[210] During the winter or spring of 1089 Alfonso VI despatched his most trusted agent, Count Pedro Ansúrez, to the court of the king of Granada. The point of the journey was to secure the resumption of the payment of the annual tribute, parias, in the aftermath of the conspicuous Muslim failure at Aledo. The Granadan king, Abd Allah, tells us that he finally agreed to pay 30,000 dinars, representing the three years of arrears since Zalaca at 10,000 dinars per annum. The taifa king of Zaragoza, al-Mustain, also resumed the payment of tribute as did the other Muslim rulers of the east, according to the same source.⁽¹⁾

Despite their paramount importance for Alfonso, the parias represented but a part of the king's objectives. Abd Allah reveals that the Leonese monarch pressed him to make formal claims to lands currently held by al-Mutamid of Sevilla, which Alfonso would recognize of course. The Granadan monarch piously rejected the prospect of cooperating with the infidel against others of the faithful. His whole course of action, he tells us, was taken under duress because the failure of the Murâbit Yusuf to leave an army in Spain after the siege of Aledo deprived him of an alternative.⁽²⁾ Nevertheless, in 1089 Alfonso could still hope at least to neutralize Muslim Spain in the continuing struggle with the Murâbits. Even at Zalaca, Muslim Zaragoza, Lérida, and Valencia had remained neutral and Badajoz may have had to have been actively coerced into participating. The failure at Aledo had subsequently tarnished the military reputation of Yusuf ibn-Tashufin, and the taifa kings were all suspicious of his ultimate intentions. Better perhaps tribute to a suzerain in León than deposition and prison at the hands of the Murâbit. In fact, during the next five years Alfonso would consistently adopt the role of defender of the independence of the Spanish taifas.

Yet the Leonese monarch had to have the parias even if their collection weakened the position of his taifa allies with their own subjects and forced them to make lame explanations to Yusuf.⁽³⁾ The parias were, we [211] surmise, essential to that condition of military superiority in the peninsula which León-Castilla had already assumed. That kingdom was, by 1086, a giant by comparison with the other political units, Christian or Muslim. Nonetheless, composed for the most part of the two mesetas of Castilla la Vieja and Castilla la Nueva, its disposable wealth from agriculture in all likelihood was less than that from any one of the single taifas of Zaragoza, Valencia, or Sevilla. Of a surplus from trade it hardly yet knew.

Under these circumstances even León-Castilla's burgeoning population could be as much problem as potential. Unless the king could mobilize large armies on a sufficiently regular basis to ensure the safety of the newly reconquered or reoccupied territories the initiative would pass from his hands. The problem was not simply that a resurgent Islam might reclaim Toledo and the valley of the Tajo and render even the new settlements of the trans-Duero precarious. It was also that other Christian hands might here and there assume the leadership in the struggle. The geographic character of the peninsula enhanced such a possibility. Already the kings of Aragón and the counts of Barcelona were exercising such a role though on a diminutive scale. But in the taifa of Valencia Rodrigo Díaz de Vivar had found a truly royal arena in which to exercise his capabilities for leadership. In Valencia that was to be a temporary phenomenon, but in Portugal a little later such a local initiative was to produce an enduring feature of the Iberian political world.

Nor were the parias essential merely to royal preeminence in the Reconquista of the frontiers. One may suspect that they had, in some degree, become first an additional instrument and then an important support of royal policy in every regard. It is not farfetched, I think, to take the case of the royal relationship with the great European power which was Cluny as paradigmatic.

At some point during the last ten years of his reign Fernando I had assumed an obligation of paying an annual cens of 1,000 gold dinars a year to that great monastic house.⁽⁴⁾ Then, in 1077 in the midst of a dispute with Pope Gregory VII, Alfonso VI had agreed to double the annual payment. None of the documents concerned mention an overt quid pro quo of course. Prayers were offered regularly at Cluny for the souls of King Fernando and Queen Sancha. Nevertheless the continuing, mutually rewarding cooperation of monarch and monastery is [212] striking. But the refusal of the taifa kings to pay the parias from 1085 created a financial strain which seems to have been reflected in the cessation of the annual payments to Cluny. Beset by a major reduction in income and increasingly heavy demands for military expenses, Alfonso had doubtless economized everywhere.

His Cluniac allies, however, were reluctant to lose permanently so handsome a portion of their revenues, and Abbot Hugh protested sharply to the king. In the aftermath of the Council of Husillos, the latter could not afford the opposition or even the neutrality of the great Burgundian house. He needed and received Hugh's support at Rome for the reconstitution of the Toledan archbishopric and primacy and for the confirmation of Bernard of Sauvetot as its archbishop. Even more acutely, after October 1088 he needed his good offices for the painful unraveling of the affair of the bishopric of Santiago de Compostela. Certainly the matter of the cens to Cluny had been discussed before and Alfonso VI had promised some resolution of the problem. Now he must act, and in 1089 he made a one-time payment of 10,000 dinars to Cluny, which amount suggests the five-year arrears from 1085 to 1090 at a rate of 2,000 a year.⁽⁵⁾ Such pressing expenses make more explicable [213] the king's necessity to squeeze the Muslim taifas for tribute even though that pressure in some degree countered his interest in detaching them from the Murâbits.

If we can extrapolate from the crown's well-documented expenses in the case of Cluny to those humbler, domestic needs that more rarely found a chronicler, we can partially understand as well the way in which the drying up of the parias would have loosened the bonds of regard and self-interest that made Alfonso effective master of the realm. In the western medieval tradition of kingship the monarch was, among other things, the great giver of gifts. The royal largesse was one of the tools of leadership, to phrase it crassly.

At the highest levels we can see this at work in León-Castilla where the rich endowment of the church-shrine of San Isidoro in León follows directly on the heels of Fernando's triumphant campaign of 1062, which institutionalized the parias of Andalucía. Again, the successful campaign of Alfonso VI in 1074, which forced the taifa of Granada to render tribute, was quickly followed by a royal visit to Santiago de Compostela and the endowment of the shrine-church there. But to understand the extent of the royal dependence on the regular payment of parias we must be able to imagine how this type of resource, stable for more than two decades between 1062 and 1085, must have inflated the scale of royal generosity to humbler churches, monasteries, familiars, magnates, counts, and castellans. The expectations so created could not be disappointed without threat of a reaction prejudicial to the best interests of the crown and dynasty.

The exaction of parias, and a treaty with the kingdom of Granada, was followed by a campaign to secure the same arrangements with al-Mutamid of Sevilla, the most energetic and powerful of the Andalucians. Such efforts failed of their purpose and the Sevillan ruler appealed yet again to Yusuf ibn-Tashufin for assistance. During the same summer, Abd Allah was beset with revolts in his own territories, which failed but had the cumulative effect of driving him toward closer cooperation with Alfonso VI.⁽⁶⁾ The Christian sources and documents tell nothing of all this, and even the Muslim sources fail to mention the attitude of the enigmatic al-Muttawakil of Badajoz.

It is entirely possible that the Leonese monarch was not personally involved in any Andalucian campaign in 1089, leaving that task to subordinates such as Count Pedro Ansúrez. The documentary record is [214] quite blank for the first half of 1089, and when we encounter the king in June or July he seems to be in Lugo in Galicia.⁽⁷⁾ By August 25, 1089, the king and the court have returned to the vicinity of León, where Alfonso exchanged properties with María Peláez. Among others, Archbishop Bernard of Toledo confirmed the royal charter, which I would take to be an original.⁽⁸⁾ His confirmation, however, makes it strictly impossible that he should have been in Braga for the dedication of the cathedral there on August 28, 1089, as has usually been believed.⁽⁹⁾

So far as can be established the royal court remained in the general area of León for the remainder of the year. On September 24, 1089, in the course of adjudicating a dispute between his sister, Urraca, and the bishop of León, Alfonso made some far-reaching decisions on land tenure in the realm.⁽¹⁰⁾ On November 9, he granted extensive lands to the church of Toledo, but the act probably took place at León or Sahagún.⁽¹¹⁾ Finally, on November 24, a private judicial action was executed, probably at Sahagún, in the presence of the court.⁽¹²⁾

While the king thus stubbornly pursued his strategic aims in the Islamic world of the south in 1089, he was no less busy in other quarters. When the archbishop of Toledo returned from Rome bearing his pallium in the late fall of 1088 he also brought with him the papal letters condemning the deposition of Bishop Diego Peláez of Compostela. The royal charter of December 27, 1088, was the last document Pedro of Cardeña confirmed as bishop of that see.⁽¹³⁾ Although Alfonso VI would comply in some degree with papal demands, the mode of that compliance must be negotiated.

Negotiations with Urban II were not be be carried on in a vacuum as it turned out. If the new pope understood little of peninsular realities in 1088 he was quickly to become more acquainted with them and would [215] draw strength from that knowledge. The early summer of 1089 found the emissaries of Sancho Ramírez of Aragón at the papal court, and Urban's letter of July 1, 1089, initiated new and closer relations between that kingdom and Rome.⁽¹⁴⁾ While it would be a mistake, as we shall see, to cast the Aragonese monarch as an opponent of Alfonso VI, the former was, in the subtleties of diplomacy, at least a minor rival and therefore a potential counter for Urban in negotiations with the latter.

The continuing advances of Aragón in the northeast of the peninsula, although modest, contributed to the concerns of the Leonese king. After suffering a setback to his ambitions in 1088 when Centulle IV, viscount of Béarn, was murdered within his territories, Sancho Ramírez not only maintained the alliance with the south French but on June 24, 1089, captured the fortress of Monzón.⁽¹⁵⁾ That feat permanently breached the forward defenses of Zaragoza in the northeast between Barbastro and Lérida.

Hand in hand with this stubborn offensive against Zaragoza went a policy ambitious in other ways as well. Count Ermengol IV of Urgel provided in his will that should he have no son or brothers, his lands would pass to Pedro, eldest son of the Aragonese monarch. Apparently, Sancho had been married to a daughter of the count.⁽¹⁶⁾ In 1090 he was also to repopulate Estella in order to manipulate the pilgrim traffic to Santiago de Compostela to his advantage.⁽¹⁷⁾ Nevertheless his relations with León-Castilla remained good. In the same year Archbishop Bernard of Toledo would be found collaborating with the abbot of Irache in providing for some mills in Puente la Reina where the frontiers of Aragón and the Leonese realm ran together.⁽¹⁸⁾ Most important of all, during the summer of 1090 Sancho Ramírez would assist Alfonso VI to force the raising of the siege Toledo by the Murâbits.⁽¹⁹⁾ But of course the more the two monarchs collaborated the more difficult it became to deny the Aragonese something like equality of status, and hence the more real became another set of limitations on the alternatives open to León-Castilla.

Further to complicate Alfonso VI's relationship with the papacy was [216] Urban II's decision to reestablish the archbishopric of Tarragona, taken in the summer of 1089.⁽²⁰⁾ True, Catalonia was even farther away than Aragón, and the reconquest of the city of Tarragona was to be long delayed. Even so the papal decision placed yet another counter on the diplomatic board and the Leonese king would be constrained to watch it as well.

All of this negotiation we can observe only most imperfectly. By the middle of 1089 Pope Urban had appointed a new legate for Spain, Cardinal Rainier, a former Cluniac who was destined to become pope as Paschal II ten years later. The cardinal was charged with the matter of Tarragona as well as that of Santiago de Compostela.⁽²¹⁾ Alfonso was doubtless informed of this action, and he appears to have visited Santiago de Compostela on January 28, 1090, to prepare the ground for the forthcoming decision.⁽²²⁾

Indeed, from this document it appears that he had already taken important steps. The future bishop-archbishop of the see, Diego Gelmírez, confirms as "maiorinus et dominator compostelle honoris." The appointment of a secular vicar to manage the temporalities of a vacant see was a regular royal practice, especially when a long vacancy was expected or planned. The device could be a lucrative one for the crown, but it could also be destructive. In this instance, years afterward the conduct of the royal vicars would be recalled with great bitterness in the "Historia Compostelana."⁽²³⁾ At any rate the very real interests of the crown would be protected for the political, military, and economic prerogatives of the bishop were much too substantial to be left unattended [217] or in unreliable hands. We cannot be positive that the town of Compostela was entirely under the control of its bishop at this time, but an episcopal señorio in the towns was common in Galicia.⁽²⁴⁾

Provision had been made, as well, for the closer governing of the entire province. Count Raymond of Burgundy confirmed the document as "imperans Gallicia." As we have already seen it is marginally possible that the French noble held that dignity earlier, but the documents are few and scattered and there are no originals among them.⁽²⁵⁾ From 1090 the notices of the count became regular although the first original instrument known to me currently is a private document of February 27, 1091.⁽²⁶⁾ The old realm of García Fernández had now become the province of Count Raymond, who would rule it with viceregal powers until his death better than seventeen years later.

In fitting preparation for this high responsibility, the Burgundian had either been married or betrothed to the king's daughter, Urraca, probably at León during the preceding Christmas season. The royal infanta confirmed the document as "comitis domini Ramundi maritata." At that time Urraca would have been between eight and nine years old, and apparently the count's sepulchral inscription gave her age as eight at the time of her marriage.⁽²⁷⁾ She would not then have reached puberty, but the public act need not have implied marriage in the sense of either sacramental service or physical consummation. The medieval church and society regard the declaration of intent embodied in betrothal as actually effecting the marriage. In fact the young princess may have continued for a time yet in the actual care of her guardian, Count Pedro Ansúrez, as surety for the good behavior of her husband.⁽²⁸⁾

By February 14, 1090, it appears that the court had returned to the León-Sahagún area.⁽²⁹⁾ At León during the latter part of March a great council would be held in the presence of the papal legate, Cardinal Ranier. No acts or canons have survived from the Council of León so we are forced to piece together its participants and substance from scattered testimony. It was already in session on March 22, 1090, when the [218] king's brother, García, died in prison and his body was brought to León and buried with great ceremony.⁽³⁰⁾

The "Historia Compostelana" reports only that in the council Pedro of Cardeña was deposed and that Santiago de Compostela continued to be governed by lay administrators thereafter.⁽³¹⁾ The former abbot of Cardeña now drops from sight, but the former bishop, Diego Peláez, long continued to struggle for what he considered his rightful position. Although his case had been revoked to Rome he apparently traveled no farther than Aragón after his release by Alfonso VI and there eventually found tacit support from the Aragonese crown.⁽³²⁾

Lucas de Túy, writing almost a century and a half later, gives a garbled report that the council legislated that the "ecclesiastica officia" in Spain should follow the rule of San Isidoro of Sevilla and that ecclesiastical books should be written in the caroline rather than the visigothic script.⁽³³⁾ What he meant by the first is impossible to say and the second was a slow process, already in train, which would be completed only gradually over the next forty years. It is doubtful that the council would have regarded the latter as worth its attention.⁽³⁴⁾

It is probable that the council discussed the organization and structure of the Iberian church. After all, Toledo had just become the first archbishopric restored since the Muslim conquest. In addition the legate himself bore among his commissions that of reestablishing the archbishopric of Tarragona. Finally, among those sees of the Christian north liberated from the Muslim yoke one, Braga, had had a brilliant history as an archiepiscopate in classical antiquity. Carl Erdmann believed that Bishop Pedro of Braga attended the council and made an attempt [219] to reclaim its ancient rights.⁽³⁵⁾ The atmosphere would have been cool, at best, to an argument based on antiquity as a model. Sees as important as Oviedo, León, and Burgos had no classical warrant. Lugo had present claims to the ancient dignity of Braga itself, and the primatial dignity accorded to Bernard of Toledo in 1088 had included jurisdiction over all sees otherwise without a metropolitan. Alfonso would hardly, given his own problems at the time, have been willing to intervene decisively on the part of far away Braga even had he been so inclined. In the long run, nevertheless, that omission was to be exceedingly costly.

The meeting of the council would also have been marked by the usual and expected generosity of the king. It may still have been in session when Alfonso confirmed the possessions of the see of Palencia on March 31, 1090. That action probably can be accepted even if the instrument by which it reaches us has been interpolated.⁽³⁶⁾ In it Alfonso VI stated that he took this action with the counsel, among others, of his son-in-law Count Raymond, and undoubtedly a major event of the council was the public recognition of the high position in the realm now accorded to the Burgundian. Given the death of his brother, García, whom Alfonso had at some point considered as his possible successor,⁽³⁷⁾ and the continuing lack of a male heir, such recognition came close to designating Urraca and Raymond as his successors although it probably stopped short of an explicit designation. Still, everyone there would have understood the implications.

In the context of such an expected devolution of the realm, Alfonso probably also consulted those present at the council on his decision to reaffirm, perpetually and solemnly, the annual cens of 2,000 dinars to Cluny. On Easter Sunday, April 21, 1090, he was to meet Abbot Hugh in Burgos and surrender formally to the latter such a written promise.⁽³⁸⁾ [220] That meeting had been concerted carefully and was designed to bolster the prestige of the king in his realm and to lay the continuing problem of the succession. Yet the price was substantial indeed.

The next notice of the whereabouts of the royal court puts it at Sahagún on July 8, 1090.⁽³⁹⁾ From that customary royal residence the Leonese monarch was to organize and coordinate a relief expedition for the city of Toledo. The Murâbit emir, Yusuf ibn-Tashufin, had decided on yet a third Spanish expedition in the spring of 1090. In June he disembarked at Algeciras and marched northeast to Córdoba where he had arrived by July.⁽⁴⁰⁾ From that city he would march north to lay siege to Toledo, the key to the control of the valley of the Tajo and the most direct line of communication between Muslim Andalucía and Muslim Zaragoza.

Then as now the only practicable route for an army lay along the line of the present national highway IV. That is east from Córdoba to Bailén and then north through the pass of Despeñaperros and, once out on the central meseta, an approach toward Toledo from the southeast. In road miles that is roughly 360 kilometers, and at the sustained rate of twenty-five kilometers a day it is unlikely that the Murâbit army could have reached Toledo much before July 20.

Toward this threat Alfonso VI seems to have adopted a strategy of attrition. The army of Yusuf started with a grueling march in the hottest part of summer through country that allowed of their easy harassment by small, mobile detachments. The early warning of their approach allowed for a complete victualing and garrisoning of what was an almost impregnable position in any event. The Christian leader evidently preferred to let the hardship of the march, the ordinary diseases associated with camp life, and the resistance of the Toledo garrison reduce the number of enemy effectives and sap their morale and only to risk a major battle on the most favorable terms. That is the best explanation for the fact that it is only on August 11, 1090, that we find him at Dueñas, south of Palencia, on his way toward Toledo.⁽⁴¹⁾

From Dueñas the Leonese army could easily choose one of two [221] passes through the Guadarrama chain. The pass of Navacerrada just south of Segovia was short but high and led down to Madrid. South of Avila the long but low pass of Arrebatacapas led down to Maqueda. Presumably both of these fortresses were still in Christian hands for we hear of no towns captured by the Murâbit forces. I believe that Alfonso took the route through Avila for Maqueda was better placed to serve as a base of operations from which to bedevil an army besieging Toledo and offered an easier line of retreat if that should prove necessary. But if Alfonso was at Dueñas on August 11, he could hardly have covered the 250 kilometers south to Maqueda before August 22.

That meant that by the time he and his ally, Sancho Ramírez of Aragón, arrived at Maqueda the Murâbit army before Toledo was already breaking camp. We have no details as to the problems of the siege, but they were apparently sufficient so that, at news of the approach of a relieving force, Yusuf decided to retire. In marked difference from his two earlier campaigns, Yusuf seems to have had no assistance whatsoever from the taifa rulers of Andalucía. He could not be sure what his prospects would be then for an orderly retreat if he should be defeated in the field. Rather than take such a risk he determined to act immediately to see that he never again had to face such a prospect.

By September 8, 1090, Yusuf was at the gates of Granada. To have traversed the roughly 400 kilometers between Toledo and Abd Allah's capital, he must have given up the siege by August 25 at the latest. Now the king of Granada became the first of the taifa rulers to be deposed and replaced by a Murâbit governor. Shortly thereafter Abd Allah's brother al-Tamin, king of Malaga, was also deposed by Yusuf. When the emir of the Murâbits left for Morocco in November he left behind a sizeable force under his cousin, Sir ibn-Abu Bakr, with instructions to continue the process of consolidating Muslim Andalucía under his rule.⁽⁴²⁾

More than the siege of Toledo itself, what speaks to the real weakness of Alfonso VI at this time is his inability to follow up the success of his plans there with a counter invasion of Andalucía. At this critical point he was unable to intervene to save his ally in Granada, and he is to be found rather back at Sahagun on September 7 and September 16, 1090, attending to internal concerns of the realm. These matters were serious for they touched a quarrel between two great magnates of the realm and companions of the king in his youth, Counts Pedro Ansúrez and Martín Alfónsez, and the powerful abbot of the royal monastery of [222] Sahagún.⁽⁴³⁾ The crown could not afford such friction at a time when the Murâbit offensive should engross all its energies.

For the first time in years there was to be no winter respite in the activities of his enemies. Sir ibn-Abu Bakr brought great energy to the position accorded him by his cousin and promptly launched an offensive against the most powerful of the Andalucian kingdoms, Sevilla. After seizing Tarifa in December to safeguard yet further his communications with Morocco, he then settled down to a siege of Córdoba, strongly defended by a son of al-Mutamid, Fath al-Mamun. The resistance of the city was fierce but by March 15, 1091, it had fallen to the enemy and al-Mutamid's son was dead. From that secure base Sir was able to begin to operate militarily on the southern fringes of the meseta of Castilla la Nueva and secured a forward position there at Calatrava.⁽⁴⁴⁾ These operations must be deduced solely from the Muslim histories for the northern chroniclers took no note of them.

One could not safely say the same of Alfonso VI, but at the very least he took no action personally to intervene in the south. The royal whereabouts are documented fairly well in early 1091, and it is clear that the court was in the vicinity of León or Sahagún continuously from January to April 18.⁽⁴⁵⁾

By mid-April Córdoba had already fallen and Sir was besieging the [223] stronghold of Almodóvar del Río on the road to Sevilla. Exactly when the castle there fell is hard to say, but by May 9, 1091, the Murâbit general had just taken Carmona and was within thirty kilometers of Sevilla itself, to which he promptly laid siege in turn.⁽⁴⁶⁾ Al-Mutamid in this extremity appealed for help to the very enemy against whom he had so often invited the Murâbit into the peninsula, but without avail.⁽⁴⁷⁾ Sometime during the summer a Christian relief force penetrated as far as Almodóvar but was defeated there.⁽⁴⁸⁾ Even then, it may be that, as so often would be the case later on, Alvar Fáñez was leading a force raised locally rather than a royal army raised more generally in the realm. In any event, Sevilla would fall to Sir ibn-Abu Bakr on September 9, 1091. the collapse of the greatest of the taifas was promptly followed by the fall of the smaller kingdoms of the southeast; Jaén, Almería, Denia, and finally Murcia late in the year.⁽⁴⁹⁾ Of Muslim Andalucía, only the kingdom of Badajoz retained its independence at the outset of 1092.

Still, in the north, Alfonso VI had barely stirred. On July 19, 1091, he was at León.⁽⁵⁰⁾ By August 5 he was at Castrojeriz with his court, perhaps on his way to Burgos.⁽⁵¹⁾ It may have been on this trip that the king granted the charter to the Riojan monastery of San Millán dated only by the year.⁽⁵²⁾ But by September 24 the court had returned to Sahagún and stayed there through November 10, 1091, which is as far as we can trace its movements for the year.⁽⁵³⁾

At this remove it is difficult to account for the seeming inactivity of the Leonese monarch in the crucial years of 1090 and 1091. One possibility [224] is that the loss of the parias of Andalucía which were certainly not paid after the advent of Yusuf in the peninsula in 1090, and the prospect of much fighting but little booty there drastically inhibited the king's ability to recruit an army of the requisite size. But there are indications as well that problems internal to León-Castilla may have combined to reinforce that effect.

There is a cryptic statement in Lucas of Túy, "Eo tempore Rex Adefonsus offendit graviter Comitem Castellanum Garsiam de Cabrera et causa placendi ipsum dedit ei Geloyram sororem suam in uxorem, et pacificavit totum regnum quod in seditionem vertebatur."⁽⁵⁴⁾ There is no further indication of time in our source except that the passage occurs immediately after the author's description of the Council of León and is in turn followed by the account of the marriage or betrothal of Infanta Urraca to Count Raymond. All of this suggests to me that the solution to the problem of Alfonso's lack of a male heir, in train since 1087 by the elevation of the Burgundian count, was resented much more widely than merely in Galicia where it prompted the revolt of that year. Marriage to the king's eldest daughter was, in the circumstances, a great prize, and the magnates of the realm could not have but watched its bestowal on Raymond with anything but rage and disappointment.

Unfortunately it is impossible to identify securely this magnate whose pacification was so important. Lucas tells us that he was a Castilian and a count. But countship was unusual in Castilla, and the only Count García who appears in the contemporary documents is García Ordóñez of Nájera whom we can rule out immediately.⁽⁵⁵⁾ If by "count" Lucas meant simply to indicate "magnate," then it seems to me the documents dictate only one possible choice, García Alvarez, son of Alvar Diaz of Oca. This scion of one of the great magnate families of Castilla will become alférez of Alfonso in 1100 and remain in that post until 1107. If this identification is accepted, then at Uclés he died defending the royal infante, Sancho Alfónsez, whose guardian he was.⁽⁵⁶⁾

This hypothesis that the kingdom was in some turmoil in 1090-91 is also suggested by ecclesiastical events. As we have seen, the Council of León did not settle the affairs of Santiago de Compostela and that very considerable plum remained vacant. Further, Bishop Diego Peláez seems to have been able to defy the command of the pope to proceed to [225] Rome for judgment, and the king was unable to enforce the papal directive from which he could reasonably have expected to profit. But far more striking was the decision of Bishop Pedro of Braga to ignore the king entirely and to seek the restoration of the archiepiscopal status of his see from the antipope, Clement III, in 1091.⁽⁵⁷⁾ This action was to lead to the bishop's deposition in fairly short order, but for him even to have imagined that he could scout the royal wishes and policy in such a manner can testify only to a period in which the debility of the crown's power was marked indeed.

Some further conception of the way in which the polities of the realm were evolving can be gained by close attention to the personnel of the royal curia between the end of 1086 and the beginning of 1092. Among the clerical members of the court the most outstanding novelty is the vault into prominence of the newly installed archbishop of Toledo. That cadre of bishops who had long been prominent at court continued to be so. The prelates of León, Palencia, Burgos, and Astorga all confirmed two-thirds or more of the eighteen known and reliable royal documents of this period. But now the archbishop of Toledo did the same.

Bernard of Sauvetot arrived in Spain, as we have seen, during the struggle between Alfonso VI and Gregory VII and became abbot of the favorite royal monastery of Sahagún in 1080. Although by no means prominent in royal diplomas he must have made himself most useful to the king for he was elevated to Toledo at the royal wish in 1086. We know little of his earlier life. Jiménez de Rada was following a now lost vita of Bernard preserved at Toledo when he informs us that his subject was studious in youth but took arms as his profession until an illness moved him to enter the monastery of St. Orens in Auch. From Auch he was summoned to Cluny by St. Hugh.⁽⁵⁸⁾ Noreen Hunt dated Bernard's entry into St. Orens about 1070 and says that he returned to St. Orens as prior about 1078 after having risen to be chamberlain at Cluny.⁽⁵⁹⁾

From the time of his advancement to Toledo he became the most trusted of royal councilors of Alfonso VI; a role he retained under the king's successor, Urraca, until his death in 1125. Because of the king's dependence on him and the fact that his archiepiscopal city remained [226] until late in the twelfth century not the center of the realm but a particularly exposed outpost of it, Bernard spent virtually all his life at court and was buried at Sahagún rather than Toledo.⁽⁶⁰⁾

The position of the Frenchman both at court and in the church was galling to the bishops of the meseta in particular, all of whom, with the probable exception of Astorga, cherished some such ambitions for themselves and for their sees. Quite possibly the new prominence of Toledo and Bernard had a part in provoking the revolt in Galicia in which the bishop of Santiago de Compostela had had such a prominent role in 1088. Certainly the new position of Toledo reacted to produce the spectacular defection of Braga in 1091.

Other changes in the makeup of the Alfonsine curia are less striking. A new majordomo appears in 1086 and will hold that office for almost ten years.⁽⁶¹⁾ This Ermegildo Rodríguez is peculiar in that he never appears in the documents of the time before or after his term as majordomo with one exception. His given name suggests the possibility that he was drawn from the region of northern Portugal where it was fairly common, whereas elsewhere it was rare indeed. He may be that "Armigii ruderigu" who confirmed the donation of the Portuguese noble, García Múñoz, to King García of Galicia on March 24, 1066. A recent historian has suggested he was Leonese but on slender indications.⁽⁶²⁾ In any event, Ermegildo's obscure origins and long tenure may indicate that the king either could not get or did not want a great magnate in this dignity.

In contrast to the stability of the office of majordomo, there was much movement in that of alférez. At the very beginning of this period, Rodrigo Ordóñez continued to hold it as he had for the past five years.⁽⁶³⁾ He was then replaced for something like a year by Alvar García, who is known only from the documents in which he held this dignity.⁽⁶⁴⁾ His name would suggest that he was a Castilian like Rodrigo Ordóñez. From the latter part of 1088 till the end of 1091 the post was held by Pedro González of the rising Lara family of Castilla.⁽⁶⁵⁾

Among the secular magnates who frequented the royal court the most prominent at this time was Count García Ordóñez, who confirmed [227] eleven of the eighteen diplomas. His brother, Rodrigo, confirmed five but was never, after leaving the post of alférez, elevated to a countship as would otherwise have been the usual practice. Perhaps the position of Count García in the Rioja was commanding enough for the royal taste. The great Leonese count, Pedro Ansúrez, confirmed eight court documents but his fellow Leonese and early companion of Alfonso VI, Count Martín Alfónsez, confirmed but four. The latter was to die in 1093 and it may be that failing health already kept him from traveling with the court. Newly prominent in the royal entourage of this period was Count Froila Díaz, who confirmed seven documents. There has been considerable confusion over his family, with Menéndez Pidal insisting that he was the brother-in-law of the Cid. Although his origins are in Asturias, more recent historians do not accept that relationship.⁽⁶⁶⁾ Froila seems to have been a count from the time of his first appearance at court on December 18, 1086.⁽⁶⁷⁾ Among the thirteen other counts of this period of whom we have notice, no one of them confirms as many as one-third of the eighteen known royal documents. The same phenomenon is noticeable in the case of the magnates of Castilla who did not ordinarily bear the comital title. Only Alvar Diaz of Oca confirmed as many as seven diplomas, and the head of the Lara family, Gonzalo Núñez, confirmed but five.

So it would seem that there is at least the possible reflection here of a loss of attraction by the royal court. For a time, the ability of the crown to make its influence felt everywhere had slackened and so attendance on the king had rather less point.

Certainly Alfonso VI's ability to control events in the east of the peninsula was at a new low. Sancho Ramírez of Aragón had a virtually free hand after 1087. He used it to improve on his conquest of Monzón in 1089 by capturing Estadilla, which protected the former's supply line, in 1091.⁽⁶⁸⁾ This campaign of strangulation against Zaragoza was also advanced dramatically in 1091 when the Aragonese monarch erected the fortress of El Castellar just twenty kilometers northwest of Zaragoza on the Ebro and athwart its communications with Tudela, the second city of the taifa.⁽⁶⁹⁾

The only real check on his ambitions came not from Alfonso but rather from the Cid, who had become the real arbiter of events in the [228] east of the peninsula from his position as effective suzerain of al-Qadir and of Valencia. Though none of the principalities of the east could be considered in isolation from its neighbors, the chief concern of Rodrigo in 1089 and early 1090 was al-Mundir of Lérida and Tortosa, whose lands directly adjoined those of Valencia. He was harrying this enemy in the territories of Tortosa when al-Mundir died in early 1090.⁽⁷⁰⁾ At this juncture Count Berenguer Ramón II of Barcelona, who exercised a protectorate over these lands, was obliged to intervene. Qn this occasion he found himself in league with al-Mustain of Zaragoza, who had never abandoned his claims to Lérida and Tortosa. The allies encountered the Cid at Tévar west of Tortosa in May or June of 1090 and were resoundingly defeated. The count of Barcelona was captured and gave up his pretensions over the former lands of al-Mundir in the course of securing his freedom.⁽⁷¹⁾

Rodrigo Díaz de Vivar was now the most powerful figure in the Levant and all of the taifa rulers there paid him parias, with the exception of Zaragoza.⁽⁷²⁾ This enormous income was lost to Alvonso VI, whose only intervention was an interview with al-Mustain in the spring of 1090 if we may believe the "Historia Roderici."⁽⁷³⁾ Instead, after his victory at Tévar the Cid advanced into the lands of Zaragoza and spent the summer there. Only a serious illness, which immobilized him at Daroca, provided a check to his ambitions and in the fall he retired to the east.⁽⁷⁴⁾ The fantastic story of a new reconciliation of Alfonso and the Cid in 1091, of a joint expedition against Granada, and of yet a third falling-out between them is recounted in the "Historia Roderici" and accepted by Menéndez Pidal.⁽⁷⁵⁾ It is totally at variance with the known facts. Simply put, the Cid had become an independent ruler.

There may also be a rise of semi-independent powers within the realm itself discernible in the documents of the period. Necessarily one speaks of matters of degree in this respect and the judgments that must be made are very subtle ones, but we have already seen how the events and threats of the years 1080-81 moved Alfonso to create the satrapies of Lop Jiménez in Alava and Vizcaya and of García Ordóñez in Rioja. In a similar manner by 1090 the king had allowed the whole of the distant province of Galicia, an area the size of Scotland, to pass under the control of his new son-in-law, Raymond of Burgundy. This royal legitimization [229] of vigorous local powers of very considerable extent around the fringes of León-Castilla was a practical, governmental device but it was also a testimony to the real weakness of the crown as it attempted to grapple, in straitened circumstances, with the enormously swollen kingdom.

This same sort of development was also occurring along the line of the Duero in what we could easily misunderstand as the interior of the realm. Count Pedro Ansúrez in 1085 controlled the entire seventy-kilometer reach of that rampart from Zamora, through Toro, at least to Tordesillas.⁽⁷⁶⁾ How long he had been so empowered is difficult to say. Probably his control was not as early as 1074 but at least from 1084 and most likely from the time about 1082 when it was becoming clear to Alfonso that he would have to concentrate on the affairs of the faltering taifa at Toledo.⁽⁷⁷⁾ That control continued through 1090 although it was by then limited to Toro.⁽⁷⁸⁾ Increasingly Pedro Ansúrez was to make his foundation at Valladollid, on the Río Pisuerga just north of its confluence with the Duero, the center of his power in the south. On May 21, 1095, he would endow the newly consecrated church there in a great ceremony attended by many of the notables of the realm.⁽⁷⁹⁾

All of these domains were quite separate from the hereditary lands of the family, north on the Río Carrión, about Carrión de los Condes, Saldaña, and Liébana. They represented not only new power along the line of the Duero but a power that extended south into the new settlements of the trans-Duero where a regular structure of local government would be rare even forty years later.⁽⁸⁰⁾ We know that he was responsible for the repopulation of Cuéllar some forty-five kilometers to the southeast of Valladolid, and it is likely that he possessed a sort of vice-regal jurisdiction in the whole eastern half of the trans-Duero from Olmedo and Arévalo to Sepúlveda.

Another Leonese magnate who rose to authority in this area was Count Martín Laíñez, whose original family holdings lay in the north of León about the castle of Aguilar.⁽⁸¹⁾ From 1088 through 1092 he was [230] regularly cited as count in Simancas.⁽⁸²⁾ Martín appears to have been a brother of that Fernando Laíñez who was the alférez of Alfonso VI and was subsequently disgraced for involvement in a conspiracy against the king.⁽⁸³⁾ In his person the family fortunes appear to recover, and before long he would regain the family holdings in the north as well.⁽⁸⁴⁾

It may have been about 1088 that responsibilities were reshuffled at the extreme western end of the line of the Duero as well. In any event by August 16, 1090, the documents reveal Pelayo Vellidez commanding "in coria et in zamora." He would still enjoy that position as late as May 18, 1092.⁽⁸⁵⁾ Here the jurisdiction of the line of the Duero is explicitly linked with an outpost in the trans-Duero at Coria more than two hundred kilometers to the south. Pelayo was a Leonese noble who had been the majordomo of Alfonso in 1084-85 so that it is not surprising to find him subsequently with a territorial jurisdiction, but he never seems to have formally held comital rank.⁽⁸⁶⁾

Finally, at the extreme eastern end of the Duero line, Gonzalo Núñez of Lara enjoyed roughly the same sort of position. In 1089 he was to be found promoting the repopulation of the district about Berlanga.⁽⁸⁷⁾
 

By the years 1090-91, then, the resettlement of that great, internal frontier of the realm which was the trans-Duero had become a task of which the direction and initiative had been delegated by the crown to a series of great magnates who could hardly become but greater in the prosecution of that undertaking. After the defeat at Zalaca, and with the gradual loss of the income from the parias, preoccupied with the dynastic problem of succession, such a devolution of leadership was both necessary and irresistible. But combined with the preemption of suzerainty in the east by Rodrigo Díaz and the assumption of something like that in the west by Count Raymond, a great danger of the further erosion of practical royal authority was posed. Not that men would cease to venerate their king of a quarter of a century, now in his fifties and without a son, but that increasingly they would look elsewhere for the assistance they sought in the pursuit of their purposes.

1. Abd Allah, El Siglo XI en 1^a persona, pp. 225-29.

2. Ibid. One recalls, of course, that all memoirs are self-justifying by definition.

3. Ibid, p. 228.

4. In what follows immediately here, I am largely rehearsing the findings of a brilliant study by Charles Julian Bishko, "Fernando I and the Origins of the Leonese-Castilian Alliance with Cluny," pp. 1-136.

5. For the document, see Bruel, ed., Recueil des chartes 4:697-98. The royal letter is undated but see Bishko, "Fernando I and Cluny," pp. 71-74, for the chronology and context.

6. Huici Miranda, Grandes batallas de la Reconquista, pp. 89-90.

7. See chapter 10, note 49.

8. AC Oviedo, Serie B., Carpeta 2, no. 12. Pub. García Larragueta, Documentos de Oviedo, pp. 267-68. Despite the present location of the document, the confirmation by the bishops of Burgos, Palencia, and Astorga in addition to Bernard and the absence of the bishop of Oviedo makes León the more probable place of its execution.

9. The private document that mentions his presence, along with that of Bishops Gonzalo of Mondoñedo, Auderico of Túy, and Pedro of Orense, still exists at Braga, Arquivo Distrital, Gaveta 2 das propriedades do Cabido, no. 138. It looks to me like a copy so the date could be mistaken. Costa, O Bispo D. Pedro 2:411, regarded it as an original.

10. AC León, reales, no. 993. Pub. Claudio Sánchez-Albornoz, "Muchas páginas más sobre las Behetrías," AHDE 4 (1927): 146-48.

11. Partially pub. Francisco J. Hernandez, ed., Los cartularios de Toledo (Madrid, 1985), pp. 11-12.

12. Prieto Prieto, "Documentos de Sahagún," pp. 524-25.

13. See chapter 10, note 61. This charter is also the first document that indicates the return of Archbishop Bernard.

14. Duan Gudiol, Collección diplomática de la catedral de Huesca 1:70-73. Kehr, "El papado y los reinos de Navara y Aragón," pp. 125-27, Alfons Becker, Papst Urban II (Stuttgart, 1964), pp. 246-47.

15. Pierre de Marca, Histoire de Béarn ed., V. Dubarat (1640, reprint Pau, 1894), pp. 428-31. Afif Turk, Reino de Zaragoza, pp. 166-67.

16. Domingo J. Buesa Conde, El Rey Sancho Ramírez (Zaragoza, 1978), p. 23.

17. Luis García de Valdeavellano, Orígenes de la burguesía en la España medieval (Madrid, 1969), p. 141.

18. Lacarra, Colección diplomática de Irache, pp. 91-92.

19. Antonio Ubieto Arteta, ed., Crónica de San Juan de la Peña (Valencia, 1961), p. 57.

20. Becker, Papst Urban II, p. 252.

21. Carlo Servatius, Paschalis II, 1099-1118 (Stuttgart, 1979), p. 20. The accommodating attitude of Urban is indicated by the fact that Archbishop Bernard had been consulted about the restoration of Tarragona aud about the choice of a new legate. See Fidel Fita, "Sobre un texto del Arzobispo Don Rodrigo," BRAH 4 (1884): 374-75, and Mansilla, Documentación pontificia, pp. 46-49.

22. AD Santiago, Fondo San Martín, no. 72, an original and an eighteenth-century copy. AHN, Clero, Carpeta 511, no. 17. The copy was published by Antonio López Ferreiro, Historia de Santiago 3:31-33 append. The document was not prepared in the royal chancery, the form of whose products it entirely lacks. It is the very particularity of the document that reassures me as to its reliability. A later forger would hardly have located its production "in domo Petri Vimarat" given that worthy's subsequent reputation. López Ferreiro, ibid., p. 167, n. 2, asserts that Diego Gelmírez added his confirmation later as does Anselm Gordon Biggs, Diego Gelmírez, First Archbishop of Compostela (Washington, D.C., 1949), p. 31, 0. 176.

23. ES 20:18. "Petrus Vimara, laicus, et regius villicus, totum honorem quem Episcopus obtinuerat sanctissimi Regis Domini Adefonsi manu suscipiens, ad propria rediit: ubi tantum crudelitatis pauperes ac divites depraedando instanter exercuit." There were a series of administrators of the see between 1090 and 1100.

24. See Reilly, León-Castilla under Urraca, pp. 333-41.

25. See chapter 10, note 22.

26. AHN, Clero, Carpeta 1.325B, no. 18.

27. Rodríquez Fernández, Pedro Anurez, p. 61, n. 69.

28. Menéndez Pidal, ed., Primera crónica general 2:646. The evidence for the count's guardianship is thus somewhat late.

29. AHN, Códices, 989B, fol. 232V, a private document. A purported royal charter to the monastery of Cardeña of the same date is rather more problematic. Berganza, Antigüedades de España 2:452. The diplomatic is somewhat irregular and it purports to have been confirmed by a Bishop Diego. There was no contemporary bishop of that name except the deposed Diego Peláez of Compostela.

30. Pérez de Urbel and González Ruiz-Zorrilla, eds., Historia Silense, pp. 123-24, establishes the coincidence without furnishing a date. It is our earliest source. The "Chronicon Compostellanum," ES 20:610, the second earliest, gives the date used here and is supported by López Ferreiro, Historia de Santiago 2:554-55, n. 1, whose date is taken from Garcia's sepulchral inscription. Among the later sources Ubieto Arteta, ed., Crónica Najerense, p. 117, gives the same day in 1091 as does Flórez, "Annales Compostellani," ES 23:321. The "Anales Toledanos" in Huici Miranda, ed., Las crónicas latinas de la Reconquista 1:343, gives the dates of 1082. Lucas de Túy "Chronicon Mundi," p. 101, also gave the year as 1091. But by the beginning of 1091 Cardinal Ranier was already back in Rome. Säbekow, Die papstlichen Legationen, p. 32.

31. ES 20:17-18.

32. Antonio Ubieto Arteta, "El destierro del obispo compostelano Diego Peláez en Aragón," CEG 5 (1951): 43-51.

33. "Chronicon Mundi," p. 101.

34. For the now generally received view see David, Études historiques, pp. 432-39. Menéndez Pidal, España del Cid 1:250-51, spoke for the older, generally xenophobic tradition.

35. Das Papsttum und Portugal in ersten Jahrhundert der Portugiesischen Geschichte (Berlin, 1928), pp. 8-9. He is supported by Costa, O Bispo D. Pedro 1:244-46.

36. AC Palencia, Armario3, no. 3; another copy I have seen only in AHN, Microfilmas, Palencia, rollo 1.659, no. 27; also AC Zamora, legajo 8, no. 3; and two late copies in Acad. Hist., Catedrales de España, 9-25-1-C-6, ff. 24r-31r, and Colección Salazar, 0-17, ff. 211r-217r. Pub. Fernández de Pulgar, Historia de Palencia 1:123-25. For a critique see Reilly, "Alfonso VI of León-Castile," pp. 16 and 33.

An agnitio dated Mar. 23, 1090, would put the king in Oviedo and should probably be redated. Another judicial document, dated simply to 1090, would also put Alfonso in that city sometime during the year. Neither is an original. Pub. García Larragueta, Colección de Oviedo, pp. 272-73, and 275-79.

37. Pérez de Urbel and González Ruiz-Zorrilla, eds., Hisroria Silense, p. 124.

38. Bruel, ed., Recueil des chartes 4:809-10, published the French copies. For a Spanish copy, see AHN, Clero, Carpeta 1.700, no. 13, fol. 3r-v.

39. AHN, Clero, Carpeta 886, no. 6. The royal charter to Cardeña dated April 27, 1090, in Berganza, Antigüedades de España 2:450, must be redated to the period 1085-86, if it is to be accepted at all, on the basis of the confirmation by Bishop Sebastian. See chapter 9, notes 47 and 48.

40. The dates cannot be established more closely. My account follows González, Repoblación de Castilla la Nueva 1:88.

41. BN, Manuscritos, 720, ff. 274r-v. A private charter records a donation to the bishopric of León in August 1090, conditional, as the donor says, "si vivens venerit de fossato" AC León, Códice 11, fol. 89r.

42. González, Repoblación de Castilla la Nueva 1:88.

43. Sept. 7, 1090. AHN, Clero, Carpeta 886, no. 9; Códices, 989B, fol. 23r-v dated to Feb. 7, 1090; pub. Escalona, Historia de Sahagún, pp. 487-88, and Prieto Prieto, "Documentos de Sahagún," pp. 525-26. Sept. 16, 1090. AHN, Clero, Carpeta 886, no. 10; and Códices, 989B, fol. 36v, is an exchange between Pedro Ansúrez and Sahagún. Perhaps about this time, certainly between December 1086 and December 1092, on the basis of those who confirm it, Pedro Ansúrez made a very generous donation to the church of León, but the charter lacks a date. AC León, Códice 11, fol. 30r-v; pub. Jorge Serrano Redonnet, "Ovetensis monete," CHE 1-2 (1944): 157-60.

44. González, Repoblación de Castilla la Nueva 1:89.

45. Jan. 1, 1091. AC León Códice 11, fol. 142r-v. Feb. 1, 1091. AHN, Clero, Carpeta 886, no. 12. Feb. 13, 1091. AC León, Códice 11, ff. 84v-85r. Mar. 31, 1091. AC León, Códice s1, ff. 104v-105v; pub. ES 35: 411-14; Manuel Risco, Historia de la ciudad y corte de León y de sus reyes, vol. 1 (Madrid, 1792), p. 395; Muñoz y Romero, ed., Colección de fueros municipales y cartas pueblas, pp. 89-93; Hinojosa, ed., Documentos de León y de CastilIa, pp. 36-3 9; and Justiniano Rodríguez Fernández, La judería de la ciudad de León (León, 1969), pp. 182-85.

There is another, slightly different version of this famous document dated to Feb. 7, 1090. AHN, Códices, 989B, ff. 199v-200v. Pub. Justiniano Rodríguez Fernández, Las juderías de la provincia de León (León, 1976), pp. 342-44. The latter discusses the two versions, pp. 243-44, n. 56, without choosing between them although he does regard them as a single document. The major difference in the latter document occurs in the opening protocol. I would be hard pressed to choose between the versions.

Apr. 1,1091. AHN, Clero, Carpeta 886, no. 16. Apr. 18, 1091. AHN, Microfilmas, AD León, Gradefes, rollo 6.311, no. 6.

46. González, Repoblación de Castilla la Nueva, p. 89.

47. Abd Allah, El siglo XI en 1^a persona, pp. 290-92.

48. Huici Miranda, ed., "Anales Toledanos," Crónicas latinas de la Reconquista, p. 358. "Arrancada sobre Alvar Hanez en Almodovar; era MCXXX." AI.-Maqqarí, The History of the Mohammedan Dynasties in Spain 2:254 and 297.

49. González, Repoblación de Castilla la Nueva, p. 89.

50. AC León, Códice 11, ff. 89v-90r.

51. AHN, Códices, 989B, fol. 159r-v; pub. Hinojosa, ed., Documentos de León y Castilla, pp. 39-40, and Prieto Prieto, "Documentos de Sahagún," pp. 526-27.

52. Serrano, Cartularo de San Millán, p. 281.

53. Sept. 24, 1091. AHN, Códices, 989B, fol. 207r; pub. Prieto Prieto, "Documentos de Sahagún," p. 528. Nov. 3, 1091. Alfonso Andrés, "Monasterio de San Juan de Burgos," BRAH 71 (1917): 119-21, published this from a sixteenth-century manuscript in the Archivo Municipal de Burgos which dated it to 1097, but the date given by Antonio de Yepes, Corónica general de San Benito 6:489v-490r, accords better with the list of those who confirmed it. Also BN, Manuscritos, 5.790, ff. 140v-41r, with the date of 1090. Pub. F. Javier Peña Pérez, ed., Documentación del monasterio de San Juan de Burgos, 1091-1400 (Burgos, 1983), pp. 3-8, from the original and copies in the Archivo Municipal of Burgos with the date of Nov. 3,1091. Nov. l0, 1091. AHN, Códices, 989B, ff. 6v-7r; pub. ES 3 6:74-76, and Escalona, Historia de Sahagún, pp. 488-89.

54. "Chronicon Mundi," p. 101.

55. For his family see Julio González, El reino de Castilla en la época de Alfonso VIII, vol. 1 (Madrid, 1960), pp. 293-94, but with the corrections of Lacarra, El poema de mio Cid, pp. 141-43. But the Count García García of the latter never appears in contemporary documents.

56. Rodrigo Jiménez de Rada, "De rebus Hispaniae," pp. 143-44.

57. Costa, O Bispo D. Pedro 1:246-48.

58. "De rebus Hispaniae," p. 137. I have done a study of Jiménez de Rada on Alfonso VI which may be published before this book. His sources are examined there.

59. Cluny under St. Hugh, 1049-1109 (Notre Dame, 1968), pp. 176-77. Juan Francisco Rivera Recio, El arzobispo de Toledo Don Bernardo de Cluny (Rome, 1962), is the only modern biography and it virtually ignores Bernard's role as chief councilor of the realm.

60. Morales, Viaje a León, p. 37.

61. Dec. 18, 1086. See chapter 10, note 13. Aug. 9, 1095. BN, Manuscritos, 18.387, fol. 300v.

62. See chapter 2, note 36, and Luciano Serrano, "Los Armíldez dc Toledo y el monasterio de Tórtoles," BRAH 103 (1933): 69-73.

63. Dec. 18, 1086. See chapter 10, note 13. It is his sole appearance during the period.

64. July 21, 1087. See chapter 10, note 20, his first appearance. Apr. 30, 1088. See chapter l0, note 45, the last.

65. Dec. 27, 1088. See chapter 10, note 61. Nov. 10, 1091. See note 53.

66. Menéndez Pidal, España del Cid 2:723-24. He was followed by Floriano Cunsbreño, Estudios de historia de Asturias, p. 109. However, see Alfonso Prieto Prieto, "El conde Fruela Múñoz, un asturiano del siglo XI," Asturiensia medievalia 2 (1975): ¡6-17, and Estepa Diez, Estructura social de la ciudad de León, pp. 245-46.

67. See chapter 10, note 13.

68. Turk, Reino de Zaragoza, p. 167.

69. Ubieto Arteta, Historia de Aragón, pp. 104-105.

70. Huici Miranda, Historia musulmana de Valencia 2:33.

71. I depend on Menéndez Pidal, España del Cid 1:376-88, for this account. He, in turn, follows the "Historia Roderici" very literally.

72. Huici Miranda, Historia musulmana de Valencia 2:38-39.

73. Menéndez Pidal, España del Cid 2:941.

74. Ibid., 1:386-88.

75. Ibid., 2:949-51, and 1:400-405.

76. May 6, 1085. AHN, Clero, Carpeta 885, no. 13. "Comes petro ensuriz in toro et in zamora et eius vigarius Otero de sellus."

77. June 24, 1074. AHN, Códices, 989B, fol. 137v; pub. Escalona, Historia de Salhagún, p. 473. "Comite petro anssuriz in Zamora." Apr. 27, 1084. AHN, Clero, Carpeta 885, no. 4. "Imperante zamora comite petro ansuriz."

78. AHN, Clero, Carpeta 886, no. 7.

79. Pub. Pulgar, Historia de Palencia, pp. 135-37, and Mañueco Villalobos and Zurita Nieto, eds., Documentos de Valladolid 1:24-54. See also Rodríguez Fernández, Pedro Ansúrez, pp. 63-68.

80. Reilly, León-Castilla under Queen Urraca, pp. 295-300.

81. José, González, "El monasterio de S. Martín de Pereda," AL 10 (1955): 9.

82. Apr. 7, 1088. Mañueco Villalobos and Zurita Nieto, eds., Documentos de Valladolid, 1:7-10. Sept. 18, 1088. AHN, Clero, Carpeta 885, no. 23. July 8, 1090. Ibid., Carpeta 886, no. 6. Sept. 16, 1090. Ibid., no. 10, Feb. 1, 1091. Ibid., nos. 12 and 13. Feb. 7, 1092. Mañueco Villalobos and Zorita Nieto, eds., Documentos de Valladolid 1:17-18. Mar. 29, 1092. AHN, Clero, Carpeta 887, no. 3.

83. See chapter 8, note 33, and May 28, 1066. AC León, Códice 11, fol. 70r.

84. Serrano, ed., Cartulario de monasterio de Vega, pp. 30-32.

85. Sept. 6, 1094. AHN, Clero, Carpeta 886, no. 7, and ibid., Carpeta 887, no. 4 respectively.

86. See chapter 8, note 3.

87. González, Repoblación de Castilla la Neuva 1:129.