contacts with the Christian north, not only literary and artistic but also technological, appear to have been important. Eulogius of Córdoba brought books back to al-Andalus from his tour of Pyrenean [289] monasteries in the early ninth century, just as manuscripts originating in Christian scriptoria in Toledo and Córdoba reached the kingdom of Oviedo. These included not only texts of the Visigothic period, but also material of more recent origin, such as the text of the Council of Córdoba, held in 839. Manuscripts continued to arrive from the south throughout the ninth and tenth centuries, following the routes of Mozarable migration to Ripoll in Catalonia ind to the Riojan monasteries of Albelda, Cardeñia, San Milláñ de la Cogolla, and Silos, heavily Mozarabic centers which became nodes of information exchange between Mozarabic and European culture.⁽²⁷⁾

In a manner functionally similar to the way in which Islamic currents were felt in al-Andalus via Morocco, western European influence impinged upon Christian Spain through France.⁽²⁸⁾ For a time after the Islamic invasion, the flow of ideas from Christian Spain to northern Europe continued to follow the same routes as during Visigothic times, though at a much diminished rate. Materials from the Toletan councils of the late seventh century and monastic rules such as those of San Fructuoso of Braga continued to reach European monastic centers in manuscripts of peninsular origin through the ninth century. In that century also, Gerona was a focus for the diffusion of manuscripts of the Visigothic Laws, in response to demand from "Hispani" settled in the Frankish kingdom, who were allowed to rule themselves according to the old Gothic laws. In the tenth century, the direction of flow changed, the circulation of European material in Spanish monasteries accelerating toward the end of the century. When the Roman rite was introduced, the Christian zone, particularly in the east, was inundated by relevant ecclesiastical literature, according to Manuel Díaz. The fact that the monastery of Ripoll was central in this process seems to indicate that the general cultural movement paralleled that of scientific activity on the part of Christians.

The fact that the see of Narbonne spanned the Pyrenees ensured that the "Spanish March" would be a center of exchange with Frankish Europe. The monasteries of Ripoll, Cuixá, and Canigou were the leaders in establishing contacts with Europe north of the Loire, the former having established close relations with the monastery of Fleury. In the peninsular west, a similar situation prevailed, thanks not only to the pilgrimage route but also to the fact that the political entity of Navarre also spanned the Pyrenees. In the eleventh century the pace of exchange [290] quickened and it was during this time that the Christian kingdoms, particularly those of the west, made the decisive shift from an Andalusi to a European cultural polarity. The influence of French monks, first from Cluny, later from Citeaux, was strongly felt. Cluniacs, as summarized by Defourneaux, directed the changeover from Mozarabic to Romanic liturgy, supplied most of the bishops, particularly in reconquered areas (for example, Bernard de Sédirac, first bishop of reconquered Toledo), organized the pilgrimage route, advised kings, and participated in exchanges of Christian and Islamic art and high culture. just as Ripoll and Cuixá were centers for the reception of Frankish cultural currents in the east, so were San Juan de la Peña in Aragón and Sahagún in Castile-León, the latter with a Cluniac abbot named by Alfonso VI.

Castro noted, correctly in my opinion, that there is no inherent contradiction in the simultaneous action of Islamic and non-Islamic models, even upon the same phenomenon. Thus the military orders could both assume a French economic and political structure but be implanted in an Islamic-style value system which made a sacred duty of fighting for the faith. Sánchez-Albornoz, in fact, makes the same point, regarding kingship in León; the notion of a territorial kingship (regalia territorial), though inherited from the Visigothic juridical tradition was nevertheless renforced by the example of Islamic norms prevalent in al-Andalus.⁽²⁹⁾

4. Balance-Sheet of Cultural Exchange

The preceding observations provide a basis for a rational evaluation of cultural exchange. In this concluding section, I will confine myself largely to a discussion of Américo Castro's ideas, with some additional observations as a gloss upon them. Castro's critique of traditional Spanish historiography is directed at the notion that Spanish culture has fixed characteristics which have remained constant from the time of the Iberians to the present. This fallacy he labels "pan-Hispanism" since it attributes "Spanish" characteristics to all groups which have inhabited the Iberian peninsula. Castro's reply to this generalization is that the culture we call Spanish is not an eternal entity, but one which came into being in the centuries following the Islamic invasion of 711. That culture was one whose idiosyncrasies can be explained by reference to the interaction of Muslims, Christians, and Jews. As I will note below, Castro's account of the process of cultural crystallization and the emergence of Spanish [291] ethnicity accords well with recent sociological explications of these processes, even though he failed to understand that cultural processes are controlled to a very great extent by social structure.

Unacceptable as historical method, however, are his continual reiterations of the singularity of the Spanish situation: "the essential uniqueness of the phenomenon," "uniquely Spanish circumstances." Such expressions follow from his assertion that all peoples are unique and that, therefore, historical parallels are always inaccurate. On this basis he concludes that the Spanish situation of cultural pluralism was "without analogy . . . in Orient or Occident."⁽³⁰⁾ Such statements have the result either of discouraging comparative study or of excusing the lack of it, and this posture is defensible only if one deals with content and not structure, or with traits and not relationships or patterns among traits. Not only is there a large literature regarding culture contact and acculturation, but there are instances of Muslim-Christian contact, notably that of the Turks and Greeks in Asia, which bring to light phenomena of direct comparability. But Castro specifically discourages looking into such situations, since such data as might be unearthed would have only "folkloric" value: "For my purpose it makes no difference that such practices [Islam-derived customs present in Spanish culture] are also found in other countries likewise touched by Oriental civilization . . . since I am not interested in folklore but merely in making evident the effects of nine hundred years of Christian-Islamic interaction on the modes of life of the Hispanic peoples."⁽³¹⁾ The stress on uniqueness, linked to the derogation of comparative method in analyzing the processes of cultural change in medieval Spain, has been a dead-weight on medieval historiography. As the sociologist Pierre van den Berghe notes, uniqueness and universality are really two sides of the same coin: "Indeed, at the limit, even the universe is unique, but uniqueness is the product of a large number of repeated, normative, and predictable elements."⁽³²⁾ In his ardor to combat the myth of an eternal Spain, Castro, by eschewing the general literature of cultural action, denied his arguments an important weapon of defense.

Both Castro's theory of history and his methodology hinge upon factors which he understands to have a "historifiable" dimension. Throughout his work he defines and redefines "historifiable," a necessary procedure since he does not give the term objective parameters. In one of the clearest passages, he describes "historifiable existence" as that which is a "creator of values."⁽³³⁾ That is, "historifiable" refers to all those processes [292] whereby value systems are created. This is the concept at its most intelligible level, and the analysis of the process of value creation in medieval Castile is the most significant aspect of Castro's work. But as nuances are added to this straightforward meaning, the concept loses explanatory power. "That which is historifiable will depend . . . on how social groups ceaselessly conceive, with their gaze turned toward the future, the ways of life in which they exist, that is to say, on how they use their biological nature, which in itself is of no interest to history."⁽³⁴⁾ Earlier in the same discussion, he states that the modes of social organization which permitted the formation of Spanish culture "were both desired and projected ... a will toward self-expression in the form of a valuable culture."⁽³⁵⁾ In sum, Castro's concept of historifiable includes four qualities: (1) an element of consciousness, of a group's awareness of constituting a particular kind of society; (2) a teleological element, the notion a group has of becoming something in the future; (3) a valuative element -- collective awareness of having accomplished something great or of enduring value; and (4) the ability to mold these first three elements into some form of artistic expression, which in turn becomes the moving force of society.

It is the element of consciousness that causes the most problems for social scientists. Castro constantly stresses that those values which matter most are those which are consciously realized: "collective awareness capable of being elevated to the national level," "awareness of forming part of a [specific] human community," "to appear now as a Spaniard ... is the result of an aim," "the large majority of [the Spanish] people willed to have done with the Jews and the Moors."⁽³⁶⁾ Such citations could be multiplied, but these suffice to illustrate two points: first, Castro conceived the basic "historifiable" process as the ongoing progress of a people toward collective self-awareness; second, his emphasis on conscious action constitutes an excessively limited view of culture and, methodologically, a rationale for minimizing the significance of the unconscious level of cultural change, which is the preponderant one.⁽³⁷⁾

The dimensions of collective self-awareness are further elaborated in Castro's conceptualization of culture, which he factors into two distinct constructs: the ".dwelling place" (morada vital) or "living structure" (vividura) of a people. The morada vital encompasses everything contributing to a feeling of selfness (ipseidad), including cultural artifacts, language, customs, values, and so forth; it more or less implies the standard anthropological definition of "culture." Vividura is the actual process [293] of feeling oneself part of a given culture, or morada vital: in other words, experience. Vividura represents the conscious level of the largely unconscious processes contributing to the formation of culture.⁽³⁸⁾ Within this broad framework, Castro offers only one mechanism for cultural diffusion and that is simple contact, convivencia ("dwelling together").

The development of the Spanish morada vital is the process that I have characterized as cultural crystallization, encompassing the entire process of culture formation from the moment of initial Christian-Muslim contact, through an extended period during which a new culture was blocked out, to the final stage of crystallization proper, when cultural boundaries were firmly established and the system as a whole became more rigid (eleventh century).⁽³⁹⁾ The crux of this process is that the recipient culture, in this case the Christian, must in some way adapt to the new situation (caused by the invasion) in such a way as to bring the structure of society and culture into accord with a new reality. It can seek new meaning in the values of the dominant culture or, if it finds these incongruent or incompatible, can create new meanings.⁽⁴⁰⁾ The first choice points to cultural borrowing, the second to reactive adaptation. In either instance, and here Castro is completely correct, the motive force for cultural change in Christian Spain was the Islamic presence.

The inevitability of adaptation occupies a properly central place in Castro's view: "The Islamic element interposed in the eighth century into Romano-Visigothic Hispania became a theme of vital moment to those who were not Muslims and were forced to adjust to those new circumstances. Thenceforward their existence would be woven from the present demands created by strange men and creeds as well as the exigencies of their own traditional habits."⁽⁴¹⁾ Insecurity, which Castro sees as the most characteristic element of the Castilian world view, was initially engendered in reaction to the cultural and political superiority of Islamic civilization and was subsequently fed by ambivalent relationships with Jewish and Muslim minorities, hated but indispensable.⁽⁴²⁾ Hostility among the three groups provided the stimulus for reactive adaptation. Castro makes it an underlying supposition that the only way for a medieval Christian, Muslim, or Jew living in Spain to feel that he was being himself was through opposing the other two ethnic groups.⁽⁴³⁾ From this flows a whole series of reactive adaptations: Christian exclusiveness in response to Jewish hermeticism; the accentuation of religion in general, in response to the theocratic nature of Semitic culture; and the antipathy of Christians [294] to manual labor in reaction to the perceived industriousness of Jews and Muslims. What Castro's notion of insecurity suggests is a kind of paring down of the repertory of norms, as values associated with Muslims or Jews came to be discarded as unworthy, a not untypical reaction by ingroups to outgroup success.⁽⁴⁴⁾

In Castro's view, Castilian culture had crystallized by the twelfth century, by which time the boundaries of Castilian ethnicity were settled and had become relatively impervious to additional Islamic influences. The influence of French culture, borne by Cluniac monks, was a sign that Islamic influences were waning, although Castro also indicates that the crystallization of Castilian culture had the further effect of rendering it more or less impervious to European currents as well.⁽⁴⁵⁾ The growth of intolerance is a further indication of the increasing rigidity of Castilian culture once its constitutive elements had been integrated. Tolerance, in this sense, was a function of the fluidity which inheres in a situation of culture creation under the impact of a more powerful alien group.

Castro's explanation of culture formation, terminological inconsistencies aside, is quite in line with the thrust of modern anthropological and sociological literature of culture contact and acculturation. Where his view is deficient, it is because he failed to see the social structure underlying cultural exchanges, the stage on which those exchanges took place. As van den Berghe notes:

"The analysis of ethnic relations must not be focused exclusively or even primarily at the cultural level; ethnic relations cannot satisfactorily be accounted for simply in terms of cultural differences, culture contact, and acculturation between groups. It is important to distinguish analytically the structural elements of ethnic relations from the cultural ones. The dynamics of group membership, solidarity, and conflict, and the network of structured relationships both within and between groups, are at least as essential to an understanding of ethnic relations as the cultural dynamics of group contact. People are not only 'carriers of culture'; they are also members of structured groups. Insofar as systems of ethnic relations are largely determined by structural asymmetries in wealth, prestige, and power between groups, an inventory of cultural differences gives one a very incomplete picture of group relations. Cultural differences are frequently symptoms rather than determinants of intergroup behavior, even in systems where the distinguishing criteria of group membership are predominantly cultural."⁽⁴⁶⁾

[295] Acculturation is a two-way process. Contacting groups tend to grow like one another, and elements are exchanged in both directions, although one pole may be stronger than the other. The direction and rate of flow may be determined by power relations, by demographic patterns, by differentials in economic, intellectual, or artistic development, and so forth. The social context of exchange also lends structure to the process. One would expect that whether different mechanisms of cultural diffusion were salient or not would depend on whether that diffusion took place across a political frontier or whether it occurred between a dominant group and an ethnic enclave. If the latter, the size of the enclave relative to the dominant group would also affect the cultural outcome. In al-Andalus through the tenth century, Romance elements were transmitted to Arabs through the conversion of indigenous peoples. Since the latter were a majority, the sheer weight of numbers meant that, through various non-formal means, considerable borrowing would occur. As conversion proceeded and the society became more homogeneously Muslim, the flow of Romance elements within al-Andalus steadily diminished as the processes of Islamization and Arabization were completed. The number and influence of Romance elements within Islamic society must have approximated a bell curve. We know that the culture of Nasrid Granada was wholly Arabophone and exaggeratedly Islamic, with the exception of specific groups, such as the military elite, who provided a market for Christian military and social innovations.

The demographic structure of intergroup relations effectively sets limits on the rate and direction of cultural flow. When Mozarabs were a majority, for example, one can expect those Arabs and Berbers who intersettled with them to have picked up Romance elements through normal daily interactions. When the enclave lost its demographic weight, however, then the dominant group's ability to acculturate minorities through formal, institutionalized means became increasingly apparent. Converts were acculturated through the molding influences of Islamic institutions, as remaining Christian pockets came under increasing pressure to convert, to learn Arabic, to conform, or to emigrate.

In Christian Spain the situation was different, and more complex. Ethnic enclaves were never a majority, except in certain times and places: e.g., Mozarabs in areas of León and post-conquest Toledo, Muslims in post-conquest rural Valencia. Except for the nuclei of Mozarab settlement, therefore, interaction with enclaves came late. Before the [296] mid-eleventh century, there was a great deal of cultural flow across political boundaries, the selective factor being determined by the differential development of Islamic and Christian societies. As the static agrarian society of the north grew and matured, there was a demand created for cultural innovation which at times took the form of cultural borrowing (imitation) and at times reactive adaptation. Such a situation cannot properly be covered under Castro's general rubric of convivencia. This latter refers to the later period of stabilized pluralism in places like Toledo, where cultural interchange between Muslims, Jews, and Christians was, in effect, managed by the latter. The early period was characterized by massive adaptation that had to affect the entire structure of Christian society, whether that adaptation was imitative or reactive in nature. The phase of convivencia (twelfth and thirteenth centuries) was characterized by extremely selective borrowing -- the translation movement is an adequate symbol. The contours of this movement were nevertheless extremely complex, and the flow was bidirectional, as Christians pressured Muslims and Jews to acculturate. Burns gives some interesting examples of formal pressures placed upon Valencian Muslims in the thirteenth century. The Christians saw the council of elders as a town council and forced its transformation accordingly. They also reintroduced the muhtasib, an office that had fallen into desuetude in late Islamic times, in the revivified form of the mustasaf, with its traditional jurisdiction but armed now with a standard, written code to execute.⁽⁴⁷⁾

All these nuances must be related to specific social (and especially demographic) variables. In Castro's explication, the processes are too telescopic to give an adequate explanation. In his telling, Jewish and Muslim societies ultimately failed in Spain because the Christians aspired "to be more,"⁽⁴⁸⁾ as if will alone and not such factors as power, wealth, numbers, or technology weighed in the victory. Ultimately, he indicates, the central theme of the Spanish middle ages was not the real "Reconquest" but an interior one, whose stage was the collective mind of Castilians, and the struggle among the three "castes" effectively took place within the Christian "caste."⁽⁴⁹⁾ This is a true statement only if the discussion be narrowed to the explanation of the emergence of group self-awareness as the key element in culture formation and the emergence of ethnic groups. But clearly many other processes were involved, many below the level of awareness.

Awareness of cultural differentiation as a reflection of the general [297] processes taking place assumed a variety of forms. We have noted Pedro Alfonso's awareness of the deficiencies of Latin scientific education. Arnald of Vilanova, less of a cultural intermediary than Pedro, masked his dependence on Islamic thought under the guise of criticism. Muslims were aware of their history of technological superiority. In a refutation of ibn García's Risâla claiming Neo-Muslim superiority, Abu'l-Talyib al-Qarawî wrote: "If you had contended for superiority with the Arabs in respect to erecting waterwheels, bending iron hooks, planting trees in enclosures, pruning overgrown branches, and working at the anvil, we would be satisfied and would acknowledge the justice of your claim."⁽⁵⁰⁾ But intellectuals are able to perceive the general nature of cultural processes and to distill them, while the workers at the anvil might well be less cognizant of the origin of their technical repertory.

Reyna Pastor asks rhetorically, with respect to Castilians of the late eleventh century, whether a people might have a level of saturation beyond which they cannot absorb alien cultural influences.⁽⁵¹⁾ The answer seems to lie in such concepts as social and cultural crystallization, which indicate that societies and cultures are most open to innovation in their formative periods, after suffering structural loss, and when, as a consequence, social and cultural norms are in a state of flux. Crystallization has the result of hardening cultural boundaries, making them less permeable. But the process of cultural change is continuous, and no culture is ever completely impervious to outside influences. Christians continued to borrow from Muslims in the late middle ages; as a result of the conquest of Granada, a whole new wave of Arabisms, many of them technological, entered the Spanish language.

Quantitative measures of cultural impact are difficult to come by and harder still to interpret. If one looks at the lexicon of the thirteenth century, the percentage of Arabisms in Castilian (as tabulated by Neuvonen) is quite low, less than half of one percent of the total vocabulary. Of the total, 14 percent entered the language before 1050, 15 percent between 1050 and 1200, and the majority entered the language in the thirteenth century.⁽⁵²⁾ These figures, of course, reflect Neuvonen's sample, which was literary in nature and encompasses more or less standardized vocabulary. Examination of regional documentation reveals many more Arabisms of a local or specialized nature. More significant is the breakdown of Arabisms by category (Table 6). Over a six-hundred-year period, the borrowing of terms related to social and administrative [298] institutions was preeminent in the process, an indication, in the first two periods, of the modeling of aless highly structured society after a more highly structured one. The figure for the third period is also high, given the fact that many of the other categories (plants and animal names, e.g.) reflect the translation movement. Also preeminent are agricultural terms. In a recent study of the contemporary agriculture lexicon in Andalusia, Julio Fernández-Sevilla expresses surprise that the number of Arabisms was lower than expected, given the Islamic agricultural heritage of that region.⁽⁵³⁾ Apparent here is the risk in using contemporary philological data as a historical source, for almost certainly there has been a steady loss of Arabisms since the seventeenth century, so that what [299] remain now are mainly the standard Arabisms accepted in the official language. Of the more historically significant survivals, many date to the Morisco, not the medieval, period. For example, almocafre (weeding hoe) appears with greater frequency in the Granadan heartland than in western Andalusia, although in Seville it appears to have been introduced by Moriscos displaced from Granada.⁽⁵⁴⁾ More significant than the descriptive terminology are the techniques themselves. Then too, vocabulary of whatever period does not reveal the full impact of culture contact upon language. Most suggestive is Yakov Malkiel's observation that "The oft-quoted circumstance that Arabic did not add a single phoneme to ancient Ibero-Romance, even if true, does not preclude the other, now frequently concealed, fact that the entire acoustic impression of Spanish underwent a revolutionary change through the introduction of polysyllabism (guadamacil, berenjena); words of comparable length without an analyzable prefix or suffix hardly existed previously in the vernacular."⁽⁵⁵⁾ In the final analysis, the quantity of the impact of Islamic upon Christian culture matters less than the quality. In discussing the diversity of the Islamic world, Maurice Lombard, using another linguistic analogy, referred to regional "Inflections" upon broader cultural styles.⁽⁵⁶⁾ An Arab Islamic culture in al-Andalus, inflected by Romance influences, a Christian zone inflected no less with eastern traits, adaptation on both sides of the cultural frontier in consonance with and reaction to the contact of cultures -- these are the distinguishing features of Spanish culture in the high middle ages.

1. Maravall, El concepto de España en la edad media, pp. 191-192.

2. Sánchez-Albornoz, "Tradición y derecho visigodos."

3. Bonnassie, La Catalogne, I, 71, 307.

4. Bolens, "Méthodes culturales," pp. 5, 14.

5. Guichard, "Les arabes ont bien envahi l'Espagne," p. 1493.

6. Hilda Grassotti, "Lo mozárabe en el norte cristiano como proyección de la cultura hispano-goda," Cuadernos de Historia de España, 33-34 (196l), 336.

7. Martínez Ruíz, "Notas sobre el arado bereber," p. 326.

8. Castro, The Spaniards, p. 180.

9. R. Menéndez Pidal, Orígenes del español, p. 555.

10. Ibid., p. 143.

11. See Makkî, Aportaciones orientales, p. 28o, and also section 3, below. Mones, "Hommes de religion," p. 71; Ashtor, The Jews in Muslim Spain, p.234; Pérès, Poésie andalouse, p. 52; Ibn Khaldûn, The Muqaddimah, III, 17.

12. Federico Corriente, A Grammatic Sketch of the Spanish Arabic Dialect Bundle (Madrid: Instituto Hispano-Arabe de Cultura, 1977), pp. 7, 34, 63, 126, 131 n. 223, 147, 148.

13. Ibid., p. 8.

14. On these incidents, see Makkî, Aportaciones orientales, pp. 255-258.

15. E. García Gómez, "Una extraordinaria página de Tîfâsî y una hipótesis sobre el inventor del zéjel," in Etudes d'orientalisme dédiées a . . . Lévi-Provençal, II, 519; Guichard, Al-Andalus, p. 504.

16. Urvoy, "Etude sociologique," pp. 248, 250, 253, 289.

17. On the distinction between formal and non-formal mechanisms of cultural change, see Foster, Culture and Conquest, p. 12; Glick and Pi-Sunyer, "Acculturation as an Explanatory Concept in Spanish History," p. 151.

18. On the ribât see Castro, The Spaniards, p. 473; Glick and Pi-Sunyer, "Acculturation as an Explanatory Concept in Spanish History," p. 152.

19. Mones, "Hommes de religion," p. 68; 'Abd al-Kaîim, Al-Andalus en el "Mu'yam al-buldân" de Yâqût, pp. 111, 128; Goitein, Mediterranean Society, I: 51-52; Verner, "La ciencia en el Islam y occidente," p. 543; Jacinto Bosch Vilá, El oriente árabe en el desarrollo de la cultura de la marca superior (Madrid: Instituto de Estudios Islámicos, 1954), pp. 16-20, 21 n. 2.

20. Makkî, Aportaciones orientales, pp. 195, 198, 280; 'Abd al-Karîm, Al-Andalus en el "Mu'yam al-buldân" de Yâqût, p. 96; EI², 1, 956-957; Mones, "Hommes de religion," pp. 68 n. 1, 71.

21. Ashtor, The Jews of Moslem Spain, I, 120.

22. S. M. Stern, "Les vers finaux en espagnol dans les muwassahs hispano-hebraiques," Al-Andalus, 13 (1948), 299-346; Dubler, La 'Materia Médica' de Dioscorides, I, 52.

23. Destombes, "La diffusion des instruments scientifiques," p. 51.

24. The 'Disciplina Clericalis' of Petrus A1fonsi, Eberhard Hermes, ed. (Berkeley: University of California Press, 1977), p. 72.

25. García Ballester, Historia social de la medicina en España, I, 35.

26. Giorgio Levi della Vida, "I mozarabi tra occidente e islam," in L'Occidente el'Islam nell'alto medioevo, pp. 677-682; Manuel C. Diaz, y Diaz, "La circulation des manuscrits dans la Péninsule Ibérique du VIlIe au XIe siècle," Cahiers de Civilisation Médiévale, 12 (1969), 384.

27. Diaz, "Circulation des manuscrits," pp. 224, 237, 391.

28. On the impact of French settlement, see Chapter 5, section 5, above. On cultural currents, see Bonnassie, La Catalogne, I, 326-340; Diaz, "Circulation des manuscrits"; Defourneaux, Les français en Espagne; J.M. Lacarra, "A propos de la colonisation franca en Aragón," Annales du Midi, 65 (1953), 331-342; Juan Ainaud de Lasarte, "Moissac et les monastères catalans de la fin du Xe au début du XIIe siècle," ibid., 75 (1963), 545-549; Anscari Mundó, "Moissac, Cluny et les mouvements monastiques de l'Est des Pyrénées du Xe au XIIe siècle," ibid., 75 (1963), 551-570; Ch. Higounet, "Un mapa de las relaciones monásticas transpirenaicas en la edad media," Pirineos, 7 (1951), 543-552; and Antonio Linage Conde, Losorígenes del monacato benedictino en la peninsula ibérica, 3 vols. (León: C.S.I.C., 1973), II, 863-997.

29. Castro, The Spaniards, p. 479; Sánchez-Albornoz, "Repoblación del reino asturleonés," p. 431. Christian kingship as a reflection of Islamic political structure is also Maravall's view (El concepto de España en la edad media, p. 386).

30. Castro, The Spaniards, pp. 233, 377, 105, 281, 243.

31. Ibid., p. 272. On parallelisms in Asia Minor, see Speros Vryonis, Jr., The Decline of Medieval Hellenism in Asia Minor and the Process of Islamization from the Eleventh through the Fifteenth Century (Berkeley: University of California Press, 1971), passim and explicitly on p. 460.

32. Pierre L. van den Berghe, Race and Racism: A Comparative Perspective (New York: John Wiley, 1967), p. 121.

33. Castro, The Spaniards, p. 405.

34. Ibid., p. 114 (emphasis mine).

35. Loc. cit. "Valuable" seems a mistranslation for "valued."

36. Ibid., pp. 39, 42, 36, 314 (emphasis Castro's).

37. The central role accorded to collective self-awareness is the major (although not the only) reflection of Wilhelm Dilthey's influence on Castro. Themes of history as self-knowledge (Selbstbesinnung) and experiential awareness (Erlebnis) are prominent in Castro's work; see The Spaniards, pp. 105-106 n. 6, 155 n. 15. On Dilthey's influence, see José Luis Gómez-Martínez, Américo Castro y el origen de los españoles: Historia de una polémica (Madrid: Gredos, 1975), pp. 120-125. Fortunately, Castro was able to contradict Dilthey's assertion that each generation forgets the experiences of its predecessors see Albert A. Sicroff, "Américo Castro and his Critics: Eugenio Asensio," Hispanic Review, 40 (1972), 28. Yet Castro also admits the functioning of unconscious processes. In a footnote, he states that colllective goals may be striven for consciously or unconsciously (The Spaniards, p. 126 n. 21); he admits that certain customs borrowed by Christians from the Muslims by reason of the latter's prestige were an "unconscious imitation (p. 272; also p. 567) ; and he recognizes that the constitutive values of the Spanish Christians, molded during the middle ages, survive today in a "collective unconscious" (p. 366). But such admissions are incidental to the main thrust of his argument.

38. See Araya, Evolución del pensamiento histórico de Américo Castro, pp. 62, 63 n. 105. Frequently, though, the two processes seem blurred, is when Castro (The Spaniards, p. 123) defines morada vital as "the creation by a people of a new awareness of its collective dimension, new with respect to the awareness expressed before by those who preceded it in that geographical space."

39. For the concept of cultural crystallization, see Foster, Culture and Conquest, pp. 222-223. Foster limits the term's applicability to the final phase of culture formation, after the blocking-out period.

40. For sociological formulations of the origins of ethnic awareness which are conceptually quite similar to those of Castro, see David Rothstein "Culture Creation and Social Reconstruction: The Socio-Cultural Dynamics of Intergroup Contact," American Sociological Review, 37 (1972), 671-678, and L. Singer, "Ethnogenesis and Negro-Americans Today," Social Research, 29 (1962), 419-432.

41. Castro, The Spaniards, p. 210.

42. Ibid., pp. 165, 204.

43. Ibid., p. 86.

44. Robert A. LeVine and Donald T. Campbell, Ethnocentrism: Theories of Conflict, Ethnic Attitudes, and Group Behavior (New York: John Wiley, 1972), p. 147: "Groups seeing themselves as unsuccessful in living up to their own ideas will manifest their greatest hostility toward outgroups that are successful in those particular ways."

45. Castro, The Spaniards, pp. 426, 523.

46. Pierre L. van den Berghe, Race and Ethnicity (New York: Basic Books, 1970), p. 150.

47. Burns, Islam under the Crusaders, pp. 386-394 (on town councils), 240 (on the mustasaf).

48. Castro, The Spaniards, p. 243 (Castro's emphasis).

49. Ibid., pp. 571, 122.

50. Monroe, The Shu'ûbiyya in al-Andalus, p. 81.

51. Pastor, Del islam al cristianismo, p. 118.

52. Neuvonen, Los arabismos del español, p. 305.

53. Fernández-Sevilla, Formas y estructuras en el léxico agrícola andaluz, p. 450.

54. Ibid., p. 318.

55. Y. Malkiel, review of R. Lapesa, Historia de la lengua española, 2nd ed., Romance Philology, 6 (1952), 62.

56. Lombard, L'Islam, p. 17.