[345] The settlement of the question of Alfonso's successor and the provisions made for the other members of the royal family seem to have allowed the life of the court to return to an orderly routine in the early months of 1108. The tranquility that prevailed there was such that it even admitted of the king's taking of yet another wife without provoking a disturbance serious enough to find a record in the documents or chronicles of the reign.

Sometime in the second half of 1107, Alfonso's current wife died. The last document that records Elizabeth as queen is dated November 17, 1107, but that is badly dated at the very least because it also cites Raymond as still count in Galicia.⁽¹⁾ In the same year the former queen, Elizabeth of France, also died, as recorded by her epitaph in the pantheon of the church of San Isidoro in León.

The king was, then, completely free to remarry. His choice fell upon another French noblewoman named Beatrice about whom we know very little. Bishop Pelayo said only that she survived Alfonso and returned to her own country after his death. The thirteenth-century Primera crónica general identified her nationality as French and added that she had no children by the king.⁽²⁾ The only fully dated notice of this new royal consort comes in a private document of the church of Astorga of May 28, 1108.⁽³⁾ Allowing time for the preliminary negotiations that attended all royal marriages and for the considerable travel involved both for the king's representatives and for the new bride and her party, it seems most probable that the marriage actually would have taken place in the week following Easter Sunday, which that year fell on April 5.

But if the person of the new queen remains mysterious, so too does [346] the point of the marriage. For the first time in his long reign the king was no longer searching desperately for a male heir. His son Sancho stood ready to succeed him and he had as well a grandson, Alfonso Raimúndez, as additional security against the future. Offspring of the marriage could not have been ruled out absolutely, but the king would already have been seventy and perhaps seventy-one at the time of the nuptials.⁽⁴⁾ In addition to being an old man, very old by the standard of times, Alfonso was also at least partially infirm. A contemporary tells us that the monarch had been ill for nineteen months at the time of his death, which would be from December 1107. We are not told the nature of the illness but only that the king remained partially active physically.⁽⁵⁾

All things considered, it seems unlikely that the desire for additional progeny was the object of the new marriage, yet surely private gratification would not have required so formal a procedure, even assuming that Alfonso still felt the itch of the flesh. What we are forced to posit is rather that the marriage secured some diplomatic or political advantage to the crown. By analogy to most previous marriages of Alfonso it seems likely that Beatrice was of some at least moderately important south French noble line, perhaps Burgundian. A line related to Abbot Hugh the Great of Cluny would have offered the most obvious advantages.

While the arrangements for the marriage were being completed the royal court was apparently in its usual winter location at Sahagún. It is difficult to be positive because of the dearth of royal documents in the last year and a half of the reign. A purported donation of Alfonso to the Castilian monastery of San Millán de La Cogolla is dated only to 1108.⁽⁶⁾ Dated in the same fashion is an agnitio of Astorga which furnishes our only other documentary mention of Beatrice as queen.⁽⁷⁾ As a consequence we must depend almost exclusively on private documents and chronicles for our knowledge of the king's whereabouts. Three private donations to the monastery of Sahagún indicate, by those who confirm them, the likelihood of the presence of the court there from February 16 to March 31, 1108.⁽⁸⁾

[347] The first and the last of these documents place Count Henry of Portugal at the royal court during this same period. That was to be expected in that the count had been a regular court figure during the winter months for the past ten years. Now, since the death of Count Raymond would have resulted in a greatly enhanced prominence there for Henry and since his royal father-in-law was contracting another marriage, it had become of even more importance to him to maintain a presence. But this means that the count's grant of exemption to the church of San Martín de Espiunca, near Arouca, on March 31, 1108, must have been executed at Sahagún despite the Portuguese who figure in its confirmation list.⁽⁹⁾ At this time the count served to attach Portuguese magnates to the Leonese orbit rather than to be drawn himself into a provincial one.

Being present there Henry would have been at least nominally consulted over the promotion of a new bishop to the very important see of Palencia. Bishop Raymond had died on January 12, 1108.⁽¹⁰⁾ The real choice lay with the king and Archbishop Bernard of Toledo, who acted with noticeable speed to fill this key bishopric. The name of the new Bishop Pedro of Palencia appears already in a private charter of Sahagún dated March 2, 1108.⁽¹¹⁾ The appointee again was one of those French clerics brought to Spain by Bernard in 1096.⁽¹²⁾ By April 11, Pope Paschal II had been informed of the new bishop, about whose elevation he obviously had nothing to do if the time elapsed is considered.⁽¹³⁾

Although he acquiesced in that determination, Paschal II continued to be less than cordial to Bernard of Toledo. This same letter informed the bishops of León, Compostela, Palencia, Nájera, and Astorga of the papal decision against Toledo, and its new suffragan at Osma, and in favor of Burgos in the matter of some disputed territories. Another papal [348] letter of November 12, 1108, conveyed yet a further decision in favor of Burgos, made directly by Paschal II without reference to the primate and papal legate in the peninsula.⁽¹⁴⁾

Whether or not any of the very active life at the royal court during the winter and spring of the year involved Alfonso's oldest child, Infanta Urraca, is more difficult to determine. Obviously she had attended the council at León in late December 1107. Later in the spring, however, Urraca was apparently in Galicia when the great emergency following the disaster of Uclés arose, as we shall see. A charter that she granted to the church of Lugo, dated January 21, 1108, suggests that she had withdrawn from court that early and was concentrating on establishing a firm base for her authority in the northwest.⁽¹⁵⁾ Unfortunately the list of those who confirmed is lost so that no indication of the place of issuance remains, and such charters were as often issued at court as in the church to which they were granted. If the infanta were in eclipse after her husband's death, events were soon to change her status dramatically.

In the spring of 1108 the Murâbit of Andalucía decided to resume their offensive, which had faltered badly since their reoccupation of Valencia in 1102. Their ultimate objective was the city of Toledo certainly, but the first and only battle revolved about the tiny fortress of Uclés.⁽¹⁶⁾

This latter was a settlement some 30 kilometers south of the Tajo and 103 kilometers east of Toledo. The Muslim sources describe it as the capital of the district and it was probably one of the few centers that saw new Christian efforts at resettlement and expansion south of the Tajo after Alfonso VI's successes in 1104 and 1105 against the Murâbit. It would have been a key position in a drive to turn the eastern flank of Toledo and to roll up the Christian positions between the Tajo and the Guadarramas.

Certainly the composition of the Muslim army suggests that intention as well. The commander of that force was Tamin ibn-Yusuf, governor of Granada and brother of the new Murâbit emir, Ali ibn-Yusuf. Tamin had at his disposal the forces of Granada and of Córdoba but [349] more significantly those of the governors of Murcia and Valencia, whose interest in the particular ends of the campaign would have been more direct.

Tamin left Granada between May 2 and May 12 and proceeded to Jaén where he was joined by the Cordoban contingent. From there he proceeded through Baeza to Chinchilla, where juncture was made with the governors of Murcia and Valencia. The united army then marched to Uclés itself where the town proper fell immediately into their hands on May 27, the Christian defenders retreating into its alcázar or citadel.

The news of their coming had preceded them of course, and a Christian response was mounted with some alacrity. The "Historia Compostelana" suggests that elements of the Murâbit force had fanned out over the meseta to some extent, burning and sacking lesser centers of Christian population and driving their inhabitants before them. The same account places the initiative for the counterattack in the hands of the young Infans Sancho Alfónsez, to whom his father had entrusted the rule of Toledo.⁽¹⁷⁾ Possibly this change too had been effected at León in December 1107. Subsequent developments make it likely that Alfonso VI was still in the north, as might have been expected if he had solemnized his marriage shortly after Easter on April 5, rather than directing events himself as the Muslim accounts imply.

In all probability the young infans had come south from León to Toledo in early or mid-April to make preparations for an ordinary opening of the summer campaigning season. He seems to have brought with him a force of moderate size. It was not at all a general levy of the realm. While other accounts are not quite so precise as the Crónica Najerense, there seems to be general agreement that seven Christian counts were slain at Uclés and that one, Alvar Fáñez, escaped.⁽¹⁸⁾ Since the counts, or in Castilla magnates, of that period who can be identified number some twenty-seven and the bishops seventeen, the heavy cavalry would have represented a little less than one-fifth of the resources of the whole kingdom.

If we estimate that each of the counts, including Alvar Fáñez who did not hold that title, and the bodyguard of Sancho at 50 knights apiece, the heavy cavalry would have numbered 400 knights plus 400 squires, and another 400 grooms or attendants would have accompanied them. In addition, the Muslim sources add that the Christians of Calatañazor, Alcalá, and Toledo, led by their alcaldes, formed part of the army. One [350] thinks that these latter contingents were composed mostly of foot in the eleventh century, but there were probably some lightly armed horse among them. Given the sparse character of early Christian settlement in Castilla la Nueva a generous estimate of their numbers might be 750 fighting men. If we add yet another 300 men involved in the supply trains of the various contingents, a total force of about 2,300 emerges. Since the subsequent disaster engulfed nearly the entire Christian army, the Murâbit General Tamin did not exaggerate too greatly when he reported to his brother, the Emir Ali, that 3,000 Christian heads were taken. The latter count may also have included some of the hapless population of the countryside around Uclés. Then too, not all of the Christian participants may have been identified and hence went unreported. The presence of the militia of Calatañazor, a town far north of the Tajo valley in the province of Soria, in particular is curious.

The battle itself developed in the usual fashion of these clashes. The Christian force struck frontally against the Cordoban contingent of the Murâbit, hoping to win by the sheer shock value of its heavy cavalry, and had some initial success as at Zalaca more than twenty years before. The Muslim army resorted to the enveloping tactics which could only have been possible to a force greatly superior in numbers. The troops of Murcia and Valencia attacked the camp and supply depot of the Christians. The student of the art of warfare of the times would be grateful to know whether the Leonese forces had arrived the night before and hence established a camp or whether such a practice of forming a camp was simply a standard battlefield procedure. I suspect the latter was the case. In any event, it evidently offered some better opportunities of resistance for a beleaguered army. The Christian cavalry now retreated on their own camp and there conducted a desperate but futile resistance to the reunited enemy forces. Unlike Zalaca the main force never broke free and was slaughtered on the field.

Only some scattered elements managed to escape. These included the infans himself and a small body of retainers who contrived to reach the village of Belinchón, twenty kilometers to the northwest, but were cut down there by the local Muslim population who took advantage of the Leonese defeat to rise against their conquerors. Alvar Fáñez, who must have managed to break through the Murâbit host with a fair-sized group of horsemen, fared better and escaped north to organize a defense of the now endangered valley of the middle and upper Tajo. So ended the bloody battle at Uclés on May 29, 1108. The full consequences of the Leonese defeat there, however, would go on working themselves out for better than a decade before stability could ultimately be restored.

[351] While Tamin returned to Andalucía the governors of Murcia and Valencia were left to conduct mopping-up operations. Part of this was the final reduction of Uclés itself where the defenders were tricked into sallying from the alcázar and then cut down. The major portion of the Muslim energies must then have been directed to securing the whole of the south bank of the Tajo between Aránjuez and Zorita. Certainly the rebellion of the Muslim of Belinchón was not an isolated phenomenon nor one whose advantages were to be neglected. While our sources of information leave much to be desired as to the particular chronology of these events, it seems to be clear that Huete, like Uclés a center of repopulation southeast of the Tajo and some thirty-five kilometers northeast of the latter city, also fell in the ensuing summer.⁽¹⁹⁾ The way was thus prepared for the Murâbit Emir Ai ibn-Yusuf himself to launch a campaign in the years 1109-1110 whose ultimate results were to reopen the direct road from Andalucía to Zaragoza and to bring that last great taifa itself into the North African empire.

Even before the news reached him of the rout at Uclés, Alfonso VI may have been moving south from Sahagún to direct the usual summer campaigns. Although the command he had entrusted to his young heir would have been exercised at the discretion of the seasoned warriors, the counts who surrounded the latter, the king would have desired to oversee events himself. But since the documentary record fails entirely between the beginning of April and the beginning of September in 1108, we can only resort to the imperfect and imprecise record of the "Historia Compostelana" for a reconstruction of subsequent events. Bishop Diego Gelmírez, we are told, hastened south with Urraca and a force raised in Galicia when he learned of the death of Sancho Alfónsez. Surely he was responding to a royal summons which would have called out every available warrior to deal with the emergency. The doughty bishop, we are told, put the enemy to flight and then fell ill. When he had recovered, he then joined the king in Segovia. There he received a long-promised donation and Alfonso VI informed him of his plans, first to visit Toledo and see to its fortifications and then to go as a pilgrim to the church of Santiago in Compostela.⁽²⁰⁾

A tentative chronology can be constructed about this account. For the news of the defeat to have reached Gelmírez in far off Galicia, a distance of something like 725 kilometers by the most direct routes, surely two weeks were necessary at the very least. Then he would have had to raise the forces available to him and make the preparations for a very [352] long march indeed. Given the rate of march of an expeditionary force during the period, and allowing an additional week for its organization, a minimum of five weeks would have been required to reach Toledo after the summons had arrived. The earliest that the army of Galicia could have appeared on the frontier then would have been the end of the first week in July. Allowing two weeks for skirmishing against such Murâbit forces as remained in the field and for punitive action against indigenous Muslim populations north of the Tajo who had risen in revolt, yet another week for Gelmírez's illness, and a final week for the trip north from Toledo to Segovia, we can estimate that Gelmírez met the king in the latter city about August 4 or 5, 1108.

Since Alfonso stated at that time his intentions to proceed to Toledo and examine its defenses we can infer that he had as yet not traveled south of Segovia but was engaged in a general oversight of repairing defenses and restoring his authority in the long, threatened eastern triangle between Madrid, Toledo and Zorita, perhaps even stretching northward toward Medinaceli. That he had not taken directly to the field himself may also testify to his generally weakened health.

A quick royal visit to Toledo may have been carried out at the end of the first week of August. There would hardly have been time, however, for the sort of great curia that concurred in the king's choice of Urraca as his heir and his decision to arrange her marriage to Alfonso of Aragón to be held there, as some have envisioned.⁽²¹⁾ The recent presence of the king in that city would nevertheless have made most unlikely the massacre of the Jews of Toledo which the "Anales Toledanos" somewhat confusedly dates to August 14, 1108. The circumstances as well as a considered chronology suggest rather 1110 as the actual time of the latter.⁽²²⁾

Before the end of the first week in September the king and his court had returned once again to Sahagún.⁽²³⁾ It is noteworthy that Count Henry was in attendance. Did the king then fulfill his intention to make a pilgrimage to Compostela? There is no evidence either for or against the possibility for there is a total absence of relevant documents between early September and the following January. I suspect that he did not for it would be strange that the author of the "Historia Compostelana," [353] given his continuous attention to everything that would glorify either that church or its bishop, should not have at least referred to the actual performance of that pious act by the king. In addition, the fall of 1108 would have represented the first substantial opportunity for Alfonso VI to turn his attention from the military crisis created by the defeat at Uclés to the even more critical political crisis spawned by the same event.

That sudden and unexpected turn in the fortunes of the realm has ordinarily and quite properly been considered in terms of the consequent lack of an adult male heir to the throne and the steps Alfonso VI took to provide against the expected weakness of the crown if his daughter, Urraca, should inherit. But such an analysis, in itself, is necessarily partial and, I believe, does less to illumine the grounds for the king's ultimate decision than we might reasonably expect. After all, Sancho Alfónsez did not perish alone at Uclés but rather surrounded by the "seven counts," the entourage provided to the hope of the kingdom by his father. Alfonso VI would no more have expected his fifteen-year-old son to rule in Toledo without the direction and advice of more mature and sober heads than he would have expected the boy to meet a challenge like Uclés out of his own limited experience. As a result, what the king and kingdom lost at the end of May in 1108 was not simply its male heir but also a sizable contingent of its most powerful and sagacious magnates.

Unfortunately the chronicles tell us little enough of their particular identities. One of the few and the earliest to identify an individual tells us "occisus est comes Garsias de Grannione cognomento Crispus et sex alii comites cum eo."⁽²⁴⁾ This clearly is Count García Ordóñez of Nájera, for more than a quarter of a century a trusted lieutenant of Alfonso. Sure enough he disappears abruptly from the documents after a last citation in a private document of the Riojan monastery of Valbanera dated only to 1108.⁽²⁵⁾ In the first half of the thirteenth century Lucas of Túy writes "occubuit Sancius Regis filius et comes Garsias Fernandi et comes domnus Martinus et alii plures."⁽²⁶⁾ The first of these is surely a mistake either of the author or of the copyist. Lucas's work never has been critically edited. The preferred reading would be "García Ordóñez."⁽²⁷⁾ The second is a valuable addition and it can refer to no other [354] than Martín Laíñez, the only count of that given name at the time. He was a great magnate of León and a major court figure for more than ten years. His last fully datable notice in the documents was in the royal charter of May 14, 1107.⁽²⁸⁾ The effect of his demise was compounded by that of his son, Gomez Martínez, who also appeared for the last time in the same document as "filius comitis," and who probably perished with his father.

Relying solely on the documents we may suspect that one of the other counts was Asturian and two were Castilians. The first of these is Count Fernando Díaz, who vanishes from sight after the royal charter to the church of Oviedo dated March 19, 1106.⁽²⁹⁾ Like most Asturians the count was never a regular figure of the court so that a two-year absence from the charters does not necessarily argue a demise before Uclés. In Asturias though, he was the greatest of the secular magnates.

Neither of the Castilian magnates was technically a count, for that title was rarely employed in that province. Nonetheless all were of such a stature that subsequent writers could easily have assimilated them to that dignity for literary effect. Both made their last appearance in Alfonso VI's grant to San Millán de La Cogolla dated simply 1108.⁽³⁰⁾ Diego Sánchez had regularly confirmed royal charters concerning Castilla since 1085. Lop Sánchez, in all probability his older brother, had done the same. Both were also nephews of Lop Jiménez, Count of Vizcaya and Alava in the 1080s and early 1090s.⁽³¹⁾

If these identifications of the magnates killed at Uclés are correct, even in part, then one sees that the problems facing the king in the fall of 1108 compounded and bedeviled the decision shortly to be made concerning a new heir. Were the succession to pass to a minor or a woman the ordinarily expected weakness of the crown under such circumstances would be magnified by the contemporary confusion and rivalry attendant upon the sudden demise of the greatest magnate of Asturias. [355] The latter had no sons, so far as we can tell, and the countship itself in Asturias de Oviedo disappeared.⁽³²⁾

Somewhat the same situation obtained in the heartland of Léon itself. Martín Laíñez, given the exile of Count Pedro Ansúrez in 1103, had been the most important single noble there. Now his sudden demise, along with the simultaneous death of his son and heir, created a vacuum which guaranteed trouble even for a strong ruler. But a strong ruler could also take advantage of that situation as was true in Asturias de Oviedo.⁽³³⁾ On the person of the heir depended the danger of near chaos or potentially great consolidation for the crown.

These considerations, while far from minor, were dwarfed by the disruption of the eastern frontier in the Rioja and Castilla. In the former the death at Uclés of Count García Ordóñez removed the major prop of royal authority there. For twenty-seven years García had held the rich lands that stretched along the west bank of the middle Ebro against the always potential greed of the rulers of the taifa of Zaragoza immediately to the south and of the rapidly growing kingdom of Aragón which held the east bank. The sudden disappearance of the magnate who had held the province for León-Castilla was bound to tempt them to fish in troubled waters.

That danger was magnified by the loss also at Uclés of a number of Castilian magnates whose personal and historic interest in the Rioja had reinforced the stability of that district. Even before that fatal battle the lord of Oca, Alvar Díaz, had died. Alfonso VI's charter to San Millán of 1108 speaks of him as already deceased.⁽³⁴⁾ Not only had his holdings controlled the main road and pass from Castilla la Vieja to the Rioja, but he was allied by marriage to the Ordóñez.⁽³⁵⁾ Again the closely contemporary death of his son, García Alvarez, whether or not the latter was killed at Uclés, would have been doubly disruptive of the political patterns that had linked Castilla and the Rioja for more than a quarter of a century.⁽³⁶⁾ Finally the deaths of Lop Sánchez and Diego Sánchez, the family who controlled the district around Haro in the northern Rioja, completed the elements necessary for near political chaos on the eastern frontier.⁽³⁷⁾

[356] In the fall of 1108 Alfonso VI's search for a political solution to the problem created by the death of the young Sancho Alfónsez was massively complicated by all of these related events. In all probability the bishops and the magnates of the realm responded to these same considerations as well. Jiménez de Rada relates the story that they advanced the candidacy of the Lara count, Gomez González, to become the husband of Urraca, the obvious successor to the king as his oldest surviving daughter. Such a marriage to the greatest of the Castilian nobles then surviving would obviously have provided a certain stability in the east of the kingdom. The court, however, seems to have expected a certain resistance to this solution on the monarch's part for they seem to have acted indirectly and through the mediation of the kings adviser and physician, Joseph Ferrizuel, surnamed Cidellus.⁽³⁸⁾ In fact the proposal was angrily rejected by Alfonso, who imposed the choice of Alfonso I of Aragón, "El Batallador" as the husband-to-be of Urraca.

Fixing the chronology of these events is extremely difficult. The military necessities of the summer of 1108 would seem to have precluded such deliberations and consultations. It is likely that when Urraca, in the company of Bishop Diego Gelmírez, joined Alfonso in Segovia in early August the king assured his daughter that she would be his choice as successor. As mentioned above, by early September the king and court had returned to Sahagún but it cannot be securely located again until the following March finds them there once again. Nevertheless it seems to me that more than common sense indicates the fall of 1108 as the period in which the Aragonese marriage was fixed upon and in which the inevitable negotiations with that king would have been begun.

In retrospect the decision made by Alfonso VI to marry his daughter to Alfonso I of Aragón appears as the gravest mistake of his entire reign. As is well known, the marriage itself was sterile and so did not, in itself, solve the dynastic problem of succession to the throne of León-Castilla. In addition, the majority of the nobility of the realm refused, after the old king's death, to accept the Aragonese as their sovereign. So too did virtually the entire episcopate, in part because Urraca and Alfonso of Aragón, having a common great-grandfather in Sancho el Mayor of Navarra, were within the canonically prohibited degrees of relationship. Pope Paschal II eventually condemned the union on this ground.

But the marriage also failed because it outraged the contemporary [357] sense of dynastic proprieties as well. Whatever guarantees of lesser status were provided, the fact was that the claim of the king's legitimate grandson, Alfonso Raimúndez, to succeed was overridden. Nevertheless, it was almost immediately to find defenders in the bishop of Santiago de Compostela and among the great nobility of Galicia. The claims of Alfonso VI's natural daughter Teresa and her ambitious husband, as well as her potential offspring, to at least a place in a council of regency were simply ignored. Although Urraca herself was to try loyally to comply with the will of her father, she was soon to find that the conjunction of these factors made the governing of the kingdom impossible. She was forced to break with her new husband as a matter of policy regardless of the controverted question of whether or not that separation was personally congenial to her. The result of that decision was a long, intermittent war with Aragón which, with truces, lasted into the reign of her son, Alfonso VII.

Yet surely Alfonso VI foresaw all or most of these dangers. Within the policies and realities that were the substance of his thought they were acceptable risks. We must not commit the anachronism of attributing to him our own nationalist ideal of a united Spain. Before all else he would have been guided by the overriding necessity of an able warrior king in the face of the ever more menacing Murâbit threat to Toledo and perhaps even to the new lands in the trans-Duero. The king of Aragón possessed this qualification in abundance. Of that the Leonese monarch was apprized particularly by his old companion in arms, Count Pedro Ansúrez, who had been cooperating with the Aragonese from his base in exile at Urgel for four years at the time.⁽³⁹⁾

Then too, the marriage he devised for his daughter followed the same policy that had governed his own marital choices. Always the king had sought to insulate the crown from the claims of his domestic nobility while also adorning the former with the prestige of a foreign alliance. The choice of his prelates and magnates, Count Gomez González of Lara, would inevitably have created pretensions in that great Castilian house at the same time it would have exacerbated the outrage of the partisans of his grandson and of his natural daughter Teresa. Alfonso of Aragón brought with him not merely the laurels of a warrior but the charisma of kingship as well.

Finally perhaps, the Aragonese monarch might be less dangerous as a sovereign than as a neighbor. He had been following a highly aggressive [358] policy of expansion in all possible directions since his accession five years before. Now, in the aftermath of Uclés, all of the western Rioja lay open to his designs as did much of eastern Castilla. Neither Urraca herself, or even married to Lara count, could be expected to hold them against the Aragonese. That would be especially so if they were handicapped by serious domestic discord. The best defense for those exposed territories, in the long run, might be a diplomatic one. It was preferable that the Batallador should hold them as a king of León-Castilla than as a king of Aragón.

Nevertheless, all of these defensible considerations led to a disastrous decision on the part of the old king. He might understand that prelates, popes, and nobles would eventually recognize the rights of children born to the marriage of Urraca and Alfonso even if they never would accept the marriage itself. What he could not know was that the marriage would produce no children. But his most critical failure lay in his inability to appreciate the political skills of his own daughter. In the next quarter of a century those were to prove more than adequate to overmatch the military abilities of her many opponents as well as such political and diplomatic acumen as they possessed. In short, he never realized the extent to which she was, in truth, his daughter.

Whatever the attitude of Urraca herself was toward Alfonso's decision, it seems likely that she participated in the negotiations. After the meeting with her father in Segovia in August 1108, no document specifies her whereabouts for six months. That she spent the fall and the Christmas season at court is probable in view of what had been the customary practice and the fact that she was central to the negotiations in train. But on February 22, 1109, the queen to be was in Galicia. There she donated the monastery of San Vicente de Pombeiro to Cluny and the grant was confirmed exclusively by the prelates of the northwest.⁽⁴⁰⁾ Bishko regarded this charter as evidence of a bid for Cluny's support against the possible partisans of her son.⁽⁴¹⁾ If her remarriage was already in prospect such a motivation is quite credible. Acceptance of the latter by the bishops of Galicia, whatever might have been their private reservations, would also be indicated by their confirmations.

By March 27, 1109, Urraca had returned to Sahagún and the court and had brought with her at least Bishop Gelmírez of Compostela, who also had confirmed the Cluny grant.⁽⁴²⁾ From this time the heiress was probably constantly in her father's company.

[359] Similarly the attendance at court of Count Henry and Infanta Teresa all through this period is to be presumed. However disgruntled by the course events were taking, they could not have afforded to be absent. Count Henry's presence is attested for September 4, 1108, for March 27, 1109, and for May 22, 1109.⁽⁴³⁾ The king would have been subjected regularly to the polite but firm opposition of the two to the course he had chosen. If his daughter was already carrying or, in 1109, had given birth to Alfonso Enríquez, their eventual successor in Portugal, her position would have been strengthened materially. Existing sources differ notably as to the infans' year of birth, but 1109-1110 seems to me the most likely time.⁽⁴⁴⁾

On the subject of marriage the king was unyielding as we know. Yet he may have been prepared to make minor concessions to assuage the couple's feelings. There is a document, dated April 25, 1109, which cites Henry and Teresa as "imperante Tiniego."⁽⁴⁵⁾ The latter was an important hamlet in western Asturias de Oviedo near two passes over the Cantabrians from that province into northern León. Such a dignity was easily conferred, given the death of Count Fernando Díaz, and might even have been useful.

Further clues to the tenor of relationships between the crown and the couple may be furnished by ecclesiastical changes in Portugal at this time but the chronology is difficult to establish. At Braga Archbishop Gerald died in either late 1108 or early January 1109. In March Paschal II at Rome was still unaware of his death.⁽⁴⁶⁾ That prelate had been the instrument of the king and the archbishop of Toledo in the province. Now Gerald would be replaced by Maurice, formerly bishop of Coimbra. The new archbishop would have a famous career, ending finally in ignominy when, as the imperial claimant to the papacy Gregory VIII, [360] he was captured and imprisoned for life. But in January 1109 when Maurice was translated from Coimbra to Braga he seems still merely to have been the protégé of Alfonso and Archbishop Bernard.⁽⁴⁷⁾

The very fact of his translation guarantees that it was the king and not the count who stood behind the appointment. The translation from one see to another was one of those few ecclesiastical processes in which the consent of the Roman pope was more than a mere formality in this period, and only the king possessed sufficient influence in Rome to secure that assent. Moreover, the confirmation by the new Archbishop Maurice in late February of Urraca's donation to Cluny indicates his willingness to accept the royal plans for her marriage. Such adhesion surely was the price of his consecration which may well have taken place at Compostela on or near that very date.⁽⁴⁸⁾

With the royal blessing, Maurice then left for Rome to secure the proper papal recognition. Possibly while he was gone a new bishop was elevated to his old see of Coimbra. Again the decision lay with the king and his great aide, the Toledan archbishop, as demonstrated by the fact that the new incumbent recognized Toledo rather than Braga as his metropolitan. Nevertheless, it is a mistake, based on the later quarrel between the two archbishoprics which erupted after Alfonso VI's death, to regard his appointment as some sort of Toledan coup pushed through in Maurice's absence.⁽⁴⁹⁾

The new bishop of Coimbra, Gonçalvo Paies, was a scion of the Portuguese magnate Paio Peres.⁽⁵⁰⁾ His appointment was a part of the careful balancing of influence in that distant province by the crown and was in all likelihood at least formally agreed upon by Maurice himself, the Portuguese magnates, and Count Henry. The generous donation made to Coimbra on July 29, 1109, by Henry and Teresa should be seen as their first bid to purchase Gonçalvo's support after Alfonso's death. Mere chance determines that it is our first notice of Gonçalvo's incumbency.⁽⁵¹⁾

All of these events demonstrate clearly that the Leonese monarch [361] continued until his death to regard the control of appointments to the episcopate of the realm as an indispensable prop of the royal authority. If he were willing to make some minor concession to Henry and Teresa in order to soften the blow constituted by his decision on the succession question, these were not to be found in the realm of high ecclesiastical appointments. The royal determination to retain such control must have further increased the frustration of his second daughter and her husband. When they left the court in the late spring of 1109 they may already have formed the resolve not to return except as its masters.

That departure must have taken place after May 22, 1109, and probably coincided with the departure of the royal court itself for Toledo.⁽⁵²⁾ A contemporary informs us that the leave-taking was anything but cordial.⁽⁵³⁾ The context of his remarks suggests that Count Henry accompanied the king to Toledo but other considerations make that unlikely. In July, after Alfonso's death, the count captured Santarem in Portugal.⁽⁵⁴⁾ Now a direct journey from Toledo to Santarem would have been hazardous in the extreme if not impossible, and a more roundabout trip could not have been completed in the time indicated. Henry and Teresa, then, refused to journey to Toledo precisely because the king intended to proclaim Urraca officially his heir and successor there. The two also remained in Portugal after his death and failed to attend his obsequies in Sahagún or the inauguration of Urraca's reign in late July.⁽⁵⁵⁾

The journey of Alfonso VI and his court to Toledo in the late spring of 1109 had two purposes. One of them was to launch a general offensive against the Murâbit aimed at recovering the losses of the previous year and to forestall any further exploitation by the latter of their successes at and after Uclés. The death of Alfonso in Toledo effectively thwarted those intentions. Only largely local forces were available in August for an unsuccessful siege of Alcalá de Henares, which had passed into Muslim hands sometime in late 1108 or early 1109. In the summer of 1109 the Murâbit Emir Ali ibn-Yusuf crossed the straits into Spain and during the 1110 waged a major offensive all along the line of the Tajo.⁽⁵⁶⁾

The second purpose was, of course, to proclaim Urraca as his heir in the old capital of the Visigoths and the primatial see of León-Castilla. [362] If the king had left Sahagún for Toledo sometime after May 22, 1109, he could hardly have arrived there before the end of the second week in June, given what must have been the progressively worsening condition of Alfonso. That this journey gave the coup de grace to his health is a fair assumption.

Nevertheless, Alfonso there proclaimed his daughter as his heir in the presence of "almost all of the nobles and counts of Spain," says an eyewitness.⁽⁵⁷⁾ Urraca herself also later fixed the place and the time, although the latter only roughly.⁽⁵⁸⁾ The royal announcement would have included at least the provisional terms of the impending marriage of Urraca to Alfonso I of Aragón, although we know that the official agreement, or carta de arras, was not executed until six months later. The argument that the young queen had already married the Aragonese has been made in an ingenious but unconvincing fashion.⁽⁵⁹⁾

The activities and whereabouts of the king of Aragón during the critical period between the battle of Uclés and the death of Alfonso VI cannot be specified very fully. Early in 1108 the Murâbit governor of Valencia had launched a major raid into Catalonia which had carried almost to Barcelona itself.⁽⁶⁰⁾ Of necessity this offensive would have unsettled the very recently established southeastern frontier of Aragón. It is not surprising then to find the Aragonese monarch at Huesca in December 1108 in the only one of his charters known from this entire period. In its final protocol he made no reference to a queen or to any claim on the Leonese kingdom.⁽⁶¹⁾

[363] On July 1, 1109, in the seventy-second year of his life and the forty-fourth year of his reign Alfonso VI died in Toledo.⁽⁶²⁾ As always under the monarchy the death of the king threw all existing political understandings into a state of flux. Therefore, despite the practical concern for the defenses of the Tajo frontier, virtually the entire court hastened north with the royal corpse. On July 21, 1109, the greatest monarch Christian Spain had seen since the Roman era was laid to rest in the royal monastery of Sahagún. In attendance were twelve bishops and seven counts of the realm.⁽⁶³⁾ They constituted a virtual roster of the notables of the court and the kingdom at the time.

The long reign of Alfonso VI had closed and the troubled one of Urraca had begun. The Murâbit Emir Ali was just about to debark in Andalucía to undertake the reconquest of Toledo. Count Henry and Infanta Teresa were sulking in self-imposed exile from court at Coimbra in Portugal. Count Pedro Ansúrez had just returned to the Leonese court as the protagonist of the royal marriage between Urraca and Alfonso I of Aragón. The far-flung kingdom that the dead king had constructed, defended, and ruled for more than four decades was about to be tested in a vortex of invasion and civil war which would endure for more than two decades.

1. AHN, Clero, Carpeta 892, no. 20; pub. Escalona, Historia de Sahagún, pp. 506-507. Vignau, ed., Documentos de Sahagún, p. 351, further confused the matter by giving the date as Nov. 17, 1108.

2. Sánchez Alonso, ed., Crónica del Obispo Don Pelayo, p. 86. Menéndez Pidal, ed., Primera crónica general 2:521.

3. BN, Manuscritos, 4.357, fol. 40r.

4. See chapter 2, note 21.

5. Sánchez Alonso, ed., Crónica del Obispo Don Pelayo, p. 84, "et quamvis esset infirmus omni die aliquantulum equitabat iussu medicorum ut aliquod levamen corporis haberet."

6. Serrano, ed., Cartulario de San Millán, p. 297. In addition the preserved text is at best abbreviated for the diplomatic is clearly not that of the royal chancery.

7. BN, Manuscritos, 9.194, fol. 103r-v.

8. Feb. 16, 1108. AHN, Códices, 989B, fol. 165v. Mar. 2, 1108. Ibid., ff. 115v-116r. Mar. 31, 1108. AHN, Clero, Carpeta 892, nos. 17 and 18; copy in Códices, 989B, fol. 219r-v.

9. Rui Pinto de Azevado, ed., Documentos Medievais Portugueses, vol. 1 (Lisbon, 1958), pp. 17-18. Azevado calls an original the Henrician charter granting the fueros of Coimbra to the settlers of Tentugal, dated only to 1108, ibid., p. 16. However, that can hardly be the case if one compares it with Henry's charter of July 29, 1109, ibid., pp. 19-20, attributed to the same notary. Neither the paleography nor the diplomatic corresponds. The Tentugal fuero should be regarded with extreme caution.

10. Fernández de Madrid, Silva Palentina 1:147, n. 1.

11. See note 8.

12. Serrano, Obispado de Burgos 2:17, n. 1.

13. Ibid., 3:132-33. The private document, dated Sept. 30, 1107, given in José María Fernández Catón, "Documentos leoneses en escritura visigótica," AL 27 (1973): 222-23, would seem to be misdated. It cites Pedro as already bishop at Palencia. It is a copy.

14. Serrano, Obispado de Burgos 3:134-35.

15. AHN, Códices, 1.043B, ff. 16v-17r. The year is given as 1107 in the text of this copy but is obviously mistaken since the donation is made "pro animi viri mei gloriosissimi ducis domni Ramundi."

16. The campaign has been discussed at length by Huici Miranda, Grandes batallas de la Reconquista, pp. 114-17; and again in his Historia musulmana de Valencia 2:21-28. He based his account on the Muslim sources which are the earliest and fullest. More recently his account has been refined and partially corrected by John E. Slaughter, "De nuevo sobre la batalla de Uclés," AEM 9 (1974-79). 393-404. For the most part my account follows Slaughter unless otherwise noted.

17. ES 20:67.

18. Ubieto Arteta, ed., Crónica Najerense, p. 118, identifies only one, García Ordóñez, count in Nájera.

19. For a critique see González, Repoblación de Castilla la Nueva 1:98-101.

20. ES 20:67-68.

21. García de Valdeavellano, Historia de España 1-2:390-91. No contemporary evidence for such an assemblage exists.

22. Huici Miranda, ed., Las crónicas latinas de la Reconquista 1:344. "Mataron a los Judios en Toledo dia de Domingo, Vispera de Santa María de Agosto, Era MCXLVI." But the feast of the Assumption fell on Saturday in 1108 whereas in 1110 it fell on a Monday.

23. Sept. 4, 1108. AHN, Códices, 989B, fol. 80r. Sept. 5, 1108. AHN, Clero, Carpeta 892, no. 19.

24. See note 18.

25. Manuel Lucas Alvarez, ed., "Libro Becerro del monasterio de Valbanera," EEMCA 4 (1951): 598-99.

26. "Chronicon Mundi ab Origine Mundi usque ad Eram MCCLXXIV," Hispaniae IIlustratae, ed., Andreas Schottus, 4:102.

27. A "García Fernández" appears in only two documents of Alfonso's reign and never as a count. Lacarra, El poema de mio Cid, pp. 142-43, has a "García Garcés" perishing at Uclés but no such name appears in the documents of Alfonso's reign to my knowledge. García Alvarez, alférez of the king perhaps as late as Dec. 27, 1107, AHN, Clero, Carpeta 249, no. 2, may be another possible reading for he disappears from the documents thereafter. However, he was never a count so far as we can tell, and he may have simply been replaced as alférez as early as Sept. 30, 1107, AHN, Códices, 989B, fol. 209r, by Pedro González. I suspect the latter document is to be relied upon.

28. See chapter 16, note 50. A subsequent notice is dated only to 1108, BN, 9.194, fol. 103r-v, in which Queen Beatrice is also cited.

29. See chapter 16, note 88.

30. See note 6.

31. Luis Salazar y Castro, Historia genealógica de la Casa de Haro, ed. Dalmiro de la Valgoma y Díaz-Varela, (Madrid, 1959), pp. 11-21.

32. Reilly, León-Castilla under Queen Urraca, pp. 286-87.

33. Ibid, pp. 300-306. At the time of the writing of that book I was unaware of the impact of the battle of Uclés itself on the subsequent development of royal authority in Asturias de Oviedo and in León.

34. See note 6.

35. Menéndez Pidal, España del Cid 1:50, n.1.

36. See note 27.

37. See notes 30 and 31.

38. Jiménez de Rada, "De Rebus Hispaniae," pp. 145-46. For what is know of Cidellus see Baer, A History of the Jews in Christian Spain, pp. 50-51.

39. That Pedro Ansúrez returned to León-Castilla as one of the few supporters of the marriage is borne out by the documents as well as the later histories. See my León-Castilla under Queen Urraca, pp. 35-37, 53, and 58-59.

40. Bruel, ed., Recueil des chartes 5:654-55, with the year given as 1079. All of the other internal evidence makes 1109 the only possible date.

41. Bishko, "The Cluniac Priories of Galicia and Portugal," pp. 318-19.

42. AHN, Clero, Carpeta 892, no. 21; and Códices, 989B, fol. 36r-v. The subsequent abbreviation of protocol of the document notwithstanding, Urraca's confirmation is indicated. The bishop's confirmation is perfectly clear.

43. See respectively, notes 23 and 42, and AHN, Códices, 989B, ff. 79v-80r. If the fuero of Henry and Teresa to the men of Azurara in Beira, Portugal, and datable only to the period 1109-1112, was issued in this year, it was probably executed at Alfonso's court as suggested by the confirmations of counts "Fernando" and "Pedro." No such counts are known for Portugal in this period. Pub. DMP 1-1:18-19; and 1-2:561-62. The text gives only the year 1102, which is impossible. Bishop Gonzalo appears in Coimbra and the couple's son Alfonso Enríquez confirms.

44. The materials are reviewed by Luiz Gonzaga de Azevedo, História de Portugal, vol. 3 (Lisbon, 1940), pp. 240-43, who comes to the conclusion that 1106 is the probable date by, it seems to me, largely explaining away the bulk of the evidence.

45. García Larragueta, ed., Documentos de Oviedo, pp. 339-41. See also the discussion in my León-Castilla under Queen Urraca, p. 76, n. 101.

46. ES 20:88-90. The bull of Paschal II, dated Dec. 4, 1108, Erdmann, ed., "Papsturkunden in Portugal," pp. 162-63, addressed to Gerald is a forgery.

47. David, Études historiques, pp. 454-55, cites a document of Dec. 22, 1108, in which Maurice was still bishop of Coimbra and one of Jan. 19, 1109, in which he was already archbishop of Braga. Almeida, História da Igreja em Portugal 1:264, says that Archbishop Gerald died on Dec. 5, 1108. However, a private document dated Mar. 9, 1109, DMP 2:183, still cites Maurice as bishop of Coimbra. The date should probably be ten years earlier.

48. See notes 40 and 41.

49. Carl Erdmann, Das Papsttum und Portugal in ersten Jahrhundert der Portugiesischen Geschichste (Berlin, 1928), pp. 14-16.

50. Mattoso, A nobreza medieval portuguesa, p. 169.

51. DMP 1-1:19-21.

52. See note 43.

53. Puyol y Alonso, ed., "Las crónicas anónimas," p. 247.

54. David, Études historiques, p. 301.

55. Reilly, León-Castilla under Queen Urraca, pp. 55-58.

56. González, Repoblación de Castilla la Nueva 1:100-102, details the events and their most probable order from the somewhat contradictory sources that relate them.

57. Puyol y Alonso, ed., "Las cronicas anónimas," pp. 120-21.

58. ES 20:115, "appropinquante sui transitus hora mihi apud Toletum Regnum totum tradidit." Jiménez de Rada, "De Rebus Hispaniae," p. 145, says that the official action occurred "in pago prope Toletum qui Magam dicitur," but he was writing better than a century after the fact. Magán is a hamlet twelve kilometers northeast of Toledo which may have been the site of the camp of the royal host.

59. See my León-Castilla under Queen Urraca, pp. 52-64, for a full consideration of the marriage and its chronology, especially p. 52, n. 26, on José María Ramos y Loscertales, "La sucesión del Rey Alfonso VI," AHDE, 13 (1936-4,): 36-99.

60. Boissonnade, Du nouveau sur la Chanson de Roland, p. 41.

61. José María Lacarra, ed., "Documentos para el estudio de la reconquista y repoblación del valle del Ebro," EEMCA 2 (1946): 481-82. April 3, 1108. AHN, Clero, Carpeta 711, nos. 2 and 3, which purports to be a charter of Alfonso I as "imperator in Castella, Gallicia," is a clumsy forgery, as the merest glance at the final protocol will demonstrate. Feb. 20, 1108. AC León, Códice 11, fol. 473r, is a private document that cites the Aragonese as already married to Urraca and ruling in León-Castilla. The date of this copy is obviously incorrect when compared with the bulk of the evidence.

Alfonso I of Aragón still needs a biographer. José María Lacarra, Vida de Alfonso el Batallador (Zaragoza, 1971), is but a preliminary sketch. Before a full biography can be done a critical edition of his charters is required.

62. For his age see chapter 2, note 21. Three different dates are given for his death by roughly contemporary sources. June 29 appears in the "Historia Compostelana," ES 20:96, and in the "Chronicon Compostellanum," ES 20:611. The two are independent. The "Chronicon Gothorum," David, Études historiques, p. 301, repeats this same date. June 30 appears in the "Anales Toledanos," Huici Miranda, ed., Cronicas latinas, p. 344, and in an inscription of a church, see Francisco Simon y Nieto, "El monasterio de San Salvador de Nogal," BRAH 35 (1899): 207-208. The date I follow is that given by two eyewitnesses, Puyol y Alonso, ed., "Las cronicas anónimas," p. 120, and Sánchez Alonso, ed., Crónica del Obispo Don Pelayo, p. 87. In this matter and a few others I have revised my own provisional judgments given in the introductory chapters of my León-Castilla under Queen Urraca, pp. 3-55, which materials should now be used with the appropriate caution.

63. See my León-Castilla under Queen Urraca, pp. 55-57, for the documentary evidence.