[185] When Alfonso VI led his army out against the combined forces of Muslim Andalucía and North Africa in the fall of 1086 he was behaving in a fashion that would continue to characterize him down almost to his death in 1109. Faced with a threat, he chose to seize the initiative. There could have been no doubt in his mind that the objective of his slowly advancing enemy would be the city aud the taifa of Toledo. The defense of that hilltop fortress, set in the horseshoe loop of the Tajo, would have been a sound move for a more cautious spirit. He had full and plenty of time to stock and garrison it and to expel the more untrustworthy of its citizens if need be. The campaigning season was already well advanced. Unless the allies were able to overwhelm the city quickly they would be faced with the rigors of a winter siege, far from their own base of supphies, and with a hostile army in the field.

As an experienced commander, Alfonso would not have been insensible to any of those considerable advantages. Nonetheless he chose to take the offensive as he would time and again thereafter. Al-Maqqarí, quoting an earlier source, reports what was supposedly the gist of the Leonese monarch's reasoning and argument in his curia. "If we wait and engage the enemy here and then lose," it ran, "we lose not only the battle but this land. But if we fight in his territory and lose, it may be that he will fear to press his advantage. He may fear to advance and leave the passes at his back. At the very least he will have to take time for fresh preparations. If we win, on the other hand, we can take advantage of our victory there."⁽¹⁾ While these may be the platitudes of military strategy in the eleventh century, we can also recognize that these are the platitudes reflected in Alfonso's actions.

Doubtless that monarch advanced other considerations of a more pohitical nature. The Muslim kings of Andalucía were both fearful and jealous of one another and of their new found ally. Their army would reflect those divisions. The slowness of the enemy advance already testified to its problems in that regard. A defensive strategy would give [186] the Murâbit leader time to consolidate his position politically and to integrate his forces more fully. An offensive strike and quick victory would destroy his credit and their unity at a blow. Moreover, not only would a cautious defense open the territories of Toledo to devastation, but the delay in forcing the taifas to resume payment of the annual parias would quickly impair the Christians' abilities to field just such an army as had been the foundation of their successes for the last thirty years.

The behavior of Yusuf ibn-Tashufin since his landing at Algeciras on July 30, 1085, would have seconded Alfonso's arguments. The Murâbit leader's first movements had been to secure Algeciras itself as a base and possible retreat before advancing to al-Mutamid's Sevilla. Welcomed there by that strongest and most able of kings of Andalucía, Yusuf summoned all of Spanish Islam to his standards.⁽²⁾ The response was not reassuring. From the northeast al-Mustain of Zaragoza and his uncle, Mundir of Lérida-Tortosa, were probably sympathetic but could not or would not answer. Al-Qadir of Valencia remained loyal to Alfonso. Even in the south the taifa of Almeria made excuses, and the reason why the allies advanced by way of Badajoz may be that al-Mutawakkil had not committed himself and so was to be coerced to do so. Abd Allah of Granada only joined the allied army when it had advanced to Jerez de los Caballeros enroute to Badajoz.

The chronology of Yusuf's movements is quite unknown. If we may safely assume that he had reached Badajoz by October 20, his slow rate of advance gave evidence of many difficulties, for the best of the campaigning season was already past. The Murâbit leader had now been in the peninsula just ten days short of three months. In that time he had advanced perhaps as much as 500 kilometers, depending on the route, through exclusively Muslim territory. A sustained march would have covered the same distance in three weeks at most.

Having arrived in Badajoz, Yusuf waited there to receive the Christian attack. The lone eyewitness, Abd Allah, who wrote an account of the campaign, relates that the Murâbit leader intentionally refrained from a farther advance and even decided against harassing the advancing enemy because he wished the battle to be fought in friendly territory and with the city at his back as a possible place of refuge. Being so new to Andalucía, we are told, he was not yet sure who were his allies and who his adversaries. The emir was later credited with an army of [187] seven thousand horse from Africa in addition to local forces and infantry. That seems clearly excessive.⁽³⁾

Since the battle of Zalaca is of such importance, it seems worthwhile to spend some time in consideration of the army with which Alfonso VI advanced to combat. The chroniclers of either side are only occasionally or incidentally of help in this respect. Like western medieval armies generally its essential component was its heavy cavalry operating in squadrons of forty to sixty men on the battlefield.⁽⁴⁾ That size unit seems to have been optimum in that it was sufficiently massive to have a predictable impact on contact and yet was the largest body amenable to control by voice or signal. The key to estimating the size of the cavalry element in the army, then, is to find some way in which to gauge the number of squadrons involved.

I would suggest that this may be done by utilizing the county and the bishopric as the likely units from which these squadrons were drawn. We know that in León-Castilla the county remained an important fiscal and legal unit and that the bishopric ordinarily comprised substantial economic, political, and legal strengths. Now there were at this moment some fifteen bishops in the realm and, as best one can determine, thirteen counts. On the assumption that in this emergency the king was able to draw upon half of the military force of the realm and was also willing to risk such a commitment, the bishops and counts together would have furnished fourteen squadrons of cavalry. If we allow two more for the four Castilian magnates who were prominent enough in the documents to suggest that each could have mounted one, another [188] for the Aragonese contingent, yet another for Alvar Fáñez's troops, and a final squadron composed of the royal bodyguard under the alférez, a grand total of nineteen cavalry troops is reached.

On such a basis, and allowing an average of forty knights per squadron because of the haste with which they had had to be summoned, Alfonso's cavalry force might have amounted to something like 750 horsemen with another 750 more lightly armed horsemen potentially in reserve. For we must remember that each knight would have been accompanied by a squire whose first duty would have been to succour his master in combat but who himself could be fed into the battle upon his lord's death or capture.

There is no evidence that an infantry force separately organized as such accompanied the army. Nevertheless the logistics of the army guaranteed the presence of foot soldiers even if they were not trained to operate in groups. For each knight and squire pair we can probably predicate a third groom-body servant. In an age when any healthy man could draw a bow to some effect and even may have had rudimentary training in handling a spear, the army was automatically supplied with skirmishers to the number of a rough seven hundred. To these must be added the men of the supply train for no army of this size would operate without one.

An absolutely minimum allowance must furnish eighteen carts to each squadron: one each for its commander's tent and for the cook and his helper, three for dry provisions, one for firewood; two each for water, for wine, and for animal fodder; four for munitions, and two for a blacksmith, his helper and his supplies. The size of the army, the nature of the terrain, and the hostility of the inhabitants combined to make all these necessary. Too great a dependence on lucky foraging could lead to disaster under these circumstances. Even then the need for abundant water for the close to three thousand head of horses and livestock dictated that the route of march should lie west along the Tajo and then a quick push south, through modem Cáceres, to the waters of the Río Guadiana above Badajoz. The carters and miscellaneous personnel involved in this supply function thus probably added another 325 men to the army. This contingent of foot would have performed perfectly normally if they served as a camp guard once battle was joined. There are no sources to prove that the carts themselves were arranged to form a rough, fortified precinct but it would be strange if they were not. Very little tactical sense is requisite to suggest placing a solid obstacle between one's self and a potential attacker or to create a continuous line of barriers by placing the shafts of one cart on the body of the next.

[189] Such then was the army with which Alfonso fought the battle of Zalaca on October 23, 1086. Based on the careful study of Huici Miranda, it is possible to assert that the tactics of the combatants were simple, straightforward, and traditional, though subsequent literature was to add much colorful and untrustworthy detail. Yusuf must have been confident in his superiority of numbers for he advanced across the Río Guadiana to the plain of Zalaca. Alfonso, in turn, seeing the enemy with a river at his back, could not resist the temptation to score the kind of major victory that position offered. He launched the traditional frontal cavalry assault designed to break and scatter the enemy. After some initial success against the Andalusian formations, he found himself unable to follow up with a decisive breakthrough.

Fighting must have become general all along the front for Yusuf was now able to execute a flanking movement and to overrun the Christian camp. That success was more psychological than material, but Alfonso must have seen that he lacked sufficient forces to break the enemy ranks and that to continue the attempt would be to invite envelopment and complete disaster. He did manage to cut his way free to a line of retreat but a running fight then developed, which probably continued until nightfall with some intensity, for the open terrain offered no good place for a temporary stand to discourage pursuit. Abd Allah indicates that most of the Christian casualties occurred during this retreat, as might be expected. He also indicates the duration of the action indirectly when he comments on those who died simply of exhaustion from the weight of their armor.⁽⁵⁾ Nevertheless, Alfonso was able to break off contact under the cover of night and to reach Coria, some 125 kilometers to the northeast, probably about October 25, 1086.

Both Alfonso VI and Yusuf ibn-Tashufin had now to assess realistically the consequences of the battle and what these latter permitted. Although it had been a Muslim victory most assuredly, Yusuf chose to retire to Seville and from thence to North Africa. He probably judged quite correctly that the season was too late for campaigning in Christian territories and that the Christian army remained intact if bloodied. The Murâbit now realized that real progress would depend on a common front by the taifa kings, which he was at present unable to effect and they themselves were unwilling even to attempt. He, at least, could hardly have been jubilant over the events of the summer.⁽⁶⁾ He did apparently leave behind a sizeable cavalry force to stiffen the resolve of his coreligionists.⁽⁷⁾

[190] The Leonese monarch seems not to have been unduly discouraged by this initial defeat. Toledo was safe for the time being and the most formidable of his enemies had departed for Morocco, perhaps never to return. But he himself had received a serious wound in the leg which penetrated to the tibia.⁽⁸⁾ That was no light matter for a man of forty-nine. Other casualties are hard to specify. The circumstances of the battle had so developed that all those on foot had surely perished. The heads of these unfortunates are those of whom the Muslim chroniclers speak, traveling in cartloads as trophies in Andalucía and Africa. But how many of those grisly mementos were the heads of knights and magnates? We can be positive only that one of Alfonso's counts, Rodrigo Múñoz of Galicia, was slain at Zalaca.⁽⁹⁾ However, Count Vela Ovéquez of western Asturias and Galicia does not appear in the documents after June 1085.⁽¹⁰⁾ It should also be noted that Bishop Vistruario of Lugo disappears after February 1085 and Bishop Ederoño of Orense after November 1085.⁽¹¹⁾ If all these were, in fact, killed at Zalaca it would indicate that the sector of the Christian ranks manned by the Galicians were particularly hard hit. Nothing in the history of the year following, however, suggests that the Leonese army was almost destroyed. A judicious guess might be that Alfonso lost roughly three hundred knights and squires together or one in every five of his mounted warriors who participated in the battle. Such an estimate assumes that the four counts and bishops indicated above were casualties at Zalaca and that the proportion of magnates killed was approximately the same as the number of knights as well.

Since Alfonso could not immediately have been assured that Yusuf was departing the peninsula, be returned to Toledo from Coria, which had been his first refuge. The city on the Tajo was a prime target for the enemy and would have had to have been put in condition for a possible siege. it seems that a general curia was held there at the end of the year, one of whose actions was the consecration of Archbishop Bernard of Toledo on November 6, 1086.⁽¹²⁾ Certainly every action was discussed and many taken, which would strengthen the city and its leadership for [191] the difficult days ahead. One of these surely was the endowment and consecration of the cathedral on December 18, 1086; a day which also was the major festivity of the Virgin, its patroness, in the old, Mozarabic liturgy.⁽¹³⁾ The curia probably also saw the formal reconciliation of Alfonso with Rodrigo Díaz de Vivar. The Cid had been summoned home and was now entrusted with the castle of San Esteban de Gormaz on the upper Duero and, more strangely, with that of Dueñas on the Pisuerga below Palencia, if the literary sources can be trusted.⁽¹⁴⁾

Sometime in January, we may safely presume, Alfonso began to move north since the immediate danger to Toledo had disappeared. It was to be a year of frenzied diplomacy for the Leonese monarch but one that has left few reliable traces in the documents. For the sense of it we are dependent upon a wide variety of literary sources for the most part.

Upon learning of the landing of the Murâbit army in the previous summer, Alfonso had surely appealed for aid from beyond the Pyrenees. That it should reach him before Zalaca was impossible, but a major French expedition did appear in the peninsula in the early spring of 1087. The French chronicles reveal that Alfonso's plea had struck a responsive chord. This army included contingents led by Duke Eudes I of Burgundy and Guillaume le Charpentier, viscount of Mélun. Well before its arrival, however, the Murâbits had left the peninsula and Alfonso had decided to continue a policy of moderation toward the Muslim of Andalucía. The French forces were accordingly persuaded to spend their energies in a siege of Tudela, the northern bastion of Zaragoza. This they did with notable lack of success, and their army began to disperse toward home in April 1087.⁽¹⁵⁾

Nevertheless the siege was apparently accompanied by some very serious diplomacy in the tents of the attackers. Alfonso VI and Sancho [192] Ramírez of Aragón both seem to have participated, and it was probably there that Sancho and his son and heir Pedro did homage to Alfonso and promised to help him in the future defense of Toledo. In return the Leonese monarch abandoned his claims to some of the old lands of the former kingdom of Navarra and agreed to the erection of a county of Navarra under Sancho Sanz, of mixed Aragonese and Navarrese parentage, over yet other of those territories. It has even been asserted that Alfonso renounced his ambitions toward the taifa of Zaragoza but, as we shall see, the immediate sequel to the siege can hardly be read in that fashion.⁽¹⁶⁾

The tents of the besiegers also witnessed some marriage diplomacy of the greatest import. It seems entirely likely that at least the preliminaries of the betrothal of Alfonso's heir and daughter Urraca, were first discussed there. But the king also had, in Elvira, an illegitimate daughter who was subsequently married to Count Raymond IV of Toulouse. It has been asserted that this magnate accompanied the expedition of 1087, but his latest historians have been unable to document that statement. What can be proven is that Elvira first appears as his wife in 1094.⁽¹⁷⁾ It may very well be that their betrothal too was discussed there. Certainly the Leonese king would have been anxious to seize every opportunity to enhance his prestige and position in the face of the new threat developing in the Muslim south.

In addition, Alfonso's allies from beyond the Pyrenees would also have been aware of that ruler's marital difficulties and of the attractive prospects their exploitation might offer to the adventurous. As we have already seen, sometime in 1077 Alfonso had decided to put away his first queen, Inés of Aquitaine, who had not given birth to any heir who had survived in approximately four years of marriage. That decision had occasioned difficulties with Pope Gregory VII, but eventually the pope had accepted the king's marriage to Constance of Burgundy in 1079, who was probably with child in the spring of 1080 when a mutual agreement was reached. Unfortunately, Urraca, born most probably in late 1080, was to prove to be the sole surviving offspring of the union of Alfonso and Constance. If other children were born to them, such royal progeny had not survived infancy long enough to have attracted the attention of the chroniclers. The lack of a male heir would have been a major disappointment already in 1081, and it would have been doubly so if Urraca's birth were attended by complications such as made it unlikely that the queen could subsequently give birth successfully. [193] Yet a second annulled marriage would have created major diplomatic problems both in Rome and in the south of France. Constance of Burgundy would therefore remain queen down to her death in 1093.

Nevertheless, the necessities of the dynasty made it imperative that the king should take a mistress. I suspect that this took place in 1081 or 1082. It has been suggested that this liaison occurred in the brief period between the annulment of the marriage to Inés and the new nuptials with Constance, but that idea has little but gallantry to recommend it.⁽¹⁸⁾ The royal choice fell upon one Jimena Múñoz, whose father seems to have been Muño Múñoz, a magnate of western Asturias and of the northern Bierzo who seems to have been styled "count" everywhere but in the contemporary documents.⁽¹⁹⁾

Everyone in the court and the kingdom, Queen Constance included, would have understood the royal desire for a male heir. The king was then in his late forties and the imperatives of the dynasty could wait no longer. But the expedient proved ultimately to be futile as Jimena gave birth to two children only, who survived, and both were daughters. Even though child marriages were not uncommon in the period, it does seem likely that Infanta Elvira had been born before 1087 if she was married by 1094. Whether or not her sister Teresa, the future queen of Portugal, had been born is less certain because we cannot surely date her marriage before 1096. In any event, royal councils must have considered the state of the king's marital affairs with wry humor and quiet desperation.

To adventurers from beyond the Pyrenees, however, the prospects were increasingly attractive. León-Castilla was a kingdom of truly regal extent. It was eminently subject to partition as its recent history had demonstrated. It was ruled by a fifty-year-old king, with but daughters as heirs, who needed allies to compensate for his recent defeat and prospective dangers. If Raymond of Toulouse was at Tudela in 1087, the circumstances would never have been better for him to negotiate a future [194] marriage with an eye to the appropriation of a less troublesome principality than the one he was to abandon in 1095 at Toulouse or the one he was to undertake to conquer from 1102 at Tripoli. But the premier prize was to fall to the house of Burgundy, whose ties with the Leonese monarchy were of longer standing.

Giving up the feckless siege of Tudela in April, the Duke of Burgundy, unlike the other French leaders, proceeded west to León. He was probably in the company of Alfonso VI at Burgos on July 21, 1087, when the king issued a charter of immunity for the possessions of a monk of San Millán de La Cogolla.⁽²⁰⁾ By August 5, 1087, the Burgundian magnate had reached the royal city of León. There he confirmed a charter granted by Queen Constance, his cousin, to the Burgundian abbey of Tournus.⁽²¹⁾ But the major purpose for which he had come to León was to complete the arrangements for the betrothal of his cousin, Raymond of Burgundy, to Alfonso's only legitimate heir, Infanta Urraca. When Raymond of Burgundy and his cousin, Henry of Burgundy, first arrived in the peninsula is obscured in a tangle of documents, but the totality of the evidence and the circumstances indicate to me that it was with the expedition of the latter's brother in 1087 and that a marriage contract between Raymond and Urraca was effected in that summer.⁽²²⁾ Such a step was an understandable progression in the [195] long-standing relationship between the dynasty of León-Castilla, the house of Burgundy, and the monastery of Cluny and also would have suitably disposed of potentially troublesome rivals for Duke Eudes.⁽²³⁾ It was nevertheless a very drastic step for it necessarily entailed the at least implicit recognition of Raymond as heir to the realm if the marriage should indeed take place and if the king should continue to want for a son.

The rebellion that broke out in Galicia in 1087 and 1088 is best understood as an effort to thwart such an eventuality. Of course the rebels' only possibility of achieving their ends lay in the liberation of the king's brother García, who had been held prisoner by Alfonso since early 1073 in the castle of Luna high upon the Río Orbigo in northwestern León. Galicia, it will be recalled, had been García's kingdom. Moreover, the one Galician prelate who was implicated in the rebellion, Bishop Diego Peláez of Santiago de Compostela, was an appointee of García.⁽²⁴⁾ We are poorly informed about the events of the revolt but one thinks that an attempt to free the royal prisoner must have been involved; otherwise the movement was doomed from the beginning. In the Europe of the monarchy even practical success in the field could only be consolidated and legitimated by appeal to that principle.

A French cleric attached to the episcopal court of Santiago de Compostela and writing better than thirty years later provides a cryptic and curious note on the matter.⁽²⁵⁾ The charge against Bishop Diego Peláez, he says, was that "Gallaeciae Regnum prodere Regi Anglorum et Normannorum et auferre Regi Hispanorum satageret."⁽²⁶⁾ Now any student of the last years of William the Conqueror of England will immediately realize that between the desertion of his heir, the threats of the Danes,
and the war with France, the personal intervention of William in the north of Spain in the last days of his reign would have been inconceivable. But our source is too good to have been utterly incorrect.

[196] What is possible is that the rebels cloaked their actions under an appeal to the rights of a daughter of the Conqueror whom García had sought as a bride years earlier and to whom he may have even been betrothed.⁽²⁷⁾ They may or may not have been in communication with the English king, but the point of the liaison would have been the respectability of their cause in León-Castilla rather than practical assistance from abroad. In either event, the death of William of England in September 1087 would have been fatal to the rebel cause as soon as news of it should reach the peninsula. Since García remained a prisoner and none of the principals involved had a sufficiency of either prestige or power in their own right to make a serious stand against the king, their case must have become desperate as soon as it lost even the pretense of representing a royal cause.

Since documentary references to the revolt are all retrospective, it has been difficult to date it precisely. Menéndez Pidal believed that it had two stages and that only the second took place in the spring of 1087 while Alfonso VI was campaigning in the south.⁽²⁸⁾ My own conviction is that the entire episode was much more compressed, that it constituted a reaction to the betrothal of the king's heir, Urraca, to Raymond of Burgundy, and that it could not have erupted before the late spring of 1087.

Conditions propitious for a revolt may have existed in Galicia from the late fall of 1086 if, as suggested above, the defeat of Alfonso VI at Zalaca also involved the death in battle of major figures in the ecclesiastical and secular hierarchy of the region. So it may have been no coincidence that, while Alfonso was busy to the east with the siege of Tudela and its concomitants in 1087, his two sisters were looking after the interests of the dynasty in Galicia. On April 25, 1087, Elvira made a donation to the church of Compostela and on May 30, 1087, her sister Urraca did the same. Both charters were issued in councils of some considerable extent.⁽²⁹⁾ Considering that it was only the grantee, Bishop Diego Peláez, who possessed sufficient prestige to launch the revolt, [197] we must conclude that he could hardly have raised the standard of García before late June of that year.

Presuming the initial success of the rebellion in that ecclesiastical señorio, we then find it spreading east along the pilgrimage road to the hilltop fortress town of Lugo. For what were at least in part private purposes of their own, the local magnate family of the Ovéquez joined the movement and stormed the town, killing the royal merino who probably led the defense. Whether or not Bishop Vistruario had been killed at Zalaca, the see seems to have been empty at the time. Such a vacancy would have disorganized the defenders of what was, for practical purposes, another episcopal señorio and have invited the Ovéquez to seize the advantage once again in their longstanding feud with the bishops of Lugo.⁽³⁰⁾ There is no sign, however, that the rebels were able to extend their successes beyond that town and its environs.

That failure is not terribly surprising. If Count Rodrigo Ovéquez and his brother, Count Vela Ovéquez, were powerful magnates in the lands about Lugo, still Galicia was full of counts and magnates, all jealous of one another's successes. Neither of the former were major court figures, although Alfonso will complain that he nurtured Rodrigo at court in the latter's youth. They therefore lacked the widespread influence and acquaintances that would have made them more dangerous. Indeed, Vela Ovéquez may himself already have died at Zalaca, further reducing the power of that clan.⁽³¹⁾ As a practical matter the whole enterprise is inexplicable except as a desperate reaction to the betrothal of Urraca, as an intrigue that involved the reigning dynasty itself in the person of García, and a threat, ultimately unconvincing, of foreign intervention.

Given the limited material resources of the rebels, the problem they posed for Alfonso VI was primarily a political and diplomatic one. Doubtless he reacted with military force as well but we cannot be sure whether or not the king felt constrained even to take the field himself. After Burgos in mid-July, we have no trustworthy notices of Alfonso's whereabouts for the remainder of the year.⁽³²⁾ Menéndez Pidal believed [198] that Alfonso had campaigned in the south in the spring of 1087, but the sole document he adduced in support of this assertion need demonstrate no more than that forces were summoned from as far as Coimbra for the siege of Tudela.⁽³³⁾ There is, in fact, no other trace of such a campaign.⁽³⁴⁾ The only offensive of note against the Muslim recorded for this year concerns Sancho Ramírez of Aragón. After the breakup of the siege of Tudela he turned his energies to the east where he overran Estada, ten kilometers to the northeast of Barbastro in the valley of the Río Cinca.⁽³⁵⁾

Alfonso apparently wintered in the region of Sahagún for the court seems to have been there on January 16, 1088.⁽³⁶⁾ From that point he moved southeast to celebrate a great curia of the realm at Husillos, some eight kilometers north of Palencia, in early March. His forces had already broken the back of the rebellion in Galicia, its leaders were prisoners in his hands, the news that William I of England was dead would have long since reached his ears, and the king was ready to make a suitable disposition of this and other matters.

On March 11, 1088, Alfonso VI granted a charter in which he endowed the monastery of San Servando, just across the Tajo from Toledo, and ceded it to the papacy but with the provision that it should be permanently administered by the abbot of the monastery of Saint-Victor of Marseilles, who would pay an annual tribute in gold to Rome for that privilege. Alfonso stated that he did this on the counsel of "domni Richardi, cardinalis rome scilicet et abbatis massiliensis ecclesie." The confirmants included the archbishop of Toledo, the bishops of Palencia, Burgos, Astorga, and Oviedo, and four counts.⁽³⁷⁾ This document [199] was surely issued in connection with the curia and council of Husillos.⁽³⁸⁾ The only document that survives beside this former is the text of the boundary, decided there, between the dioceses of Burgos and Osma. In addition to its confirmation by the king, by the cardinal legate, and by the archbishop of Toledo, it was also confirmed by seven bishops, by four bishops-elect, by three abbots, by three abbots-elect, by four counts, and by seven other magnates, here styled "princeps."⁽³⁹⁾

The very fact that this delimitation of boundaries was taken up indicates that a certain tension already existed between the new archbishop of Toledo, who administered the territories of Osma, and the bishop of Burgos. That dispute had to be subsumed in the more general political settlement which was the business of the curia. The military problem of the revolt in Galicia had already been solved. As we are told by the Historia Compostelana, Bishop Diego Peláez of Santiago de Compostela was brought before the bishops and legate in chains and caused to surrender his pastoral ring and staff and to declare himself unworthy of the episcopate. Probably as a consolation to the bishop of Burgos and his supporters, Abbot Peter of the Castilian monastery of Cardeñas was chosen as the new bishop of Santiago and confirmed the document mentioned above as "bishop-elect."⁽⁴⁰⁾ At this same time Count Rodrigo Ovéquez was probably exiled to Zaragoza.

The problems of Galicia were further addressed as the council presided over the choice of a new bishop, Pedro, in Orense. There is no mention of the bishop Juan Alfónsez who had appeared in the charters of the infantas Urraca and Elvira in 1087, so perhaps he was a creature of Bishop Diego Peláez and shared in the latter's disgrace.⁽⁴¹⁾ It is conceivable that the plot had had ramifications that reached even so far as Coimbra in the southwest for Bishop Paternus was now replaced in that see and city by Bishop-elect Martín.⁽⁴²⁾

Other ecclesiastical appointments seem to have been more routine. [200] Though the monasteries themselves were of great importance, no special significance seems to have attached to the selection of Abbot Diego at Sahagún, or Abbot Juan at Oña, and at Cardeña, of course, the new Abbot Pedro replaced his namesake, now bishop-elect of Santiago de Compostela. More mysterious is the appearance of one Sigefredus as bishop-elect of Nájera.⁽⁴³⁾

By April 7, 1088, the council had ended and the court had moved on to Valladolid.⁽⁴⁴⁾ Then, on April 30, 1088, Alfonso made yet a further endowment of the Toledan monastery of San Servando.⁽⁴⁵⁾ The charter was confirmed by Queen Constance, Archbishop Bernard, six other bishops, seven counts, and the royal majordomo and alférez, indicating another general curia by our definition. While there is nothing to indicate the place of issuance, Sahagún rather than Toledo is the likely scene. It appears that neither Cardinal Richard, his traveling companion Archbishop Peter of Aix, nor Bishop Gomez of Burgos, all of whom had been present at Husillos, were now at court. Perhaps they had departed to consecrate the new monastery church at Silos, which they are known to have effected that year.⁽⁴⁶⁾ Since the monastery of San Servando was now Cardinal Richard's to administer it does seem that he would have figured in this charter had he been present. Evidence for the presence of the court at Sahagún exists for May 11, 1088.⁽⁴⁷⁾

From that central location Alfonso VI seems to have traveled up into the hills of Galicia. At Lugo on June 18, 1088, the king confirmed the possessions of the church of Lugo in a charter that also recited briefly the history of the rebellion of the Ovéquez.⁽⁴⁸⁾ Unfortunately this document has been redrawn from an earlier authentic grant and there is no [201] good way to check the accuracy of its account.⁽⁴⁹⁾ From what we already have established about the outbreak of the rebellion in 1087 and its defeat before March 1088, it would seem that the return of Count Rodrigo Ovéquez from exile in Zaragoza and his final, futile stand in a castle at Ortigueira on the edge of the Bay of Biscay occurred during the late spring of 1088. After his benefaction at Lugo, Alfonso continued down the pilgrim road to Santiago de Compostela to see the installation of the new bishop, Pedro of Cardeña, there.⁽⁵⁰⁾ At this same time he probably also granted a charter to the Compostellan monastery of San Martín de Pinario, the notice of which is dated only to 1088.⁽⁵¹⁾

There in the city of Santiago the news would have reached him that the Murâbits had once again landed in the peninsula. Yusuf ibn-Tashufin landed at Algeciras in May or June of 1088 and undertook his unsuccessful siege of Aledo.⁽⁵²⁾ This second descent of the emir of the Muslim was a reaction to the very active defensive strategy employed by Alfonso following the defeat at Zalaca. To protect his eastern flank the Leonese monarch had despatched Rodrigo Díaz de Vivar to Zaragoza in 1087. Presumably this followed the lifting of the siege of Tudela in April whose only result would have been the agreement of al-Mustain of Zaragoza to resume payment of the parias and to return to his former posture of vassal of the Leonese realm. The Cid and al-Mustain now marched to the relief of Valencia, another of the vassal taifas of Alfonso VI in the east. There they were successful in forcing the ever-recalcitrant Mundir of Lerida to raise the siege and withdraw.⁽⁵³⁾ By the autumn [202] of 1087, then, the buffer kingdoms of Zaragoza and of Valencia had returned to their allegiance to León and had resumed the payment of the all-important tributes.

The recovery of Leonese fortunes had taken place behind the shield offered by the Christian advance post at Aledo. This castle situated in the mountains of the southeast threatened both Muslim Cartagena and Murcia. Because of its position the rebel ibn-Rashiq was able to defy al-Mutamid of Sevilla and to maintain himself as an independent ruler in Murcia. The rebel received the recognition of Alfonso VI and paid him parias. This state of affairs had become intolerable to al-Mutamid in particular, and his appeal had resulted in Yusuf appearing once again in Andalucía.⁽⁵⁴⁾ The siege of Aledo which lasted the entire summer of 1088 is an excellent demonstration of the advantages of the defense over the offense during the medieval period, particularly when the besiegers are badly divided among themselves.⁽⁵⁵⁾

Alfonso VI must have been well aware of the difficulties of the Muslim for he himself acted with great deliberation. Huici Miranda believed that he had campaigned in the south on the headwaters of the Guadalquivir at Ubeda and Baeza in the spring of 1088 but, as we have already seen, the Leonese monarch was fully occupied in the north until summer had well begun.⁽⁵⁶⁾ After his return from his tour of Galicia in the late spring he must have spent some time organizing the repopulation of Segovia.⁽⁵⁷⁾ That site was of major importance for communications between the pacified north and the newly conquered south for it guards the northern end of the 1,800-meter pass of Navacerrada through the Guadarramas which leads down toward Madrid and Toledo. The resettlement of Segovia represented a long step south in the trans-Duero, the town lying some fifty kilometers southwest of Sepúlveda and the same distance southeast of Cuéllar. Thus its repopulation was of major importance to the elaboration of communication and mutual support both in the trans-Duero and between it and the meseta of Castilla la Nueva.

This effort would seem to have preoccupied the king for most of the summer for as late as September 18, 1088, he was at Sahagún with his court, presumably directing the movement of supplies south.⁽⁵⁸⁾ Even if [203] this was the eve of departure for Aledo, Alfonso's army could hardly have covered the roughly 350 kilometers and have arrived at Toledo before the beginning of October. To advance from that place to Hellín where he was supposed to have joined forces with the Cid would have meant another 280 kilometers of travel and so he cannot have arrived there before about October 15, 1088. By this time, however, the Muslim forces before Aledo were decamping amidst mutual recriminations and Alfonso was content to see them withdraw, discouraged and having accomplished nothing. Yusuf ibn-Tashufin would return to Africa more than ever convinced that success in Spain would first require a drastic reordering of Muslim Andalucía.

The Leonese monarch returned to the north, in the first instance to Toledo.⁽⁵⁹⁾ It would then have been possible for him to have proceeded to the far northeast, but we know of no reason for his having been there. At any rate a charter of Alfonso's to the monastery of San Millán de La Cogolla would seem to place him in that region on November 25, 1088.⁽⁶⁰⁾ The king and court are found in a place more to be expected when the former rewarded a faithful retainer, at León in all probability, on December 27, 1088.⁽⁶¹⁾

The frustration of the plans of the Murâbits in 1088 was far from constituting a solution to the problems of Alfonso of León as the shock waves of the defeat at Zalaca continued to buffet the pillars of his realm. Indeed the events surrounding the relief of Aledo throw new light on one of those problems that was developing with particular speed. While the sources for the incident leave much to be desired, it is clear that all are agreed that Alfonso VI had summoned the Cid to rendezvous with him on the march to the relief of Aledo and that the Cid had failed to do so. Though the literary sources would explain that failure [204] as an honest mistake, they nevertheless agree in affirming that it eventuated in an open rupture between the king and his vassal. From the documents of the time it is clear that the break was never repaired for after the spring of 1088 Rodrigo Díaz will never again appear in a royal charter. As is usual with the literature of the Cid, the motivations are represented as personal and Rodrigo Díaz is cast as the loyal, if misunderstood, vassal, and Alfonso as the jealous and overbearing lord.⁽⁶²⁾ We need not be concerned with the niceties of literary creativity.

Both the relationships between the king and his vassal and the exact role of the latter in the eastern peninsula had long been ambiguous. The Cid's activities in that arena had nevertheless redounded to the advantage of the Leonese monarch even if indirectly. Yet Alfonso must have been concerned about the independence of action that Rodrigo enjoyed, and the king's employment not of the Cid but of Alvar Fáñez in a crucial role at Valencia in 1085-86 may be a measure of that concern. But the despatch once again of the Cid to the east in 1087 set in train a process that would see Rodrigo cease to be a subject and become instead an ally.

In the aristocratic society of the eleventh century every adult male of the nobility and gentry was by definition a warrior. All of them had the ability to conduct private wars proportional to the strength and cohesiveness of their family ties, to the extent of their estate's resources, and to their personal prowess which drew to them a mesnada, or warrior band. Major campaigns, however, ordinarily remained in the preserve of the crown, which alone could levy on the wealth of the church, whose fisc far exceeded the estates of any noble, whose charisma was assured by virtue of office, and who retained control of the public resources of warfare at this period. These latter, the fossata, or right to military service, and the fossataria, or right to a monetary payment in lieu of military service, were levies reserved to the crown, to the counts and bishops at the behest of the crown, or at worst available to rebels who had or claimed the support of some member of the reigning dynasty.⁽⁶³⁾

The extent of any or all of these resources, for king or noble alike, at any given time before the modern period is impossible of determination. It can only be observed that, in a marginal and almost totally agricultural society such as was the Christian north of Iberia in the eleventh [205] century, all were easily exhausted and renewable only as the slow and unhurriable rhythms of nature allowed. But the disintegration of the Caliphate of Córdova at the beginning of the century had, paradoxically, provided the Christian monarchs with a whole new means of financing military operations without wasting their own resources or provoking unrest by levies on those of their subjects. The wealth of the relatively prosperous Muslim taifas furnished a strong enough lure to facilitate the raising of a volunteer army, given the increasing population and its pressures north of the Duero, motivated by the prospect of booty. The successful subjection or the regular harassment of the taifas then produced the payment of those parias which allowed the maintenance of the same armies independently of what we may call "feudal" resources.

Again, the parias were ordinarily the prerogative of the king alone and further enhanced his relative superiority over the greatest magnate. What seems peculiar to the latter career of the Cid in the circumstances following Alfonso's defeat at Zalaca is his ability to appropriate the parias of Valencia to himself. Acting first as an agent of his king, with a mesnada swollen by his established reputation as a warrior, he was early set on the road to becoming an independent prince by the victories that allowed him to support that army out of the surplus of Valencian society.⁽⁶⁴⁾ Despite the confused chronology of the sources, it seems clear that Rodrigo Díaz was, formally or informally, exacting tribute essentially for the support of his own army by 1088.

Alfonso would have grudgingly accepted that practice in an emergency. But real autonomy quickly imposes its necessity of independent assessment of aims and of priorities among aims. We cannot tell, and in truth it is not essential to our understanding to know, if the Cid missed his rendezvous with the king in 1088 by simply misunderstanding, or because at some point he decided that Aledo would withstand the siege without him, or because he considered his continued presence vital in the Levant. What Alfonso VI could not accept was the very fact that his subject should have presumed to rest satisfied with such a state of affairs or judgment. The king's reaction signals the already extant position of his former subject as an independent power, albeit a not unfriendly one.

Yet another problem with its ultimate genesis in the defeat at Zalaca continued to trouble the realm and its master at the end of 1088. In the summer of that year Archbishop Bernard of Toledo traveled to Rome [206] to seek the approval of the accumulated ecclesiastical business of three years. Chief among that prelate's concerns, and those of his royal master, were to secure papal approbation for the restoration of the primatial see of Toledo, for the selection of Bernard as its new archbishop, and for the decisions of the Council of Husillos, including the deposition of Bishop Diego Peláez of Santiago de Compostela and the selection of Pedro of Cardeñas as bishop in his stead.

All of these innovations and adjustments had resulted from the initiative of the crown as a close examination of the documents makes clear. "Ego," Alfonso says, "concedo sedi nietropolitane, scilicet, Sancte Marie urbis Toletane honorem integrum ut decet abere pontificalem sedem secundum quod preteritis temporibus fuit constitutum a sanctis patribus" in the charter of endowment of the new primatial dignity.⁽⁶⁵⁾ In the one act of the Council of Husillos that remains to us the introduction speaks of "presidente domno Ricardo, vicario sancte Dei Romane Ecclesie," but its decision is reached "cum concilio et consensu supradicti catholici regis" and the document itself appears as a royal document issued by "Ego Adefonsus divina preordinante gratia hispaniarum rex supra notatam divisionem fieri volui manuque propria firmavi" while the papal legate merely confirms the royal act.⁽⁶⁶⁾ Concerning the affair of the bishop of Santiago, the language of the "Historia Compostelana" is straightforward and unembarrassed, "Quem episcopum praedictus Rex Alfonsus expullit ab Ecclesia B. Jacobi."⁽⁶⁷⁾

Now Alfonso wished simply that all these royal determinations should be duly ratified by the Roman authority, but his earlier encounters with that spiritual power had educated him in some degree as to the necessity for propriety of form at least. To that salutary caution probably was due the sudden reappearance on the scene of Abbot Richard of Saint-Victor of Marseilles, papal legate for Spain, after an absence of seven years. Although one could argue that his commission had expired with the death of Gregory VII who had appointed him, no other legate had been appointed at least. It looks very much as if someone at court remembered his pliability in 1081, arranged for him to be summoned again to court, and that he himself was willing to be complaisant, especially if rewarded with the revenues of the Toledan monastery [207] of San Servando. So in some rough sense the decencies of ecclesiastical law were observed even in the breach. The archbishop of Toledo was then despatched to Rome to shepherd all this potentially controversial business through the papal offices. On the way he stopped at Cluny to fortify yet further the royal case by securing the blessings of Abbot Hugh for the effort.⁽⁶⁸⁾

As luck would have it, the fortunes of the reformed papacy were in great disarray. Following on the death of Gregory VII on May 25, 1085, there had been an extended vacancy in the papal office as the reformers debated strategy and tactics. From the Castle of San Angelo, Guibert of Ravenna argued the merits of his claim as Clement III. When finally the reform party had elected Desiderius, abbot of Monte Cassino, in May 1087, the new Victor III only lived until September of the same year. After more hesitation the reformers turned to Eudes, former prior of Cluny, who was elected in March 1087 and took the name Urban II. When, therefore, Bernard of  Toledo appeared before him, Urban was ill-situated to alienate any support for his own claim to be rightful pope against that of the imperial candidate. Moreover, his own sympathies as a former Cluniac must have been automatically with the arguments of a fellow Cluniac supported by the venerable Hugh the Great himself.

Thus the series of decisions embodied in the papal letters drawn up on October 15, 1088, at Anagni were largely a legitimization of what had already transpired. Bernard was given the pallium as a token of the papal recognition of his election and consecration and the primatial rights of his see were explicitly confirmed. Another letter advised Abbot Hugh of Cluny of this action. A third letter similarly alerted the archbishop of Tarragona and the other Spanish archbishops of the reinstitution of the primacy.

The fourth letter, addressed to "Alfonso, king of Galicia," strikes a different note. It begins with a recitation of the Gelasian formula of the two powers but does go on to inform the king of the acceptance of the restored primacy of Toledo and of Bernard as its archbishop. Coming then to the matter of the deposition of the bishop of Santiago de Compostela, Urban rejects both the process and its results. He expresses sadness that the king should have seized the bishop and had him deposed as entirely contrary to the canons, "a te captum, et in captione ab episcopali dignitate depositum, quod canonibus noveris omnino contrarium." [208] Nor are these actions to be excused by the invocation of the authority of a papal legate for, Urban tells the king, Abbot Richard had no such authority, having been deprived of his authority by Pope Victor III. In a neat touch, the pope orders the bishop of Santiago to be restored to his see "per Toletanum archiepiscopum dignitati" and signals his willingness to entertain the king's complaints against the prelate by requesting that Bishop Diego be sent to Rome along with a royal embassy so that the matter may be canonically adjudicated.⁽⁶⁹⁾

With the hapless subjects of the crown, Rome was less gentle. A letter to the clergy and people of Compostela forbids them to accept or obey Pedro of Cardeña as their bishop and imposes an interdict on the diocese of Santiago until the deposed Bishop Diego shall be set at liberty. A companion letter to Pedro of Cardeña informs the bishop of his deposition and summons him to Rome for judgment. The Liber Pontificalis confirms the fact of the interdict levied at this time.⁽⁷⁰⁾

Urban II nevertheless was far from seeking a general quarrel. In addition to his acceptance of Alfonso's candidate for the primatial see of Spain, the pontiff was also willing to overlook the questionable role of the king's chief instrument in the entire affair, Cardinal Richard. Of February 20, 1089, Urban confirmed, to "karissimo fratri Ricardo," the privileges and possession of his monastery of Saint Victor of Marseilles, including the right of administering the papal monastery of San Servando outside Toledo.⁽⁷¹⁾ The precarious character of the reform party's hold on the papacy at this point is reflected in the fact that apparently Urban had no available records of what his predecessors had or had not done in regard to Spain. Otherwise, he would not have written on October 15, 1088, to "Terraconensis et caeteris Hispaniarum archiepiscopis" about the restoration of the primatial see of Toledo when in fact none of the other archepiscopal sees had been restored. In addition, the pope and his chancery had only the haziest knowledge of even the general state of political affairs in Spain as reflected by the fact that the letter to Alfonso VI could not even accord the latter his proper title.⁽⁷²⁾

But if the debility of the papacy offered advantages and opportunities [209] to the Leonese king, Alfonso could not have been pleased with the state of affairs by the beginning of the new year in 1089, by which time the papal letters of October had surely been received. Generous though Urban II had been, at least from the papal standpoint, the monarch of León-Castilla must have regarded the demand for the reinstatement of Bishop Diego Peláez of Santiago as the papal defense of a proven traitor. Even worse, his own subjects would see in his accession to the papal order a major reverse to, and limitation on, the royal authority. Full compliance would be prohibitively costly from the royal viewpoint, but outright defiance was unthinkable in the circumstances of the moment.

Ominously, the Murâbit power of North Africa had again intervened in peninsular affairs for the second time in two years. León-Castilla continued to wage an aggressive and highly successful defense against the combined forces of Andalucía and North Africa, but the initiative had clearly passed to the Muslim in the south. Just as unequivocally the taifa of Valencia in the east had become a tributary of the Cid and the profits of its trade and agriculture were being assessed to support the forces of that Castilian adventurer, not those of the crown.

At the same time the new primate of Toledo and the bishop of the see of Burgos, a diocese born of the Reconquista, without a classical tradition, and scarcely emergent from its progenitor at Oca, continued to jostle one another in what was to become a major test of strength. The crown could hardly afford to take sides in that struggle. Meanwhile, the erstwhile abbot of Cardeña had been summoned to Rome for judgement and the king must decide whether or not to let him go. The see of Santiago lay under a papal interdict and its ambitious new Romanesque cathedral was scarcely more than a handful of piers, melancholy and silent in winter rains of the northwest.

Worst of all, the king still had no male heir. That simple fact made all of the great magnates restive. The logic of dynastic monarchy in the Middle Ages, in Spain as elsewhere, made possible a rational policy for noble and bishop only so long as a male heir existed. That known, one knew who one's enemies were likely to be, whose friendship was to be accorded most useful; in short that the familiar rhythms of political life would perdure. Without a male heir suddenly anything became imaginable. The fall of the great and meteoric ascent of the modest became ever more possible as the relatively fluid rules of marriage and wardship, regency and minority promised to rule the game. The vast majority of Alfonso's subjects dreaded such a prospect scarcely less than the king himself. It was the worst of times.

1. History of the Mohammedan Dynasties in Spain 2:281

2. Throughout the account of this campaign I follow Ambrosio Huici Miranda, unless otherwise noted. This classic account reviews the sources with a sure hand. Where I supplement it is in the interest of highlighting topics not germane to that author's purpose.

3. Abd Allah, El siglo XI en 1^a persona, p. 202. This is the source that inspires the greatest confidence by its general sobriety. For numbers on the Muslim side, see Ambrosio Huici Miranda, Colección de crónicas arabes de la Reconquista, vol. 4 (Tetuán, 1955), p. 101.

4. For general purposes see John Beeler, Warfare in Feudal Europe, 730-1200 (Ithaca, 1971), pp. 169-80. Philippe Contamine, La guerre au Moyen Age (Paris, 1980), pp. 118-19, 135-36, and 140, is useful for establishing reasonable sizes for such armies and overall resources of the crown. For the Iberian peninsula see my remarks and bibliography in León-Castilla under Urraca, pp. 274-77. The size of the operational held unit is almost never specified except by indirection as when al-Kardabus speaks of sixty knights scouting Almeria, in History of the Mohammedan Dynasties, p. xxxiii append. C, or the Cid is mentioned in the Cantar as going into exile accompanied by seventy lances. Military obligations owed by towns sometimes specify a contingent of sixty. See DMP 1-1:439. The chroniclers seem to take it for granted that armies will be structured about a body of heavy cavalry numbering between 300 and 750 when the need for dramatic effect is not preponderant in determining their language. Antonio Ubieto Arteta, el, Crónica Najerense, p. 110. Enrique Flórez, ed., "Historia Compostelana," ES 20:122. Menéndez Pidal, ed., Primera crónica general 2:502. Antonio Palomeque Torres, "Contribución al estudio del ejército en los estados de la Reconquista," AHDE 15 (1944): 205-351, is largely concerned with legal obligations.

5. Abd Allah, El Siglo XI en 1^a persona, p. 203.

6. Ibid., pp. 203-205.

7. Ibn-Kardabus, "Kitab al-Iqtifa," p. xxxiii append. C.

8. Elías Gago and Juan Eloy Díaz-Jiménez, "Los restos mortales de Alfonso VI y de sus cuatro mujeres," BRAH 58 (1911): 43-47, believe that a more serious wound came twenty years later.

9. Sánchez Alonso, ed., Crónica del Obispo Don Pelayo, p. 62.

10. June 25, 1085. AHN, Códices, 1.044B, fol. 92r.

11. Feb. 5, 1086. See note 49. Nov. 25, 1085. See chapter 8, note 18.

12. "Los Anales toledanos I," in Huici Miranda, ed., Las crónicas latinas de la Reconquista (Valencia, 1913), p. 343. Rivera Recio, pp. 133-34, had some reservations since consecrations were ordinarily performed on Sundays and this was a Friday, but the circumstances were surely difficult enough to justify an exception.

13. In contrast to my earlier opinion, I have finally come to the conclusion that this most-copied of all eleventh-century charters is a pseudo-original. Pub. José Antonio García Luján. ed., Privilegios reales de la catedral de Toledo, 1086-1462, vol. 2 (Toledo, 1982), pp. 15-20. The diplomatic of the notary Sisnandus Astruariz is rendered in a fashion typical for him except that in the intitulario Alfonso is given as "Esperie" rather than "Ispanie imperator." However, a Bishop Pedro of Nájera confirms instead of that Sancho who continued to preside in the see. May 6, 1087. Serrano, ed., Cartulario de San Millán, pp. 268-69. July 21, 1087. Ibid., pp. 269-70. Also, a Bishop Cresconius confirms for Coimbra just about six years too soon as we shall see subsequently. The document was doubtless copied from an original, which it probably embellished in other, less obvious ways.

14. Menéndez Pidal, España del Cid 1:341-45, and Huici Miranda, Grandes batallas de la Reconquista, p. 85.

15. J. Saroïhandy, "La légende de Roncesvaux," in Homenaje ofrecido a Menéndez Pidal, vol. 2 (Madrid, 1925). pp. 200-62. Boissonnade, Du nouveau sur la Chanson de Roland, pp. 33-35.

16. Ubieto Arteta, "Homenaje de Aragón a Castilla," pp. 7-28.

17. John H. Hill and Laurita L. Hill, Raymond lV de Saint-Gilles (Toulouse, 1959), p. 19.

18. José M. Canal Sánchez-Pagín, "La Infanta Doña, hija de Alfonso VI y de Gimena Múñoz a la luz de los diplomas," AL 33 (1979): 271-79.

19. Our contemporary source is tantalizing as usual. Sánchez Alonso, ed., Crónica del Obispo Don Pelayo, pp. 86-87. "Habuit etiam duas concubinas, tamen nobilissimas, priorem Xemenam Munionis, ex qua genuit Geloirain, uxorem comitis Raimundi Tolosani" The connection was made as long ago as the sixteenth century. See Ambrosio de Morales, Viaje a los reinos de León, y Galicia, y principado de Asturias, ed., Enrique Flórez (1765; reprint, Oviedo, 1977), p. 112. Augusto Quintana Prieto, "Jimena Múñiz, madre de Doña Teresa de Portugal," RPH 12 (1969): 223-80, tried heroically to bring Jimena out of the shadows. Bishko, "The Cluniac Priories of Galicia and Portugal," p. 325, believed her still alive as late as 1120. Segl, Kónigtum und Klosterreform in Spanien, p. 160, had doubts about this last.

20. Serrano, Cartulario de San Millán, pp. 269-70. Another Alfonsine charter to San Millán proper, dated only by year to 1087, was probably executed at this time. Ibid., pp. 270-71. Also likely of the same period was the king's confirmation of an earlier charter of his brother, Sancho II, to the monastery of Silos. See Férotin, ed., Recueil des chartes, p. 29. Two other charters, which would place Alfonso VI elsewhere during the spring of 1087, are clearly forgeries. Apr. 25, 1087. See Reilly. "Alfonso VI of León-Castile," p. 14, n. 92 and 93. May 14, 1087. See chapter 8, note 18.

21. Notice in Ernest Petit, Histoire des Ducs de Baurgogne, vol. 1 (Dijon, 1885), p. 406. Pub. Pierre François Chifflet, Histoire de l'abbaye royale et de la ville de Tournus (Dijon, 1664), pp. 331-32, with a prefatory section mentioning the expedition and siege.

22. Menéndez Pidal, España del Cid 1:340-41, agrees. For the general background see Marcelin Defourneaux, Les français en Espagne aux XI^e et XII^e siécles (Paris, 1949). Apr. 18, 1073. BN, Manuscritos, 5.790, fol. 111r, where Raymond is already cited as son-in-law of Alfonso is clearly to be redated to 1103 on the basis of earlier confirmants. Oct. 17, 1077. AHN, Clero, Carpeta 557, no. 15, is a fourteenth-century notarial copy in which Raymond's probable later confirmation as count of Galicia has been incorporated as a contemporary one. May 13, 1084. Pub. Santiago Montero Díaz, ed., "La colección diplomática de San Martín de Jubia," Boletín de la Universidad de Santiago de Compostela 7 (1935): 62, is a badly dated thirteenth-century copy of a private charter as is clear from the reference to Alfonso reigning in Toledo. Dec. 11, 1085. A notice in Prudencio de Sandoval, Antigüedades de la ciudad y iglesia catedral de Túy (1610; reprint, Barcelona, 1974), clearly refers to the famous charter of Feb. 11, 1095, which has troubles of its own as we shall see later. Aug. 17, 1086. Acad. Hist. Catedrales de España, 9-25-1-C-3, ff. 13r-14r. A donation of Raymond as count of Galicia properly dated to Aug. 21, 1096. Oct. 28, 1086. Acad. Hist., Colección Salazar, 0-16, fol. 551r-v. A sixteenth-century copy of a private donation which cites Raymond both as count of Galicia and spouse of Urraca, it is probably misdated. Mar. 30, 1087. AHN, Códices, 1.047B, fol. 1r; also Acad. Hist., Sobreira Salgado, 9-21-6-117, no pagination. Pub. Montero Díaz, "Colección de San Martín de Jubia," pp. 63-64. Again, this is a private document which could be accepted as reflecting in some fashion the news of the negotiations before Tudela, refashioned later to make Raymond already count and royal son-in-law. Apr. 25, 1087. See note 20.

23. Apr. 18, 1088. AHN. Clero, Carpeta 960, no. 2, indicates that some provision was made for his younger brother, Henry as well. "fuit maiorinus de comite enric."

24. See chapter 2, note 63.

25. Reilly, " 'Historia Compostelana,' " pp. 80-83.

26. ES 20:254.

27. See chapter 3, notes 45-50.

28. España del Cid :34-47. Hilda Grassotti, "La ira regia en León y Castilla," CHE 41-42 (1965): 29-32, wrestles with his chronology without achieving one of her own.

29. AC Santiago. Tumbo A; ff. 34v-35r, and fol. 34r respectively. Both have been published by López Ferreiro, Historia de Santiago de Compostela, 3:25-27 and 27-30; and also appear in Estefanía "Memorias," pp. 213-14, dated to 1093 and 185-87. This latter is an unpublished manuscript of the Acad. Hist., 9-27-2-E-50. Both charters are confirmed by an otherwise unknown bishop of Orense, Juan Alfónsez, whose consecration may have preceded the first of them since the bishop of Braga, who was metropolitan of Orense, was in attendance as well. The charters are somewhat irregular in form but the infantas did not employ chancery clerks invariably.

30. For a summary of town government in Santiago de Compostela and in Lugo, see my León-Castilla under Urraca, pp. 333-38. Alfonso VI had issued a judgment for Bishop Vistruaro and against the Ovéquez as far back as Feb. 28, 1078. AHN, Códices, 1.196B, ff. 37r-38v. Pub. ES 40:417-22, and López Peláez, El señorio temporal de los obispos de Lugo 2:111-14.

31. See note 10. López Ferreiro, Historia de Santiago 3:155-58, overestimated the family's strength and followed an impossible chronology.

32. Aug. 7, 1087. AHN, Códices, 1.044B, ff. 30r-31v, contains a brief notice of a donation he supposedly made to the church of Mondoñedo in Galicia. The dating formula apparently gave the place of issuance as Toledo. It seems unlikely that Alfonso would have left the duke of Burgundy to his own devices at León. Another document from the same chartulary, AHN, Códices, 1.044B, ff. 79v-80r, places the bishop of Mondoñedo in Galicia on Aug. 1, 1087. Nov. 26, 1087. Pub. Fidel Fita, "San Miguel de Escalasa. Inscripciones y documentos," BRAH 31 (1897):479-481. This private document is confirmed by a number of court figures but the royal alférez, Gomez Gonzalez, appears five years too soon.

33. España del Cid 1:345-46. For the document, dated Mar. 15, 1087, see PMH, Diplomata, pp. 404-405.

34. Huici Miranda, Grandes batallas de la Reconquista, p. 85.

35. Ubieto Arteta, Historia de Aragón, pp. 94-98.

36. AHN, Códices, 989B, fol. 231r. AHN, Clero, Carpeta 885, no. 24, dated only to 1088 may be from the same period.

37. AHN, Códices, 989B, fol. 66v; 996B, fol. 53r-v, and another copy on fol. 54r. Pub. Martin Guérard, ed., Cartulaire de l'abbaye de Saint- Victor de Marseille, vol. 2 (Paris, 1857) pp. 184-86, and also by Fidel Fita, "El monasterio toledano dc San Servando en la segunda mitad del síglo Xl. Estudio critico," BRAH 49 (1906): 281-83. The latter mistakenly lists "Sisnandus conibriensis episcopus cf." for "Sisnandus conibriensis consul cf."

38. For the practical and linguistical identification of the same in this period, see Reilly, León-Castilla under Urraca, pp. 253-59.

39. AC Burgos, vol. 48, no. 1. Pub. Férotin, ed., Recueil des chartes de Silos, pp. 41-43; Fidel Fita, "Texto correcto del concilio de Husillos," BRAH 51(1907): 410-13; and Serrano, Obispado de Burgos 3:76-78, all of whom regarded it as an original. It is not, but there is a deliberate attempt at archaism in the script. The boundaries, which are clearly set out to the advantage of Burgos and extend the latter's claims as far south as Sepúlveda, should not be accepted as those of the council, lot the list of those confirming is probably reliable.

40. MC 20:16-18.

41. See note 29.

42. Paternus is cited as late as March 1, 1088, as bishop but that document is suspect. PMH, Diplomata, pp. 419-20.

43. Apr. 20, 1088. Pub. José María Lacarra, ed., Colección diplomática de Irache, pp. 89-90, which lists the see of Calahorra as still vacant. Some minor church business concerning Nájera also done at Husillos is mentioned in Serrano, ed., Cartulario de San Millán, pp. 226-27. The migration of this episcopal see from Nájera to Calahorra, and sometimes to the monastery of Albelda, vastly confuses the record.

44. Mañueco Villalobos and Zurita Nieto, eds., Documentos de Valladolid 1:7-10.

45. AHN, Clero, Carpeta 885, no. 5; and Códices, 987B, fol. 11v, both dated to 1084. AHN, Códices, 996B, ff. 58r and again on 69r-v, both dated to 1088. The confirmation of Alvaro García as alférez makes the latter reading preferable. Fita, "Monasterio toledano de San Servando," pp. 286-90, published it and changed the date to 1089, but that is 100 late for the alférez. In the eighteenth-century Andrés Merino, Escuela paleográfica (Madrid, 1780), pp. 39-40, published it with the date I prefer.

46. Luciano Serrano, El Real Monasterio de Santo Domingo de Silos (Burgos, 1926), p. 30.

47. AHN, Códices, 989B, fol. 117v; pub. Escalona, Historia de Sahagún, p. 486. A private charter confirmed by many court figures.

48. AHN, Códices, 1.196, ff. 12v-13r; also 363B, fol. 101r-v; AG Simancas, Sección de Gracia y Justicia, legajo 1,672, Lugo, no. 11; pub. Es 40:422-25.

49. The conflation was a bumbling job and the evidence abounds. See the more modest Alfonsine charters of July 21, 1088. AHN, Códices, 1.196, ff. 13r-14v, ff. 65v-66r and 66r-67r, and 363B, ff. 13r-14V, ff. 102-104v; pub. ES 40:182-83, and Estefanía, "Memorias," pp. 189-92. June 17, 1089. AHN, Códices, 1.196, ff. 64v-65r, pub. Estefanía, "Memorias," pp. 205-208. A supposed cession of Count Rodrigo Ovéquez of all possessions to the church of Lugo. July 25, 1089. AHN, Códices, 1.196B, fol. 14v, 363B, ff. 104v-105r; pub. López Peláez, El señorio temporal de los obispos de Lugo 2:115-18. Another limited donation by Alfonso VI. Finally, Feb. 5, 1086. AHN, Códices, 1.196, fol. 14v-15r; copy in Acad. Hist., Catedrales de España, 9-25-1-C-3, ff. 42r-44r. This earlier concession by Infanta Elvira is also relevant.

50. July 21, 1088. See note 49. This relatively trustworthy charter of Alfonso to Lugo was confirmed by Pedro. Given subsequent events, his confirmation is not likely to have been supplied by a forger.

51. Manuel Rubén García Alvarez, "Catálogo de documentos reales de la Alta Edad Media referentes a Galicia (714-1109)," Compostellanum 11 (1966): 318.

52. Ubieto Arteta, ed., Crónica Najerense, p. 117, dates the campaign to 1088. Menéndez Pidal dated the campaign to 1089, but see the extended discussion of the chronology in Huici Miranda, Grandes batallas de la Reconquista, pp. 85-94, and Historia musulmana de Valencia 2:15-29.

53. Huici Miranda. Historia musulmana de Valencia 2:8-11.

54. Huici Miranda, Grandes batallas de la Reconquista, pp. 94-95.

55. The states of mind of the besiegers are vividly described by the king of Granada, who was personally present. Abd Allah, Siglo XI en 1^a persona.

56. Huici Miranda, Historia musulmana de Valencia 2:11.

57. Sánchez-Albornoz, Despoblación y repoblación del valle del Duero, pp. 384-85.

58. AHN, Clero, Carpeta 885, no. 23. An original private document confirmed by Alfonso and many of his court.

59. "Historia Roderici," in Menéndez Pidal, España del Cid 2:935.

60. Serrano, ed., Cartulario de San Millán, pp. 275-76. It purports to be a charter of exemption. There is no notary given, and the diplomatic conforms to that of none of the notaries known to be of Alfonso's chancery at that time. In the dating formula it is said to have been given after the return from Aledo but in "campo de Conchiella in Monte Aragon," which would put the king near Huesca! It is also dated to Nov. 25, 1089. In addition, there is the most peculiar confirmation "Infante Garsea, prolis Sancio Naierense, in Toleto sedcntem, confirmans." Changing the date of the siege of Aledo to 1089 to conform to this strange document does not resolve the problems but only compounds them. That a forger simply invented all this incidental information is unlikely, but one wonders what his model or models indeed said. Also in 1088, there is a charter of Infanta Elvira to the monastery of Oña in the same general area dated simply by year. BN, Manuscritos, 720, fol. 196r. Pub. Isabel Oceja Gonzalo, ed., Documentación del monasterio de San Salvador de Oña, 1032-1284 (Burgos, 1983), pp. 23-24, from another copy in the Real Academia.

61. AD León, Gradefes, no. 8.

62. For a scrutiny of this literature and its shortcomings as history, again see Huici Miranda, Historia musulmana de Valencia 2:12-29.

63. Luis García de Valdeavellano, Historia de las instituciones españolas (Madrid, 1968), pp. 613-28, remains a useful survey of Spanish military terminology, practice, and law in the middle ages.

64. We may take the assertion of the "Historia Roderici," in Menéndez Pidal, España de Cid 2:932, that the Cid led seven thousand men at arms as the usual medieval hyperbole but undoubtedly it reflects the author's assessment of his forces as "royal" in size.

65. See note 13. The fact that this diploma was subsequently interpolated is not germane here for such language would have become less liable to have been used as the progress of the Gregorian reform increased the sophistication of everyone's language.

66. See note 39. Again, that the document has been tampered with does not affect the credibility of these formulae. See also Engels, "Papsttum, Reconquista und spanisches Landeskonzil im Hochmittelalter," pp. 44-45.

67. ES 20:254.

68. As Bernard's superior Hugh had been consulted and had approved enthusiastically the former's promotion to the see of Toledo. Marius Férotin, "Une lettre inédite de S. Hughes abbé de Cluny à Bernard d'Agen, archevêque de Toledo," Bibliothéque de l'École des Chartes 61 (1900): 339-345, and 63 (1903): 682-86.

69. Demerrio Mansilla, ed., La documentación pontificia hasta Inocencio III, 965-1216 (Rome, 1955), pp. 39-45, for all of these four letters. The presence of Bernard in Rome was also noted in the Liber Pontificalis. See Rivera Recio, Iglesia de Toledo, p. 136, n. 37.

70. These last two letters are described by P. Ewald, ed., "Die Papstbriefe der Brittiscben Sammlung," NA 5(1880): 357-58. Louis Duebesne, ed., Le Liber Pontificalis, vol. 2 (1886; Paris, 1955), p. 293.

71. Guérard, ed., Cartulaire de l'Abbaye de Saint- Victor de Marseille 2:205-207. Cited, Mansilla, ed., Documentación pontificia, pp. 45-46.

72. See note 69.