Read an Excerpt
Islamic Spain 1250 to 1500
By L. P. Harvey
The University of Chicago PressCopyright © 1990 The University of Chicago
All rights reserved.
Geographical Environment and Historical Context
Problems of Terminology
If we wish to discuss the Muslims of Spain at any period in their history, we encounter a series of terminological problems. What are we to call the Muslims themselves, and the land they inhabited? The difficulties arise in part because the geographical and ethnic terms available to us themselves arose in the course of nine long centuries of contact and conflict, and while reflecting that conflict, they do not always help to throw much light on it. Moor (moro), for example, is a historical term which is authentic in the sense that it occurs in source materials of the period, but it is a term we can rarely use nowadays. It is not merely geographically imprecise, leaving us uncertain whether the person it describes is of North African origin or simply a Muslim, it is ambiguous with regard to the value judgment it implies Often Moor conveys hostility, but there are contexts where Muslims refer to themselves as Moors with evident pride. Even such an apparently straightforward term as Spain places pitfalls in our path. Does Spain include or exclude those places not under Christian rule? (And does it imply the inclusion or exclusion of what is now Portugal?). If we seek to escape from the difficulty created by Spain by using such an apparently unambiguous expression as "the Iberian Peninsula," we find we have created a new problem, because the inhabitants of the peninsula cannot be called Iberians (that is an expression limited to ancient history, or to the contrived ecumenicism of modern times).
If we adopt the Arabic name for the Iberian Peninsula, al-Andalus, we do not avoid all difficulties, because in English (and indeed in other modern European languages) the word is used of Islamic as distinct from Christian Spain, but since in Arabic the jazirat al-Andalus was the whole peninsula, there is always the possibility that the broader sense may be intended.
In the Iberian Peninsula in the Middle Ages two major civilizations came into conflict. The words we use to describe that conflict inevitably reflect it in their very semantic structures, and there is no way available to us to talk about what went on in neutral terms, nor is there any way of achieving flat uniformity except from the most biased of partisan viewpoints. What we must do is be aware of the ways in which our vocabulary may lead us astray.
Spanish possesses a rich vocabulary to designate the minority religious and cultural groups that existed in the peninsula over the centuries, and English has adopted many of these words. It seems at first sight that we have a tidy set of terms to meet each logical possibility. Besides straightforward Christians and Muslims, we need terms for (a) Christians living under Muslim rule, (b) Christians who are converts from Islam (or descendants of such converts), (c) Muslims living under Christian rule, and (d) Muslims who converted from Christianity (or their descendants). We do indeed find (a) Mozarab (mozárabe), (b) Morisco, (c) Mudéjar, and (d) muladí, but it is important to realize that these are not terms universally applicable over the whole chronological range of peninsular history, and each one of them has a problematical aspect. The question of the proper use of Mozarab(ic) affects a multiplicity of groups over a very long period of time (say, from 711 to ca. 1200). Some would apply it to all subject Christians without exception; the extreme opposite viewpoint restricts the word to those culturally Arabized Christians from the south who took refuge in León in the eleventh and twelfth centuries. The resolution of the problem of how the word Mozarabic ought properly to be employed must be left to specialists; the sole purpose of mentioning the word here is to point out that it gives rise to disagreement and controversy. Of the other three terms mentioned, it should be noted that two, muladí and Morisco, are subject to chronological limitations. By and large muladí is used only of converts from Christianity to Islam in the days of Umayyad rule. (Later the term elche is sometimes used, but it seems to imply that the converts in question are outsiders to the community that has adopted them.) Morisco in the special sense indicated at (b) is not in use until well after 1500 (modern historians may use it from 1500 or so, but it is doubtful whether it actually came into use until about mid-century: before that the Moriscos avant la lettre were simply referred to as nuevos convertidos).
The word Morisco illustrates particularly clearly the dangers inherent in vocabulary. To make the point very briefly (and there would be a great deal more to say), by employing this term Morisco (rather than Spanish Muslim, say) we are tacitly accepting and approving of the forcible reclassification of this group of Muslims as something other, although all the documentary evidence indicates that most (not all) of them continued until the end to be crypto-Muslims who would have rejected the conversion if they had been at liberty to do so. The very use of the word seems to signify assent to the marginalization of the people so designated. Not to use it, however, is to court misunderstanding, as it is a standard part of the Spanish historical vocabulary.
The term that will concern us most in the course of the present study is Mudejar: "a Muslim who, after the surrender of a territory to a Christian ruler, remained there without changing religion, and entered into a relationship of vassalage under a Christian king" (a slightly modified translation of the entry Mudéjar in the Diccionario de la Real Academia Española). Are we, by referring to such Muslims as Mudejars, also endorsing their marginalization? As we shall see, the problem is really more complicated than that posed by Morisco.
Against the view that Mudejar is a term that relegates those so described to an inferior status, it may be argued that in recent critical usage the word has occurred in neutral and purely descriptive senses, and indeed some ascribe to it a strongly positive value judgment. In art history, Mudejar architecture has been much esteemed; even in literary criticism, López-Baralt has discerned positive aspects in "Mudéjar" writing "from Juan Ruiz to Juan Goytisolo" (the latter a twentieth-century author preaching what he sees as a mudejarismo for our times).
A difficulty facing those who would advocate a revaluation of Mudejar lies in its etymology. Etymology is no absolute bar to upward mobility, of course, and words which have begun as terms of opprobrium may sometimes be adopted as proud banners. (In British history, the rise of Tory out of the bogs of Ireland provides a particularly striking example of the phenomenon.) Yet etymology can be a useful guide to early usage and is not to be ignored. Let us see what it tells us about Mudejar.
The phonetic changes the word has undergone mean that its origins are by no means transparent, and erroneous guesses have abounded. The attempt to associate the word with Arabic dajjal, "he who leads astray," an epithet applied to Satan, can be discounted at the outset as an obvious clumsy product of anti-Islamic polemics. To associate it with Spanish dejar "to leave" (presumably because these Muslims were "left behind") implies that dejar was borrowed in a rather strange way: we have no evidence that the Arabic of Spain adopted dejar as a loan word. The obvious place to look for light on this problem is in Arabic texts themselves speaking of the Mudejars; usually they simply have Muslims, but a few, notably al-Wansharishi's Kitab al-Mi'yar, use the expression ahl al-dajn, "people who stay on" (al-Wansharishi 1908:XII, 198). As will be seen below (chapter 4), al-Wansharishi is an author who strongly disapproves of the very idea of Mudejar status, and we can be sure that in his usage it carries a charge of opprobrium. Mudajjan, the participial form, would have the same sense as ahl al-dajn, namely "having stayed on."
There is another aspect of the semantics of this word. Cognate to mudajjan are forms such as mudajin and dawajin which are used of "domesticated" or "tame" animals, particularly poultry. It would seem that a semantic link exists between the animal, which passes under human protection and acquires "domesticated" status, and human populations, which pass under the protection of a dominant power and become safe subjects. To be tame and domesticated is not a characteristic of which any man will boast. What we have here is probably a taunt, perhaps at first directed by free Muslims at subject coreligionists. Since the new "subject Muslim" status was one for which no existing term was available, the name stuck and even passed into Spanish usage.
Can the term Mudejar be part of our modern historical vocabulary? "Muslim subject to a Christian ruler" is a clumsy expression; a word is needed, and it would be perverse not to employ the one that is to hand, but we must be aware that there are contexts in which it will not function as a bland and colorless adjective, that it can be an emotive and possibly, on Muslim lips, an offensive epithet. (The theological implications of the acceptance by Muslims of Christian rule are discussed in chapter 4.)
Problems of Demography
How people are to be counted has always been a problem, and one of the difficulties experienced by medieval historians is that sometimes their sources will be making honest if unsuccessful attempts to reflect reality as it was perceived, while at other times they will be using numbers because of the symbolic or magical properties attributed to them. The preeminence of mathematics in our culture and the reliance we put on statistics place historians under pressure to find ingenious ways of extending back into the past sequences of demographic statistics of the same kind as we expect to have available for the twentieth century. That is rarely, if ever, possible.
We are far from possessing reliable statistics for any of the periods or areas that concern us. Occasionally, and for certain areas, we know a great deal, but alongside the extraordinarily detailed information which some medieval archives do preserve, there are black holes of ignorance which will never be filled. It is always necessary to remind ourselves that thirteenth-century statistics are not the same as twentieth-century statistics (which are not the same as the full and objective truth).
In general, of course, the nearer we approach the present day, the more information we have. We would expect to have more information about the sixteenth century than about the fifteenth, but there is also the special circumstance that during that period the "Morisco question" was on the agenda of Spanish statesmen and administrators. When the Final Expulsion of the Muslims came in the early seventeenth century, the consequent movement of populations was documented as fully and as bureaucratically as the Final Solution of the Jewish question in the twentieth. Thus the careful and comprehensive studies of Muslim populations in Spain in the final period have been made on the basis of full and apparently reliable statistics. Henri Lapeyre, in his Geography of Morisco Spain (Géographie de l'Espagne Morisque), was able to assemble a remarkably complete set of figures for almost all the Muslims who were expelled in 1609–1611, and to extend this coverage to earlier periods of the sixteenth century.
Lapeyre's survey can often be used to project even further backwards into the later Middle Ages information about the distribution and relative sizes of Muslim settlements, but it has to be employed with all due precaution, for the sixteenth century was a period of particularly large movements of population (affecting the Muslims of Granada more than others, but leaving few areas untouched).
We may form some idea of the degree of uncertainty with regard to population statistics at about the outset of our period by comparing two estimates. J. N. Hillgarth estimates as follows: "Reckoning back from an estimated 7,000,000 or 8,000,000 in 1482, it is conjectured that the population of Castile-León in 1225 was about 4,000,000 (another estimate is 3,000,000) and that the conquests added 300,000 more" (Hillgarth 1976: 1, 30). J. O'Callaghan (after remarking on the hazardous nature of population estimates) cites Vicens Vives' estimates of the population of "Castile, the largest of the peninsula states, at about the middle of the thirteenth century, as four or five million people, including about 300,000 Muslims and Jews" (O'Callaghan 1975: 459). Estimates for the Crown of Castile, we see, vary between 3 million and 5 million. As O'Callaghan points out, "For the crown of Aragon there are a number of statistical records that scholars have used to advantage. On the basis of the census of households of 1359, the population of Catalonia has been estimated at about 450,000 people, including about 18,000 Jews and 9,000 Muslims, located mostly in the south. Vicens Vives believed the population had dropped from about 500,000 in the thirteenth century because of emigration to Valencia, Majorca, Sicily, Sardinia and Greece. The estimate for Valencia is about 300,000, for Aragon about 200,000 and for Majorca about 45,000. This would give a total population of one million in the crown of Aragon" (460). He goes on to put all the figures together into a global estimate for the peninsula as between 6,200,000 and 7,200,000 before the Great Plague. The proportion of Muslims in these total populations is likewise not known with accuracy. Both O'Callaghan (1975:462) and Hillgarth (1976: 1, 32) quote the estimate given by Sobrequés that about half the population of the crown of Aragon were Muslims. "This very high figure is less improbable than some others given for Jewish population, but it is reached by the same process of arguing back from figures of later centuries, in this case from those calculated for the Moriscos expelled in the seventeenth century. Only in the Kingdom of Valencia (and its extension southward to Alicante) did Mudejars constitute a large majority of the population: in Majorca they may have formed 50 per cent. They were very numerous in Aragon proper" (Hillgarth 1976: 1, 32).
As for Granada itself, the figure of 200,000 given by Aragonese envoys at the Council of Vienne in 1314 (O'Callaghan 1975: 460) is undoubtedly far too low, although we have no reliable figures. Hillgarth declines even to conjecture what it might have been; Ladero gives an estimate of 300,000 for the years before 1492 (Hillgarth 1976: 1, 30).
If we add to the population of approximately 5,000,000 in Castilian lands, the 1,000,000 for Aragon, we have perhaps 6,000,000 in total (with a considerable margin of uncertainty, as we have seen). The population of Muslims may have been the 300,000 for the Crown of Castile. (O'Callaghan [1975: 459] gave 300,000 for Muslims and Jews together; his estimate seems low bearing in mind the numbers of Muslims remaining behind at least until the Mudejar revolts of 1264.) For the Crown of Aragon one-half of the total population of 1,000,000 is clearly too high, but 300,000 is probably on the low side bearing in mind the dense settlements in Aragon proper and in Valencia. For the Kingdom of Granada we can only guess at least 300,000 or more. The total for Muslims in the whole peninsula must have approached 1,000,000. Perhaps one in six of the total population in the whole peninsula may have been Muslims. (Navarre with its steady small population of about 100,000 in all, and perhaps 1,000 Muslims, does not really affect our estimates as these figures fall well within our margins of error.)
If these very rough estimates are anywhere near the mark, they give rise to considerable problems when we seek to relate them to estimates for earlier periods in the Middle Ages. Thomas Glick's survey of these problems may be cited verbatim: "Assuming that there were seven million Hispano-Romans in the peninsula in 711, and the numbers of this segment of the population remained level through the eleventh century, with population growth balancing out Christian migration to the north, then by 912 there would have been approximately 2.8 million indigenous Muslims (muwalladun) plus Berbers and Arabs. At this point Christians still vastly outnumbered Muslims. By 1100, however, the number of indigenous Muslims would have risen to a majority of 5.6 million" (Glick 1979: 35).
Excerpted from Islamic Spain 1250 to 1500 by L. P. Harvey. Copyright © 1990 The University of Chicago. Excerpted by permission of The University of Chicago Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.