[102] We do not know when Diego Gelmírez was born. He was entrusted with important administrative responsibilities as early as 1093, was ordained priest and elected a bishop in 1100, and died at an advanced age in 1140. These indications, such as they are, would suggest that he was born about the middle of the 1060s. This is imprecise, but as near as we are ever likely to get. The Historia Compostellana contains two passages which shed a little light on his family background.⁽¹⁾ Diego was a native of Galicia and came of a respectable family (natus probis secundum seculum parentibus). His father Gelmirio could best be described as a miles -- a word whose connotations at this period it is not easy to grasp with any confidence, but which, from the pen of a Frenchman writing in Compostela in the 1120s, probably implied not simply the profession of arms but a certain social rank guaranteed by the possession of property and a certain code of behaviour appropriate to it. Gelmirio had been in the service of bishop Diego Peláez, from whom he had held the castle of Torres del Oeste, and on whose behalf had administered for many years the districts of Iria, La Mahía and Postmarcos; the land bounded on the south side by the river Ulla and the Ría de Arosa, and on the north side by the river Tambre and the Ría de Muros y Noya, between the town of Compostela and the Atlantic. Gelmirio -- the name was derived from the Germanic Hildemir⁽²⁾ -- was evidently a man of complete trustworthiness in the bishop's service. The area he governed was the heartland of [103] the honor, or temporalities, of Santiago and the castle of Torres del Oeste was of great strategic importance. He was probably of no very exalted social rank. Diego's panegyrists would have been unable to resist drawing attention to his parents' nobilitas, had they possessed it. As it was they could describe them only as probi --'respectable'. Gelmirio, then, was a man of the middling sort, in all probability a native of the area near Compostela, conscientious and competent in his ministerial role; a rising man, who might hope to do better by his children than his father had done by him. If they could be confident of a good start in life, they knew that they would still have to work hard to go far. It was perhaps from his family that Diego got his self-confidence, his ambition and his energy.

Diego had four brothers, Munio (or Nuño), Gundesindo, Pedro and Juan. Another brother, also named Pedro, is mentioned in the Historia: possibly he was a half-brother, the issue of a second marriage by Gelmirio. We do not know of any sisters, though it is possible that Diego's nephew, yet another Pedro, was a sister's son. Diego must early on have been destined for an ecclesiastical career. He received his education in the cathedral school of Compostela.⁽³⁾ Our ignorance about the Compostela school in the middle years of the eleventh century should not lead us to underestimate the quality of the education it could provide. Bishop Pelayo of León (1065-86) could boast of the education he had received there.⁽⁴⁾ Diego never became a man of very wide or deep intellectual cultivation, but he had been given a good grounding in the scriptures and in law, he became a patron of learning in a modest way, and by the time he reached middle age he had assembled a private library of at least fifteen volumes -- not a negligible collection.⁽⁵⁾ After his schooling at Compostela Diego spent some time at the royal court of Alfonso VI. The king's daughter Urraca referred to this many years later when she reminded Diego in 1116 that her father 'brought you up from your adolescence' (te ab ipsa educavit adolescentia).⁽⁶⁾ We can rarely be sure precisely what writers of this period intended by the term [104] adolescentia: here I think the implication is that Diego was attached to the royal household while he was in his teens. It would have been a time to watch and to learn, to make himself useful and agreeable to those whose interest might be enlisted in his favour. The sequel permits us to assume that he acquitted himself well. By the time he was 'grown-up' -- adultus,another vague term -- he was back in Galicia in the household (curia) of his patron bishop Diego Peláez.⁽⁷⁾ This must have been before the year 1085, when the Galician rebellion in which the bishop was implicated broke out. It was perhaps at this time that the young man was rewarded with a canonry in the church of Santiago.⁽⁸⁾

How Diego managed to survive the wreckage of his patron's fortunes we shall never know. We have no materials relating to his life between 1085 and 1090. One can think of more than one reason why he should have been reluctant, later on, to talk about this period or to let his biographers write about it. Speculation is fruitless, but it is fair to surmise that the young Diego saw his career at stake and fought hard to save it. In fighting hard he may not have been able to keep his hands clean. His extreme reluctance in later years to travel through the kingdom of Aragon, where Diego Peláez lived out his exile, might just possibly indicate that his anxiety not to see the man who had given him his first leg-up in his career was not untinged by guilt.

His great opportunity came after the appointment of Raymond of Burgundy to the county of Galicia. Apparently on the advice of the cathedral community of Compostela, he chose Diego as his chancellor and confidential adviser (cancellarius et secretarius).⁽⁹⁾ We cannot date this appointment accurately: it must have occurred between the time of Raymond's nomination as count of Galicia in the spring of 1087 and Diego's first spell as administrator of the diocese in 1093-4. There survive five charters given in Raymond's name which were drawn up under Diego's [105] direction.⁽¹⁰⁾ In these documents he described himself variously as the count's scribe or notary (scriptor, notarius). His service in the peripatetic household of count Raymond during these years broadened his experience to embrace activities beyond the confines of the merely secretarial. In 1094-5, for example, he accompanied the count's army on its campaign to Coimbra and Lisbon, in the course of which he had a narrow escape from the Saracens.⁽¹¹⁾ In later years he would refer to his travels as a young man: his attendance at the courts of Alfonso VI and then count Raymond would certainly have kept him on the move.⁽¹²⁾ As to his duties, we know so little of the count's chancery that nothing may be said beyond that they were concerned with the drafting and production of official documents. One generalization, however, may safely be made. During this period such service to the crown (or, as in Diego's case, to a prominent member of the royal family) was normally rewarded with ecclesiastical preferment. We can see this exemplified in Germany, France and England, and it is probably only because no serious work has yet been devoted to the chanceries of Fernando I and Alfonso VI that we cannot yet see it in Spain.⁽¹³⁾

Diego's first substantial reward came in 1093 when Raymond appointed him administrator of the temporalities of the church of Santiago. We should remember that the bishopric had been vacant since the deposition of Pedro de Cardeña in 1090. By long-established custom the kings of León-Castile exercised what in England would have been called regalian right during such vacancies.⁽¹⁴⁾ That is to say, [106] they took the temporalities of the vacant see into their own hands and diverted the revenues therefrom into their treasury. Alfonso VI had first appointed Pedro Vimara to administer the temporalities -- the honor -- of Santiago. His stewardship proved so rapacious that he was deprived of office and replaced by Arias Díaz. According to the Historia Compostellana his administration proved no less oppressive than that of his predecessor: every rank of society groaned under his yoke, and so harsh lay his hand upon the canons of the cathedral church that they even went short of food and clothing. The death of Arias Díaz was a merciful release. It was at this point that count Raymond, having sought advice from the bishops of Lugo, Mondoñedo, Orense and Tuy, directed the elders (quosdam seniorum) and the people(populum) of the church of Santiago to submit a nominee whom he might appoint to the vacant post. At a meeting held under Raymond's aegis, which was also attended by members of the Galician nobility (principes) the unanimous choice was Diego Gelmírez. We may make what we will of this. We are told that the electors recalled how just and moderate had been the administration of Diego's father Gelmirio in a similar charge. While we need not doubt this, we may also suggest that count Raymond played a rather larger part in the appointment of Diego than the Historia's account implies.

Diego administered the honor of Santiago for a year.⁽¹⁵⁾ At the end of it a new bishop was found for the see in the person of Dalmatius, a monk of Cluny who was at that time exercising supervision over the Cluniac houses of Spain.⁽¹⁶⁾ He was appointed, we are told, by Alfonso VI, count Raymond and the infanta Urraca; but they acted on the advice (consilio) of the clergy and people of Compostela, with the permission of abbot Hugh of Cluny, and with the approval of the Roman church. This last provision was most important: [107] Urban II had ceased attempting to bring about the restoration of Diego Peláez; Dalmatius was persona grata to both king and pope. But he did not last long. Towards the end of 1095 he attended the council of Clermont, and shortly after it acquired from the pope an important privilege exempting his see from the jurisdiction of any metropolitan and placing it directly under the authority of the church of Rome.⁽¹⁷⁾ But he died soon after his return to Compostela in March 1096.⁽¹⁸⁾ Once again, at the request of the clergy and people of Santiago. Diego Gelmírez was appointed by king, count and infanta to administer the diocese; by the summer he was describing himself as the vicarius of St. James.⁽¹⁹⁾ Meanwhile, the death of Dalmatius had been the signal for the ex-bishop Diego Peláez to make another bid to recover his see. He went to Rome to lay his case again before the papal court. No ruling was forthcoming, and hence no new episcopal appointment could be made. It was for this reason that Diego Gelmírez's second administration of the honor lasted for so long, about four years (1096-1100). Towards the end of this period Alfonso VI sent an embassy to Rome in an attempt to hasten a decision in the long-drawn-out suit. His messengers found Urban II dead (29 July 1099) and Paschal II elected pope in his place (13 August). The new pope swiftly brought matters to a conclusion. He ruled that Diego Peláez had been justly deposed and in a letter of 29 December 1099 ordered a new election to be held.⁽²⁰⁾ The tribulations of the see of Compostela were nearly at an end.

For the greater part, therefore, of the last decade of theeleventh century Diego Gelmírez had had charge of the temporalities of the see of Santiago de Compostela. He sometimes referred to himself as vicarius of the see;⁽²¹⁾ once [108] as honores sancti lacobi dijudicans;⁽²²⁾ once -- possibly -- as maiorinus et dominator Compostelle honoris.⁽²³⁾ The Historia Compostellana called him protector et defensor, elsewhere gubernator et dominus; and to his predecessors Pedro Vimara and Arias Díaz it referred respectively regius villicus and maiorinus.⁽²⁴⁾Most of these terms, 'lord', 'governor', 'protector' and the like, are imprecise. Only two of them may be regarded as technical terms denoting administrative office. In Spanish documents of this period villicus usually meant the steward of an estate. Maiorinus was the Latin form of the title rendered in the Romance vernacular as merino. At this period it was the task of the merino to administer justice and to enforce royal rights within a given area; he was, below the level of count, the principal royal official in local administration. Diego's vicariate of the lands of St. James would have involved him not only in the supervision of estate-management but also in the administration of justice, in providing for defence and peace-keeping, and in the collection of the king's public revenues. When we recall that in addition to these responsibilities Diego was, between 1093 and 1096 (at the least), a prominent member of count Raymond's secretarial staff, and that in the winter of 1094-5 he had gained experience of military action we can gauge the extent of his energies and talents and judge how well fitted he was to assume, as a bishop, responsibilities still wider.

The Historia Compostellana furnishes us with an elaborate account of Diego's election to the bishopric of Santiago, which yet contains some dark corners. We may suspect that this is deliberate. The general course of events is clear enough, their interpretation less so. It is fairly obvious that strings were being pulled behind the scenes, but we cannot be sure who was doing the pulling, nor in which direction. Narratives of episcopal elections in eleventh- and twelfth-century Spain are exceedingly rare: this is one of the fullest we possess.⁽²⁵⁾ We must see what we can do with it.⁽²⁶⁾

[109] When Paschal II's letter of 29 December 1099 reached the clergy of Compostela, Diego had already left -- for Rome. He had gone there on pilgrimage (orationis gratia), we are told, but it is difficult to believe that his journey had no other motive.⁽²⁷⁾ This suspicion deepens when we learn that at the papal curia he very probably met Guy, the archbishop of Vienne. Now Guy was the brother of count Raymond of Galicia. What more natural than that Raymond's client should have solicited the interest wielded by his patron's brother, a churchman of outstanding renown who was later to become pope as Calixtus II, on his own behalf at the curia?⁽²⁸⁾ We should bear in mind too that as a result of the papal privilege acquired by Dalmatius in 1095 the pope's concern with the Compostela election was an intimate one; for thenceforward the bishops of Compostela were to be 'subject to none except the metropolitan bishop of Rome . . . as special suffragans of the Roman see'. Certainly Diego succeeded in impressing the pope. Paschal ordained him subdeacon and sent him back with a letter to the canons of Compostela (18 March 1100) telling them with a characteristically scrupulous regard for the rights of electors that he regarded Diego as a fit person to be promoted to holy orders (ad sacros ordines).⁽²⁹⁾ Perhaps Diego would have liked the pope to have gone further. Nevertheless, the letter of [110] 18 March came pretty close to being a directive to the electors.

Meanwhile, at Compostela, the canons had prudently put off acting upon Paschal's letter of 29 December 1099 until Diego should have returned. He for his part hastened back armed with the papal letter of 18 March, and on 1 July 1100 he was -- though unwilling (!) -- elected bishop of Santiago de Compostela. Where was he elected, and by whom? The Historia implies that the election was held at Compostela, but never states it clearly -- which is a little odd, but not necessarily suspicious. Who precisely the electors were is hard to determine. The Historia offers us two separate and in some respects divergent accounts. The author of Book I, chapter 8 -- Nuño Alfonso, canon of Compostela and later bishop of Mondoñedo -- writing as a member of the cathedral chapter, states:⁽³⁰⁾

visis domini papae istis subsequentibus literis [i.e. the letter of 18 March 1100] eum nolentem, atque renitentem cum nobilioribus totius Gallaetiae, et assensu regis Adefonsi, et comitis Raimundi qui nobiscum laudantes aderant, in episcopum elegimus.

A rather different slant is given by the author of Book II, chapter 2.⁽³¹⁾ This was Gerald of Beauvais, a Frenchman who became a canon of Compostela about the year 1112 and who was writing his account about ten years later: his testimony [111] is therefore not that of an eyewitness or participant. He wrote as follows:

Cumque postea clerus et populus ecclesiae beati Jacobi eligeret sibi episcopum, consilio eiusdem regis A. et comitis Raymundi atque Gallaeciae principum praedictus Didacus electus est in episcopum. Prius tamen quam a laicali manu acceperat ecclesiam et honorem ecclesiae beati Jacobi, reddidit B. Toletano archiepiscopo et Romanae ecclesiae legato coram episcopis Gallaeciae qui interfuerunt electioni eius, videlicet D. Auriensi, P. Lucensi, A. Tudensi, G. Minduniensi, et principibus Gallaeciae, clero, atque populo Compostellae.

This is very tendentious stuff. It would be preposterous to believe that so skilled a political operator as Diego would ever have given so easy a handle as this to his arch-opponent of Toledo. The great interest of Gerald's account, however, lies in what it has to reveal about the movement of ideas at Compostela between the first decade of the century, when Nuño wrote, and the third, when Gerald was writing.⁽³²⁾ [112] Does the second passage shed any more light upon Diego's election? It confirms that the king, count and nobility were consulted before the election was completed. It gives us the additional and credible information that the Galician bishops were present too.

To sum up the foregoing, it looks as though Diego was the candidate favoured by the royal family. Given his previous career, this is hardly a matter for surprise. There might have been one or more rival candidates among the Compostelan clergy. Possibly there was a faction at Compostela which wanted Diego Peláez back. By a remarkable burst of initiative he succeeded in mobilizing papal interest in his favour. Every possible precaution was taken over the election, which in itself suggests that it might have been more of a touch-and-go affair than our reports of it (from the winning side only, we should remember) permit us to see. No doubt there were awkward moments for Diego on 1 July 1100. But, taken all in all, it was a triumph of shrewd management. The bishop-elect of one of the most famous and wealthy sees in Christendom must have retired to bed a happy man that night.

Diego was not consecrated as bishop until the following Easter Day, 21 April 1101.⁽³³⁾ Why was his consecration delayed for nearly ten months? By the terms of the privilege of 1095, reinforced by the papal letters of 1099, future bishops-elect of Compostela were to be consecrated by the pope alone.⁽³⁴⁾ However, Diego did not wish to travel through the kingdom of Aragon for fear of the ex-bishop Diego Peláez and his supporters (propinqui). He therefore besought the king to write to the pope with the request that Diego might be consecrated in the kingdom of León-Castile (apud nos). The king's letter was conveyed to the papal curia by two canons of Compostela, Hugo (later bishop of Porto) and [113] Vincent. It was evidently supported by a letter from three Galician bishops -- Pedro of Lugo, Alfonso of Tuy, and Gonzalo of Mondoñedo -- and the clergy of Compostela. Paschal II proved agreeable. In letters of 14 October 1100 he announced to the Galician clergy and the king that he had instructed bishop Godfrey of Maguelonne, in southern France, to go to Spain to consecrate Diego; or, if he could not do it, the ceremony was to be performed by bishop García of Burgos.⁽³⁵⁾ Having received these letters the two canons set out on the return journey. But then disaster struck. They fell seriously ill. Vincent died, and Hugo was laid up sick for a long time. When no word came back to .Compostela, further letters from king and clergy were sent to the pope, carried on this occasion by Nuño Alfonso (the author, as we have seen, of this section of the Historia) and Munio Gelmírez, the bishop-elect's brother. Paschal II's replies, renewing his permission, were issued on 25 March 1101. The envoys retraced their steps. In the meantime Hugo had recovered and made his way back to Spain, presumably picking up the bishop of Maguelonne on the way. It was at Godfrey's hands, doubtless assisted by the Galician bishops, that Diego Gelmírez received his consecration on 21 April 1101.⁽³⁶⁾ His long episcopate had begun.

The cathedral community at whose head Diego now found himself was probably a somewhat demoralized one. The see had been without sustained episcopal guidance for about fifteen years. It was committed to a very expensive building programme at a time when its two principal sources [114] of revenue were threatened: the king's agents had encroached upon its landed endowments, and we may assume (though we cannot prove it) that political disturbance within Spain -- of which more presently -- had discouraged pilgrims and thus reduced income from that quarter.⁽³⁷⁾ A clique of reformers had recently replaced an ancient and cherished liturgy with a new and strange one of which conservative Galician churchmen were suspicious. An assertive archbishop of Toledo who enjoyed the support of the pope was hostile to a Galician tradition of de facto autonomy in ecclesiastical affairs.

A less self-confident man than Diego might have been daunted. As it was, his boisterous energy and self-assurance sustained him. But more was needed than these qualities alone. To breathe new life into his church he had to make plain to friends and enemies alike what manner of man he was and what attitudes he would take up on the large questions of his day. He had to raise as it were a standard round which supporters might rally. He had to show men 'where he stood'. The key to an understanding of his career lies in his intense ambition for the glory of St. James. This would endear him to the community which clustered about St. James's shrine. But devotion alone, however single-minded, however lacking in conventional scruple, was not enough. By temperament and early experience Diego was a man of the world. He had the wit to see that his world was changing and the intuition to grasp how change might be made to serve his apostle's -- his own -- purposes. He would have to persuade, cajole or browbeat his colleagues into seeing it too, grasping it for themselves. He had to reconcile conservatives to change. He had to show his chapter that moderate reform need not alarm. These things will become clearer as we proceed. For the present, let two episodes from early in his episcopate stand as portents.

In the autumn of 1102 Diego carried out a visitation of the properties held by the see of Compostela to the south of the river Miño, in the county of Portugal and the diocese of Braga.⁽³⁸⁾ There were many of these, but we need concern [115] ourselves only with one -- the church of St. Fructuosus (now São Fructuoso de Montelios) on the outskirts of Braga, built by the saint in the seventh century, harbouring his relics after his death, and granted to the church of Santiago in 883. The body of Fructuosus was especially dear to the people of Braga. The saint was their 'defender and patron' (defensor et patronus). Fructuosus was to Braga what James was to Compostela. Diego formed the plan of carrying off the body to Compostela -- and did so. It was a bold manúuvre, conducted so swiftly and secretly that there was not time for any serious opposition to it to take shape. Diego had anticipated opposition; not surprisingly, for though legally he may have been acting within his rights, morally speaking he was perpetrating an outrage.⁽³⁹⁾ The episode typifies his lack of scruple. It may have other lessons for us, but to these we shall return in a later chapter.

Two years later Diego paid his second (and last) visit to Rome. It was almost certainly during the course of this journey that he met an archdeacon of Lucca named Gregory distinguished for his legal interests. Some years later, probably in 1110, pope Paschal II promoted Gregory to the college of cardinals. It was as cardinal-priest of S. Crisogono, at some point before his death in 1113, that Gregory published a canon-law collection known to scholars as the Polycarpus, which he dedicated to Diego Gelmírez.⁽⁴⁰⁾ No doubt Gregory had been collecting and sifting his materials for years: but it needed, as he politely explained in the dedicatory letter, the urging of Diego to get him to sit down and knock them into shape. Diego had realized the need for an up-to-date manual of canon law in Spain. This was a characteristic concern of ecclesiastical modernizers: [116] we might compare the introduction of a new corpus of canon law into England by archbishop Lanfranc and his circle a generation earlier. Up-to-date the Polycarpus certainly was. Gregory drew heavily upon the compilation of Anselm of Lucca and the Collection in 74 Titles, the two most authoritative collections of the recent past, and he included pronouncements of Gregory VII, Urban II and Paschal II, and conciliar material from as late as Piacenza (1095). The scope of his book was comprehensive and its orderly arrangement made for ease of reference. While seven of the eight books into which the work was divided were generally conservative in tone -- they owed much to the Decretum of Burchard of Worms -- the first book, which was devoted to the primacy and special rights of the see of Rome, was strongly papalist. Diego's request for this law-book is one of several indications that he was firmly on the side of change, of ecclesiastical reform.

For Diego, therefore, the management of the church of Santiago de Compostela presented challenges and opportunities. These must be set against a background composed from the public events which were preoccupying the peoples of León-Castile during the first decade of the twelfth century. This was one of the grimmest periods in the kingdom's history. There can have been few among the subjects of Alfonso VI who could, without deceiving themselves, have looked to the future with any optimism. To understand this state of affairs we must retrace our steps a little.⁽⁴¹⁾

On 6 May 1085 Toledo had surrendered to Alfonso VI, and on 25 May he made his formal triumphal entry into the city. Coincidentally, 25 May was also the date of the death, in exile, of pope Gregory VII whose thinking may have [117] influenced Alfonso's decision to lay siege to Toledo in 1081. The ancient capital of the Visigothic kings was Alfonso VI's most famous conquest. To characterize it in this way is well enough, provided we do not fall into the error of supposing that it crowned a career devoted to conquest since his accession in 1065. Toledo marked a new departure. The city was the nucleus of a principality, the kingdom of Toledo, which was one of several successor-states to the defunct caliphate of Córdoba. These small Muslim principalities are known to historians as the reinos de taifa, the 'party' or 'faction' kingdoms. The most important of the remaining taifa kingdoms were Zaragoza, Albarracín, Valencia, Badajoz, Seville, Granada and Almería. Rich, factious and vulnerable, the taifas had fallen prey to the diplomacy of the Christian rulers of northern Spain during the mid-century years. Their subjection had been expressed in the annual payment of very large sums of gold by way of tribute. These tributes were known as parias. The influx of gold into León-Castile and Catalonia -- for Aragon and Navarre never profited to such an extent as these two -- had far-reaching effects. It made the kings of León-Castile very rich. The income from parias was probably the most important source of their public revenues in the second half of the eleventh century. Some of this wealth was passed on to the churches within their kingdoms, for example to Fernando I's foundation of San Isidoro de León. Some of it found its way into the coffers of their aristocracy. The paria regime made possible the spectacular careers of adventurers such as Rodrigo Díaz, el Cid. Some of the riches of the Islamic south found their way across the Pyrenees to more distant beneficiaries, notably to the abbey of Cluny, as we saw in chapter II.

Why it was that Alfonso VI decided to kill one of the geese that laid these golden eggs by conquering Toledo has never yet been satisfactorily explained. But if the reasons for this shift in policy remain unclear, its consequences are plain. Alarmed, the rulers of the taifas of Seville, Granada and Badajoz determined to seek help from across the Straits of Gibraltar against the Christian king. After some hesitation and with misgivings which, as events turned out, were to be only too well-founded, they entered into negotiations [118] with Yusuf ibn-Tashufin the Almoravide ruler of Morocco. The Almoravides were adherents of an Islamic sect which had originated in the basin of the Senegal in the 1040s, had spread like wildfire among the Berber tribes of northwest Africa known as the Sanhaja (the ancestors of the Tuareg), and had thrown up a precarious state which was to dominate the area for a couple of generations. The growth of Almoravide power was itself an indirect result of the collapse of the Spanish caliphate. One of the main aims of Córdoba's foreign policy in the previous century had been to prevent, by careful distribution of subsidies, the emergence of any strong political unit in north-west Africa. In fact, the bases of Almoravide power were ill-founded. But this was not apparent in 1085. The rulers of the taifas knew the fame of the Almoravide armies. They also knew that the Almoravides were uncouth, uncultured fanatics. They did not like the sound of them, but they needed their help.

Yusuf accepted their invitation. He crossed the Straits with a large army in June 1086 and marched by way of Seville towards Badajoz. Alfonso VI was at the time engaged in besieging distant Zaragoza. He hastened across Spain and offered battle at Sagrajas, a little to the north of Badajoz. There, on 23 October 1086, Yusuf inflicted a resounding defeat upon him. Then, obedient to his agreement with his employers, Yusuf took his army back to Morocco. We must beware of exaggerating the impact of Sagrajas. For Alfonso VI it was humiliating, but not disastrous; it inspired anxiety, but not panic. One of the expressions of that anxiety was the king's appeal to France -- specifically to Burgundy, his wife's homeland -- for military aid. One of the results of this appeal was the coming to Spain in 1087 of duke Eudes of Burgundy and his cousin Raymond. But the campaign anticipated never took place, for there was no enemy to fight. The Almoravides remained in Morocco throughout the year. Alfonso VI and his French reinforcements frittered away their time in an abortive siege of Tudela in southern Navarre. In the following year, 1088, the Almoravides did come back, but they did not repeat their success of two years before. They besieged the castle of Aledo, between Murcia and Lorca, one of the most southerly of Alfonso's outposts, [119] for four months, but the king managed to relieve it and Yusuf returned to Africa. Alfonso's trusted lieutenant Alvar Fáñez was collecting parias as usual from Seville and Granada in the autumn. The year 1089 was quiet, but in 1090 Yusuf returned to Spain and laid siege to Toledo itself for about a month during the summer. It was successfully defended by Alfonso VI with the aid of king Sancho Ramírez of Aragon.

On this occasion Yusuf did not go back to Morocco. His dealings with the taifa kings had shown him how great was the gap between his religious ideals and theirs; had taught him too how precarious was their power. He claimed furthermore that they were not playing fair by him, but negotiating behind his back with the very man against whom they had engaged him to fight, the king-emperor of León-Castile. He did not dally. On his return from Toledo he occupied Granada and Malaga. In March 1091 his troops took possession of Córdoba, in May it was the turn of Carmona, then Seville was laid under siege. Alfonso VI sent an army under Alvar Fáñez to relieve the city, but it was defeated at Almodóvar del Rio, and Seville fell to the Almoravides in September. Almería went the same way, then Murcia. In 1092 Aledo was lost. Valencia would have followed had it not been for the prompt action of the Cid who hastened over to frighten away an Almoravide force and lay siege to the city himself. Of the three rulers who had summoned Yusuf to their aid in 1086 the only one now remaining in power was al-Mutawwakil of Badajoz. In desperation he turned to Alfonso VI, ceding to him the cities and associated territories of Santarem, Lisbon and Cintra in the spring of 1093. But when in the following year the Almoravides attacked Badajoz they conquered it so quickly that Alfonso VI had no time to gather an army to go to its relief. They went on to overrun the lands ceded to him in the preceding year. Raymond's Portuguese campaign in the winter of 1094-5, referred to earlier in this chapter, was an abortive attempt to recover them.

The only bright spot in 1094 was the Cid's conquest of Valencia in June and his defeat of an Almoravide army sent against him at Cuarte in October. The next two years were [120] a little more cheerful. Alfonso VI regained Santarem in 1095, and Pedro I of Aragon took Huesca in 1096. But in 1097 Yusuf, who had been absent from Spain for some years, returned to take command of his armies. He made for Toledo, but was intercepted by Alfonso VI at Consuegra where he defeated the Christian forces. Discouraged, however, from persevering against Toledo he sheered off to the north-east where he encountered another army under Alvar Fáñez near Cuenca: this too he defeated. In 1099 Toledo was once more attacked, and its surrounding territory ravaged. In the same year the Cid died (10 July) and the Almoravides laid siege to Valencia. Alfonso VI was diverted in 1100 from its relief by yet another attack on Toledo; later in the same year his son-in-law Henry of Portugal was defeated by the Almoravides. In the spring of 1102 the king supervised the evacuation of Valencia: Jimena, the Cid's widow, retired to end her days at the monastery of Cardeña, and Jerónimo, his bishop of Valencia, was translated to the Leonese diocese of Salamanca. The city was occupied by Yusuf's armies in May. Later in the year (or possibly in 1103) they took over the taifa kingdom of Albarracín. The only taifa state now remaining was Zaragoza.

In 1104 Alfonso VI captured Medinaceli. Perhaps heartened by this the Castilians dispatched a raid into Andalusia in the spring of 1105; but their troops met defeat at Almoravide hands on their way home. In September 1106 Yusuf died and was succeeded by his son Ali (1106-43). After suppressing a revolt in Morocco he crossed to Spain early in 1107, eager to continue the war. The heaviest blow fell in 1108. In May Ali's brother Tamim captured the town of Uclés and laid siege to its castle. A Castilian relief force went to its aid. Tamim inflicted a decisive defeat upon this force. As a result of the battle a number of strong points near Toledo fell into Almoravide hands, including Alcalá de Henares. In the following summer the old king died (30 June 1109). During the period of political turbulence which followed his death Ali launched a further attack on Toledo. Talavera and Madrid were sacked; so was the monastery of San Servando outside the walls of Toledo; the city itself was closely invested for two weeks and heroically defended by [121] Alvar Fáñez and archbishop Bernardo. In May 1110 Zaragoza finally passed into Almoravide hands.

The period between 1091 and 1109 was one of grave peril for the kingdom of León-Castile. It was not simply that Alfonso VI's conquests between the Duero and the Tagus were threatened. The Almoravides had re-united Islamic Spain as in the days of the caliphate. They were fanatics who believed that it was their mission to extend the reach of the faith by means of holy war. They were persecutors of the Christian communities which came under their rule. Alfonso for his part was so harassed by the constant necessity to look to his southern frontier that he had little time to devote to the pressing need for the resettlement (repoblación) of the conquered territories. In consequence his most imposing conquest, Toledo, remained a vulnerable outpost in a virtual desert. He was short of manpower not only for resettlement but also for his armies. The papal letters of October 1100 and March 1101 contained orders that Spanish soldiers and clergy were not to go to the Holy Land -- presumably the reference was to the crusade of 1101 -- but to defend their own land against the Almoravides: we may take it that the pope's prohibition was prompted by the king.⁽⁴²⁾ Finally, of course, the absorption of all the taifa kingdoms by the Almoravides had cut off the supply of parias. A savage blow was thereby struck at the health of Castilian public finance, whose consequences were to dog the crown throughout the last years of Alfonso's reign and for the whole of his daughter's.

Who was to succeed Alfonso VI on the throne of Leon-Castile? This question dominated the internal political life of the kingdom during Diego Gelmírez's early years as bishop of Santiago de Compostela. The brother, García, whom Alfonso had long before dethroned died in prison in 1090: there were no other close male relatives. Although Alfonso VI was much married he never succeeded in fathering a legitimate male heir. When Raymond of Burgundy came to his court in 1086-7 the king had three daughters. The only legitimate one was Urraca, the offspring of queen [122] Constance of Burgundy: as we have already seen, she was betrothed to Raymond in 1087 (and possibly married to him in the same year -- we do not know the exact date of their marriage). By a mistress named Jimena Múñoz he had already had two other daughters, Elvira and Teresa. The former married Raymond IV of St. Gilles, count of Toulouse, before July 1094. Teresa, as we know, married Henry of Burgundy, count of Portugal, perhaps in 1095 (again, we have no sure chronology). If we can believe the evidence of a set of annals transmitted in the earliest manuscript of the Historia Compostellana, the relevant part of which seems to have been composed in or soon after 1126, Raymond was permitted to entertain great expectations. 'King Alfonso had caused him to come from Burgundy to Spain and had promised him all his kingdom with a sworn oath.'⁽⁴³⁾

In 1091, however, an event occurred whose consequences were to unsettle Raymond's hopes. When in March the city of Córdoba fell to Yusuf its previous governor Fath al-Mamun was killed in the fighting. He left a young widow, Zaida, who took refuge with her father-in-law al-Mutamid, ruler of the taifa of Seville. The Almoravides invested Seville in June and took it in September. At some point in the year, perhaps when Alvar Fáñez was in Andalusia vainly trying to relieve the city, Zaida made her way as a refugee to the court of Alfonso VI. Later legend has been at work on this encounter. We hear how her plight moved and her beauty bewitched the elderly king-emperor. A more prosaic version has it that her liaison with the king brought him a string of fortresses on his southern frontier; but this story too has difficult features. All that matters for us is that, probably in 1093, she bore the king a son, who was christened Sancho. The existence of Sancho profoundly altered the situation. Alfonso doted upon the son of his old age, and we may take it that as he grew up he showed signs of promise. By 1102 his name featured among the subscriptions to royal diplomas. In 1105 and 1106 his name headed the witness-lists, coming [123] immediately after the royal subscription, above those of Raymond and Urraca. By 1107 it was clear that he was the heir apparent. We also know that he was given charge of Toledo, the new urbs regia of Alfonso VI (as of his Visigothic predecessors), the site of the principal royal treasury and, threatened as it was, the symbol of Alfonso's Christian imperium.

What was the king's design? We cannot be sure. But he may, like his father Fernando I before him, have been thinking in terms of partition. The regnum which Raymond had (perhaps) been promised could have been Galicia, with the attached territories of Salamanca, Zamora and Avila which he was busily trying to resettle during these years. If the regnum of Toledo was to go to Sancho, perhaps Castile was to go with it. Who then was to get León? Perhaps Henry of Portugal. He certainly had designs on León at a later date (in 1110-12), and a recently discovered charter of 1105 has been interpreted as indicating the existence of such designs as early as that date. Should this have been so, Henry's lands would have been awkwardly cut into two, a county of Portugal and a kingdom of León separated by Raymond's Galician and Extremaduran territories.

Where so much remains uncertain one thing is clear. Count Raymond resented Sancho's claims and was apprehensive of Henry's ambitions. The question of the Spanish succession began to involve other interested parties outside the Peninsula. Raymond turned for assistance to abbot Hugh of Cluny, who sent as his envoy to Spain one Dalmatius Geret, a native of Burgundy whose family was well known to the count. Cluniac interests could not but be involved in the succession question. In simple material terms Cluny depended on the rulers of León-Castile for a large proportion of her annual income. With the supply of money now diminished, if not cut off entirely, owing to the expansion of the Almoravides at the expense of the taifas, the chill winds of economic stringency were blowing about the vast and as yet unfinished building programme initiated by abbot Hugh largely on the strength of the Alfonsine subsidy. (Urban II had consecrated the new abbey church of Cluny in 1095, but the works were still far from complete.) Cluniac finance [124] was bound up with political stability in León-Castile, the essential pre-condition for Christian initiatives against the Muslims. What exactly the abbot commissioned Dalmatius Geret to do we cannot tell. All we have is a document recording the results of his mission. This takes the form of a letter addressed by Raymond and Henry to abbot Hugh, carried to him by Dalmatius. The letter is undated, and none of the several attempts to date it has been altogether convincing: my own view is that it belongs to the years between 1105 and 1107.⁽⁴⁴⁾ It records a secret agreement between the cousins relating to the disposal of the kingdom after Alfonso VI's death. With its details we need not concern ourselves. Its most important provisions were, first, that Henry would assist Raymond to succeed to the throne of León-Castile against all other claimants; second, that they were to share the contents of the royal treasury at Toledo between them in the proportion of one-third to Henry and two-thirds to Raymond; and third, that Henry was to hold certain territories (either Toledo or Galicia) from Raymond as vassal from king. Although the infante Sancho was not named in the letter its provisions were clearly directed against him. In short, Raymond sought to buy off Henry in order to ensure his own succession to the undivided kingship of León-Castile. It is important to note that Raymond now had a male heir. His wife Urraca had given birth to a boy on 1 March 1105. The child was christened Alfonso. This was Alfonso Raimúndez, later to be the emperor Alfonso VII (d. 1157). Henry and Teresa had as yet no son.

This was the position in the summer of 1107. (We know that Dalmatius Geret was back in Burgundy by 13 August, which provides a terminus ante quem.) Quite suddenly, all was changed. Raymond, who seems not to have been a healthy man, was taken ill and died, probably on 20 September. The immediate result of Raymond's death was the holding of a council at León in December to review the question of the succession. This was attended not only by [125] the lay and ecclesiastical notabilities of Alfonso's realm but also by archbishop Guy of Vienne, the late count's brother and the friend, as we have seen, of Diego Gelmírez. The Historia Compostellana supplies us with an account of what took place at León, but confines itself to the provisions that were agreed upon in the event that Urraca should marry again. After her second marriage in 1109 these provisions became of critical importance. But unless the marriage that finally took place was already being discussed at León in December 1107 -- which is unlikely -- it is clear that the arrangements reported by the Historia were only a small part of the deliberations which occurred there. In the absence of other narrative sources it is tempting to use the evidence of the Historia as a guide to what transpired at León. This temptation should be resisted. Instead we should look to the strictly contemporary and less controversial evidence of the royal charters. If we can discover how Raymond's widow styled herself in official documents we may have some clue as to the decisions reached that winter. Three charters granted by Urraca, to different beneficiaries, have survived from the period between Raymond's death and Alfonso VI's. (None is an original.) The earliest is dated 13 December 1107 and may pre-date the León meeting. In it Urraca styled herself totius Gallecie domina; in other words, she assumed that she had succeeded her husband as lord -- or lady -- of Galicia. The next was issued on 21 January 1108, after the council. Urraca styled herself tocius Gallecie imperatrix. (It would be imprudent to read too much into the more exalted title. Raymond had styled himself imperator of Galicia in a charter of 17 March 1107.) The conclusion I would draw is that Urraca's rule in Galicia had been confirmed by the king at the assembly in León. Finally, in a charter of 22 February 1109 she styled herself once more totius Galletie domina. But this is not all. The charter of 21 January 1108, a grant to the cathedral church of Lugo, was made

for the soul of my husband the lord duke [sic] Raymond who lived but a short time with me and departed from this frail world before me, and for me for whom a like fate awaits, and for my son Alfonso that he may live [reading vivere for iure] and rule in happiness.

This is little enough to go on. Still, it looks as though such changes as were necessitated by Raymond's death were kept to a minimum. Sancho, presumably, remained the king's heir apparent. Urraca was to keep Galicia and her son Alfonso was to exercise rule over some -- unspecified -- territories after her. Henry of Portugal had lost his ally but not, we may suppose, put aside his hopes.

At about the same time as the León council the king-emperor was stricken with the illness which was to prove his last. He was bedridden for the remaining seventeen months of his life. To what extent Alfonso was mentally as well as physically incapacitated is an important question to which we do not know the answer. If he had had a stroke he may well have been. His Jewish physician Joseph Ferrizuel, nicknamed Cidellus ('little boss'), seems to have played an important role as an intermediary between the king and his advisers during Alfonso's last illness.⁽⁴⁵⁾ It is not beyond the bounds of possibility that he played a major part in the decision-making of those last frantic months.

Frantic they certainly were. In the summer of 1108 the cruellest blow of all was struck. The Castilian army which was defeated by Tamim at Uclés in May was led by the young infante Sancho; and in the fighting he was killed. Alfonso VI, old and mortally ill, had lost his son and heir. The whole matter of the succession had to be reconsidered. The Almoravide enemy was at the gates. Time was short. Our sources remain meagre, allusive and untrustworthy. Valiant attempts to unravel a coherent story have been made, but they carry conviction only when the reader is lulled into suspending his very proper belief that bricks cannot be made without straw. It is clear enough that there was panic, intrigue and divided counsel. It is also clear that a plan, attributed to the king, finally emerged. Daring in its simplicity, it was also daring in its recklessness. León-Castile needed [127] a king of proven military experience. The man chosen was the ruling monarch of neighbouring Aragon, Alfonso I, known to later generations as el Batallador, 'the Battler'. He had succeeded his half-brother Pedro I in 1104, and was by now a man of about thirty-six. He was unmarried and therefore available as a husband for the widowed Urraca. He enjoyed a great and deserved reputation as a soldier. In these respects, he was exactly the man to be the saviour of León-Castile. As against this, he was not well liked among the aristocratic circles of León and, especially, Castile; and in the eyes of churchmen he was barred from marrying Urraca by consanguinity, for the pair shared a great-grandfather in Sancho el Mayor of Navarre. But Alfonso VI, or the faction which acted in his name, was undeterred. The Aragonese marriage project went ahead. The precise terms of the arrangement are, it need hardly be said, unclear. Urraca claimed in 1110 that 'not long before his death my father the emperor Alfonso handed on to me the whole kingdom (regnum totum tradidit)'. In the earliest surviving official document issued by the queen, a charter dated 22 July 1109 of which the original survives, she styles herself totius Yspanie regina; which would seem to bear out her claim.⁽⁴⁶⁾ Did Alfonso VI design that Alfonso el Batallador should be simply a king consort, his role primarily a military one, without independent political power, though doubtless with some territorial endowment? It is not impossible. But we simply do not know. We do not even know for certain when the marriage took place, though it was very probably in the early part of October. This is soon enough after Alfonso VI's death on 30 June for us to be confident that it had been planned while the old man was still alive.

Conceived and rushed through by frightened men in a hurry, the Aragonese marriage raised more problems than it was intended to solve. What would be the attitude to it of the bishops? What provision, if any, had been made for [128] Henry of Portugal? Or for the young Alfonso Raimúndez? How would Pedro Froílaz de Traba, the boy's guardian, and the rest of the Galician nobility react to it? Diego Gelmírez's career as bishop of Santiago de Compostela was to be intimately affected by the way in which these questions were gradually to be answered over the next several years.

1. HC, pp. 19-20, 254.

2. Holders of the name may be traced from the Suevic and Visigothic period onwards: an archpriest of Orense of this name attended the Third council of Toledo in 589; a ring inscribed ILDIMER, apparently of the sixth century, is preserved in Pontevedra Museum; Gildimirius subscribed a charter of c. 690; Ildemirus the priest subscribed another of 818; and the name is not infrequent in documents of the tenth and eleventh centuries.

3. HC, p. 254.

4. R. 10 November 1073.

5. See below, Appendix D.

6. HC, p. 212.

7. HC, p. 254.

8. HC, p. 20.

9. HC, p. 20. Diego claimed later on that he had been Raymond's confessor as well (HC, p. 458), but this could have been at any time before Raymond's death in 1107.

10. R. 13 November 1094, 25 February 1095, 24 September 1095, 1 May 1096 and 21 August 1096. The penultimate of these is incorrectly dated 1106: Reilly, Urraca, p. 23, n. 42, has proposed the emendation to 1096.

11. HC, p. 360. R. 13 November 1094 and 25 February 1095, both drawn up by Diego for Portuguese beneficiaries, were granted during this campaign, probably in or near Coimbra.

12. HC, p. 474.

13. But see the study of the career of Sampiro, royal notary and finally bishop of Astorga (d. 1042) in J. Pérez de Urbel, Sampiro: su crónica y la monarquía leonesa en el siglo X (Madrid, 1952), pp. 11-1 25. For some slightly later examples from León, see Episcopate, pp. 80-3.

14. R. A. Fletcher, 'Regalian right in twelfth-century Spain: the case of archbishop Martín of Santiago de Compostela', Journal of Ecclesiastical History 27 (1977), 337-60.

15. R. 28 January 1090 was subscribed by Diego Gelmírez as maiorinus et dominator Compostelle honoris. This has suggested to some authorities that he had been entrusted with the administration of the temporalities already at that much earlier date. This is possible but to my mind not very probable. The charter itself is not entirely above suspicion.

16. HC, p. 20. R. 13 November 1094 was subscribed by Dalmatius, thus providing us with a terminus ante quem for his promotion.

17. HC, pp. 21-3 (JL 5601): for further discussion of this privilege see below, ch. VIII.

18. The date of his death was probably 16 March 1096: see Vones, Kirchenpolitik, p. 100, n. 2. It is likely that he travelled from France back to Galicia by sea, since he can be traced at Saintes in the early weeks of 1096: Cartulaire de Saint-Jean d'Angély, ed. G. Musset (Paris-Saintes, 1901), I, no. clxxxiii.

19. R. 21 August 1096.

20. HC, pp. 25-6 (=JL 5810, 5811). Diego Peláez had already given up hope of reinstatement before the death of Urban II, for he was back in Aragon by July 1099 when he subscribed one of the charters of Pedro I of Aragon.

21. R. 21 August 1096, 19 May 1097.

22. R. 28 March 1098.

23. R. 28 January 1090 - though this charter (as we have seen) presents problems.

24. HC, pp. 18, 19, 23.

25. For others, see Episcopate, pp. 77.

26. Vones, Kirchenpolitik, pp. 100-48, provides what is now the fullest investigation of Diego's election and consecration. The work reached me after this chapter had been drafted. Dr Vones displays great ingenuity in his handling of the sources for these mysterious episodes, but in the last resort I find him not quite persuasive. However, certain of his emphases are valuable, and I have introduced changes into my text in deference to them. Our final positions remain a certain way apart. The focus and scale of his book permitted treatment at considerable length, and the reader who seeks a thorough airing of these questions should turn to it.

27. HC, p. 26. There is no mention of him in R. 16 January 1100, an important grant by Alfonso VI to the church of Santiago, which suggests that he had by then already departed. The main purport of Dr Vones's argument is that Diego had already been elected and consecrated a bishop: this I cannot accept.

28. We cannot be absolutely certain that this meeting took place, but it looks likely. On one of his two visits to Rome (1100, 1104) Diego met Guy: HC, p. 272. Guy is not known to have been in Rome in 1104. Further, Guy had been appointed papal legate to England in 1100 and it is fair to assume that he visited the pope to be briefed for his (in the event, abortive) legation.

29. HC, p. 27 (=JL 5822). The letter hints interestingly but obscurely at some domestic opposition to Diego among his cathedral colleagues. Dr Vones is sceptical of its authenticity.

30. HC, pp. 26-7.

31. HC, p. 255.

32. This takes us to the kernel of my unease about Dr Vones's interpretation of these events. Where he sees Diego's election and consecration as having been highly irregular, and the authors of the Historia Compostellana as trying artfully to conceal this, I would see an electoral procedure of a type fairly common in the kingdom of León-Castile at the turn of the century but which churchmen of a reforming temper might later find it embarrassing to contemplate in retrospect. Ironically, it was Diego who was more responsible than any other agent for encouraging this movement of ideas. As for the authors of the Historia, I have elsewhere described them as 'more muddled than cunning'; and would hold to that opinion. See my review of Dr. Vones's book in the American Historical Review 86 (1981), 1084-5.

33. HC, pp. 27-31 (=JL 5839, 5840, 5860, 5861).

34. The clergy of Compostela's previous metropolitan, Braga, may have put up a fight against this. Vones, Kirchenpolitik, pp. 135-47 has some useful observations on this subject, though I would not go so far as he does in drawing conclusions from the very meagre evidence available.

35. For the bishopric of Maguelonne's connections with the papal see, JL 5375, 5377 (of 1088). By the terms of JL 5653 the see of Burgos had been granted the privilege of exemption in 1096.

36. We are not told who consecrated Diego. I assume that it was Godfrey of Maguelonne (1) because the letter instructing him to do so (14 Oct. 1100) does not survive among the records of Compostela, i.e. it was presumably kept by its addressee; and (2) because the second letter sent to him (25 Mar. 1101) was preserved at Compostela, i.e. the second group of envoys presumably found that Godfrey had already set out on his mission when they reached Maguelonne and went on their way with the letter. It is impossible to believe that the second group could have travelled from Rome to Compostela via Maguelonne and that the arrangements for an episcopal consecration could have been made in the short space of time between 25 March and 21 April. The only possible conclusion is that Hugo had recovered from his illness and completed his mission, as suggested in the text.

37. Building operations had been temporarily suspended (as we shall see in Chapter VII) but the programme itself had not been abandoned.

38. For what follows see HC, pp. 36-42. For the literary genre to which this account belongs see P. J. Geary, Furta Sacra: Thefts of relics in the central middle ages (Princeton, 1978). The legal background to the episode is investigated thoroughly in Vones, Kirchenpolitik, pp. 219-59.

39. It still arouses strong feelings. 'Diogo Gelmires, para roubar o corpo do santo, que os fiéis veneravam como patrono, foi o primeiro a violar-lhe o túmulo que os próprios Arabes haviam respeitado': A. de J. da Costa, O bispo D. Pedro e a organização da diocese de Braga (Coimbra, 1959), p. 19.

40. All previous work on the Polycarpus has been superseded by U. Horst, Die Kanonessammlung Polycarpus des Gregor von S. Grisogono. Quellen und Tendenzen (Munich, 1980), to which invaluable work, a copy of which Dr Horst most kindly sent me, I am indebted for what follows.

41. Some narrative of political and military events is unavoidable here of Diego's later doings are to be rendered intelligible, but I have kept it to a minimum. By opening the following account in the year 1085 I start in the middle of the story, with all the attendant risks that this entails. It is no part of my purpose, neither can I spare the space, to survey, however briefly, the complex sequence of events. which had occupied the previous half-century or so since the extinction of the caliphate of Córdoba in 1031. The essential background may be found -- to name only recent works in English -- in J. F. O'Callaghan, A history of medieval Spain (Ithaca/London, 1975), chs. 5 and 8; A. MacKay, Spain in the middle ages, from frontier to empire, 1000-1500 (London, 1977), chs. 1 and 2; D. W. Lomax, The reconquest of Spain (London, 1978), chs. 2 and 3.

42. See especially HC, pp. 88-9 (=JL 5863).

43. HC, p. 611, a statement which like others relating to the succession must be regarded as highly tendentious. The question of the succession to Alfonso VI has attracted much attention. The most recent treatment is that of Reilly, Urraca, chs. 1 and 2, with references to all earlier literature.

44. This was the dating favoured by Pierre David, 'Le pacte successoral entre Raymond de Galice et Henri de Portugal', Bulletin Hispanique 50 (1948), 275-90: though it is not without its difficulties it seems to me the most plausible of the various possibilities.

45. Y. Baer, A history of the Jews in Christian Spain (Philadelphia, 1961), pp. 50-1, 68 and references there cited. But note the cautions of Reilly, Urraca, p. 52, n. 25.

46. HC, p. 115, AC León, no. 1002 is I believe the original of R. 22 July 1109 (though Professor Reilly regards it as a closely contemporary copy): the diploma is notable for its particularly impressive witness-list, which included the archbishop of Toledo and ten other bishops, the abbots of three important monasteries, and a bevy of high-ranking laymen such as Alvar Fáñez, Pedro Ansúrez, Gómez González, Froila Díaz, Pedro Froílaz de Traba and others.