(This article appeared originally in Cuadernos de Historia de España 47 (1968), 31-135 and 48 (1969), 30-116 and appears in LIBRO with the kind permission of Dna. María Estela González de Fauvre, editor of the journal.)
 

The decade 1068-1077 sees the emergence into full documentary attestation of two rival alliance systems in Christian Iberia, each [54] closely linking a peninsular monarchy with a foreign ecclesiastical center: Aragon and Rome, Leon-Castile and Cluny. In 1068, in Rome, Sancho Ramirez places himself and his Jacan kingdom under the feudal suzerainty of St Peter; in 1077 Alfonso VI, confronting Gregory VII's claims to secular authority over the Hispanic Empire, follows his father Fernando in seeking confraternity with Cluny, while reviving the lapsed census in doubled amount. Despite Kehr's judgment that "la historia exterior y interior de España esta determinada entonces y en adelante por la rivalidad entre los dos reinos de Castilla y Aragón", ⁽³³⁵⁾ little has been done to examine this conflict as a whole, particularly from the standpoint of the divergent policies followed below the Pyrenees by the Gregorian popes and the Cluniacs in their relations with the two opposed peninsular powers. Progress continues to be impeded by assumptions no longer tenable, at least in their crude traditional form, of the Burgundian origins of the Gregorian Reform and the supposed identity of doctrines and aims characterizing abbey and Holy See in the mid-11th century, despite the numerous studies that have shown how Cluny accepted royal and feudal control over ecclesiastical benefices, including the iglesia propia ; sought a heavily monastic renewal of discipline, spirituality and custom in close collaboration with kings, magnates and episcopate; and maintained friendly ties with the Emperor Henry IV throughout the struggle over Gregory VII's decrees on lay investiture. ⁽³³⁶⁾

In Spain the often clashing interests of pope and abbot can readily be seen, as David has demonstrated, in the crisis of 1077-1080, when Cluny's support for Alfonso VI's opposition to Gregory VII led the wrathful pope to denounce the Burgundian queen, Constance, the niece of the abbot of Cluny, and to demand immediate recall of the monk Robert, Hugh's adroit envoy to the Leonese-Castilian court. ⁽³³⁷⁾ Even earlier, in 1073, another Cluniac, Cardinal Gerald of Ostia, although a legate in papal service, seems to have been chiefly responsible for blocking Alexander II's plan for a pro-Aragonese -- and thus implicitly anti-Leonese-Castilian -- French crusade in Spain under the leadership of Count Ebles de Roucy.⁽³³⁸⁾

What is manifest is that during the half-century after 1077 papal protection of Aragon's independence and expansionist aims, and Cluny's championship of Hispanic imperialism and repeated interventions in Leonese-Castilian national affairs -- for example, in the recruitment of military aid after Zalaka, the negotiation of the successoral pact between Counts Raymond and Henry partitioning the Alfonsine Empire, the possibly decisive role of Abbot Ponce de Melgueil in the civil wars under Urraca over the Aragonese marriage -- rank among the basic factors then affecting Iberian political evolution. In its main lines this dual system of competing monarcho-ecclesiastical alliances is visible enough; but its origins remain profoundly obscure. How far back in time should the dichotomy be placed? If Fernando I's friendship with Cluny, dating from the visit of Frater Galindus around 1053, long precedes the Aragonese turn to Rome, the same cannot be said of  the king-emperor's conjunctio , when his admission to societas and institution of the golden stipend transform mere mutuus amor into authentic alliance. Here we confront the root problem: which of the two partnerships, the imperial union with Burgundy or Aragonese [55] vassalage to Rome, came first? Did the one arise in direct reaction to the creation of the other new, deeply feared combination?

It is from this standpoint that particularly close scrutiny requires to be given the celebrated though sparsely documented international military expedition conducted in 1064 against the Muslim stronghold of Barbastro in Baja Ribagorza.⁽³³⁹⁾ For if this enterprise, often taken to be the direct prototype of the First Crusade, can be identified as the work of Cluny or Rome, or of both, it should be possible to obtain from it insight into the activities of the two religious centers in the context of the Reconquista and of Iberian inter-state rivalries. Unfortunately, at the present time our knowledge of the Barbastro War is subject to more uncertainty than at any time since Reinhart Dozy in 1860, and more fully in 1881, rescued it from neglect on the basis of the notices in the chronicles of Ibn-Hayyan and Amato di Monte Cassino.⁽³⁴⁰⁾ The Dutch Islamicist, reasoning from the analogy of Clermont, and convinced that Ibn-Hayyan's reference to the "commandant de la cavalerie de Rome" proved the presence of the papal gonfaloniero Guillaume de Montreuil and Italo-Norman troops alongside those from France and Spain, considered papal sponsorship of the war was central to its understanding. ⁽³⁴¹⁾ The same thesis reappears in the far more detailed reconstruction of Barbastro first presented by P. Boissonnade in his well known work on the Chanson de Roland (1923) and further expanded without significant interpretative change in a long paper published in 1932 in the Revue des questions historiques.⁽³⁴²⁾

For Boissonnade the expedition against Barbastro is above all a French crusade, inspired by Cluny and launched through Cluny's persuasion by the papacy of Alexander II, the purpose of which is to preserve a hard-pressed Aragonese kingdom from imminent invasion and possible destruction at the hands of the Muslims, following Ramiro I's shattering defeat and death at Graus on 8 May 1063. ⁽³⁴³⁾ Graus, in this Hispanic prelude to the Palestinian gesta Dei per Francos, serves as an Iberian Manzikert, with King Sancho Ramírez -- like the legates of the Emperor Alexius Comnenus at Piacenza -- appealing in desperation for papal and Frankish succor; and Cluny responds without delay because of the abbey's long-standing "tendressa" for the slain Ramiro. In turn, Cluniac influence over the Gregorian papacy assures Alexander II's call for a holy war in Spain; the ties of blood and marriage linking Aragon to the great noble families of southern France - -e.g., the Aragonese king's [alleged] marriage with Félicie, the sister of Count Ebles de Roucy ⁽³⁴⁴⁾ -- intensify trans-Pyrenean enthusiasm for the papal summons; the attack centers upon the stronghold of Barbastro as Jaca's indispensable gateway to the lower valley of the River Cinca towards Zaragoza and the Ebro plain; and the Aragonese triumph cues its climax when the victorious crusaders [supposedly] subject the conquered city to the authority of Sancho Ramírez. ⁽³⁴⁵⁾

This still generally accepted version of the Barbastro War stands however in urgent need of revision. Boissonnade's naive  analogies with the First Crusade, unfamiliarity with the true motives and aims of the Reconquista at that time, and undue concentration upon the French leaders and their followers to the relative neglect of the Iberian, above all Catalan, forces involved, result in serious [56] distortions. Even more serious is his limited cognizance of the Hispano-Christian and Taifa military and financial context, the failure to orient the war properly in the peninsular political conjuncture of the third quarter of the 11th century, which leads him to focus the story so entirely upon Aragon's supposed critical need for help against the Moors as to ignore the equally vital Catalan and Leonese-Castilian factors. On the papal side, P. David has strongly attacked the hypothesis of Alexander II's involvement on three grounds, arguing: (i) that down to the Council of Mantua in June 1064 the pope was too deeply engaged in fighting against the anti-pope Cadalus to be able to launch the crusade and lacked the support of Cardinal Hugo Candidus, whose service as his legate in Spain cannot in fact be proved before 1067; (ii) that the four extant fragments of Alexandrine bulls relating to a plenary indulgence for combatants in a holy war and to protection of Jews in southern France against attacks by crusaders moving into Spain belong not in 1063-1064 but in 1073, when this same pontiff was definitely promoting a Frankish expedition across the Pyrenees under the leadership of Ebles de Roucy; and (iii) that Ibn-Hayyan's commander of the cavalry of Rome is merely the Christian generalissimo Duke Gui-Geoffroi of Aquitaine, a contention which refutes the Dozy hypothesis of Guillaume de Montreuil and an Italo-Norman force in the papal service. ⁽³⁴⁶⁾

For the present inquiry the immediate question is the validity of Boissonnade's belief in Cluny as the initiator of the pro-Aragonese crusade, a thesis accepted by Lambert, Defourneaux and others, but which four editions of Menéndez Pidal's La España del Cid between 1929 and 1947 found it possible to pass over in silence. ⁽³⁴⁷⁾ If with David we eliminate papal sponsorship of Barbastro, then the foreign ecclesiastical organizer that alone would explain the number and diverse geographical provenance of the French contingents has to be identified as Abbot Hugh, an alternative which David himself conspicuously fails to mention. ⁽³⁴⁸⁾ Yet, as we know, this is an unacceptable solution. Far from cherishing lifelong affection for the Burgundian monks, Ramiro I both before and after Tafalla (1043) emphatically abstained from encouraging the Hispani at Peña or from emulating his father Sancho el Mayor as Cluniac socius and benefactor. ⁽³⁴⁹⁾ In 1063, some twenty years of silence and indifference towards the abbey on Jaca's part, and a decade of Burgundian friendship with the Leonese-Castilian king-emperor which could only have been politically unwelcome to Aragon, make it impossible -- even on the dubious assumption of Cluniac enthusiasm for the holy war -- that the newly ascended Sancho Ramírez, in the crisis after Graus, would spontaneously and successfully solicit Cluny's aid against the Muslims. In this light, indeed, it becomes obviously desirable to re-examine, as cannot be attempted here, David's case against the possibility of contact between Jaca and Rome as early as 1063, whether through Hugo Cándidus or other intermediary, a position that postpones formation of the Aragonese-papal alliance until 1067, on the eve of Sancho Ramírez' vassalic submission to Petrine tutela ; and it may yet appear that Gregory VII's bull Apostolica sedes may not be as historically erroneous as Kehr believed in its attribution to Ramiro I of the commencement of his kingdom's ties with the papacy.

[57] There need be no doubt that Cluny played some sort of influential role in the course of the Barbastro War; what becomes patent is that this activity does not center in Aragon but in the Pyrenean kingdom's two powerful rivals -- to the east, the Catalan counties, notably Urgel; to the west, the Hispanic Empire of Fernando I. It will be advisable to commence with the Catalan zone, since what limited Cluniac documentation we have relates to this.
 

(i) Cluny and the Catalans in the Epoch of Barbastro. The obscure beginnings of Cluny's definitive establishment in Catalonia, as distinct from her previous penetration of the Spanish March in the form of customs and spirituality, have never been closely examined; but it quickly becomes evident that they fall in the years of and just after the Barbastro War, and are to be connected with the effort of certain leading Catalan magnates to introduce a Burgundian physical presence in the region by means of priories. In the painfully small dossier of relevant texts, the earliest and most illuminating is the cartula donationis et inuestiture magne rei of 25 November 1066 by which the great Urgelian baron Arnau Mir de Tost and his wife Arsendis confer upon the abbey a large double benefaction in the Vall d'Ager, which lies in Bajo Urgel below Tremp and the Montsech, to the west of the Noguera Pallaresa.⁽³⁵⁰⁾ Some half-dozen years earlier Arnau, with his wife and son Guillem, had ceded their Church of San Pedro de Ager, the principal ecclesiastical center of the comarca and then organized as a chapter of regular canons, to the Holy See, directing themselves first to Pope Nicholas II, presumably in 1059, and adding as well a monetary gift of 5000 Valencian gold solidi; and again, apparently in 1062, they repeated the donation to Pope Alexander II, this time with a supplementary benefaction of 3000 Valencian sueldos and 10 Negro slaves. We possess bulls of both popes, dated respectively 15 April 1060 and 17 April 1063, accepting the transfer, receiving San Pedro in tutelam sancti Petri et proprietatem , and exempting the church from all episcopal jurisdiction on condition that it pay a quinquennial census to Rome of 10 solidi aurei. ⁽³⁵¹⁾ But now in 1066 the parents -- Guillem had meanwhile died -- drastically alter their original intent and undertake to convert San Pedro into a Cluniac monasterium regulare under a prior from Burgundy.

To this end, they grant to Abbot Hugh and his monks, in full allodial ownership, the church of San Pedro within the castle, and the great castle of Ager itself, along with not only the castles, villages, fortified places, churches and lands belonging to the latter, but also feudal authority over all milites, cavallarii , homines and mancipia . . . utriusque sexus resident within its términos. Henceforth the donors and their heirs are to hold the territory of Ager as homines comandi , i.e., vassals, of the Burgundian abbot: de manu abbatis Cluniacensis et de manu prioris Aggerensis quem ibi miserit uel constituent abbas Cluniacensis ; and the castellans of Ager's lesser castles are similarly to be fideles of the abbot salua nostra fidelitate et posteritate nostra. ⁽³⁵²⁾ The church continues to be subject to the potestas sancti Petri et domni papae Romani, with the Cluniac abbot bound to maintain the quinquennial census in the bulls of Nicholas II and Alexander II; but the [58] obvious aim is to call into existence not alone a Cluniac priory but in addition a Cluniac feudal señorío in western Catalonia over which the abbot will exercise a suzerainty to be formally conferred upon him during a projected early visit of Arnau and Arsendis to Burgundy, when the benefactors would doubtless be admitted as socii . ⁽³⁵³⁾

This novel scheme, with its paired religious and secular objectives and special relevance to the problem of Cluny's association with the war of 1064, has attracted little attention, chiefly no doubt because the act of 25 November is the sole extant witness to what was manifestly a transient episode in Hispano-Cluniac annals. ⁽³⁵⁴⁾ Whether the quick collapse of the project was due to Abbot Hugh's reluctance to ratify his legate's acceptance of political responsibilities, or to papal disapproval of the church's subtraction from immediate subordination to St Peter, or to objections on the part of Arnau Mir's daughter and presumed heiress, Letgarda, arid her husband, the viscount Pons Guerau of Gerona -- to whom Ager passed in 1071 and whose son converted the comarca into the viscounty of Castellbó ⁽³⁵⁵⁾ -- the fact is that by 1068 the church at Ager had become, with Arnau Mir's consent, an abbatia nullius of canons, directly dependent upon the Apostolic See, and so remained for centuries. ⁽³⁵⁶⁾ Neither in the archives of Cluny nor in those of San Pedro itself, as now preserved at Barcelona and Lérida, does any further trace of this abortive cession apparently survive. ⁽³⁵⁷⁾ How are we to explain this unprecedented attempt to found a Burgundian dependency on Iberian soil and to establish the mother abbey as a Catalan political power? Its most obvious inspiration is the desire to convert San Pedro into a center of Cluniac intercessional prayer for the soul of Guillem Arnau, the son of Arnau Mir and Arsendis, who -- despite Sanahuja's ascription of his death to between 15 April 1060 and 6 March 1061, an hypothesis that ignores Guillem's appearance as donor in the Alexandrine bull of 17 April 1063 ⁽³⁵⁸⁾ -- pretty clearly was a casualty of the fighting of 1064 (possibly dying on 25 November, the day of the donation act) or in the counter siege of spring 1065.

From two other quarters, furthermore, we have confirmatory evidence that for some time before November 1066 Arnau Mir had been preparing to bring the Burgundian monks to Bajo Urgel. The year before, there had been buried at San Pedro de Ager Count Ermengol III of Urgel, the highest ranking peninsular participant in the Barbastro War, who commanded the Iberian forces involved, and after the city's fall was installed as its alcaide and received one third of the entire municipal territory. ⁽³⁵⁹⁾ Ermengol died in the counter-siege of spring 1065;⁽³⁶⁰⁾ and both the choice of San Pedro for his tomb as against the more natural Seo d'Urgell, and the fact that on 12 April 1066 his widow Sancha bestows a number of properties upon the Church of Ager, ⁽³⁶¹⁾ unmistakably anticipate imminent introduction of the Cluniacs into an intended center of intercession for the two most illustrious Catalan victims of the war. Secondly, there is the testimony in Arnau Mir's charter of 1066 that it was drawn up in the presence of two monks of Cluny, of whom one, Frater Henricus (who also acts as scribe) was obviously a religious high in the abbot's confidence and serving as his deputy in the negotiations, much as in 1053 Frater Galindus had been Hugh's envoy to the court of Fernando I.

[59] With these consistent indications that the monastico-political decision incorporated in the diploma was arrived at as early as the summer of 1065, there needs also to be taken into account, if we are to understand the link with the Barbastro War and Cluny's possible collaboration with the Catalans in this enterprise, Arnau's previous record of many years as indefatigable combatant against the Saracens, and his liberation of the Vall d'Ager -- an achievement the baron himself cites with visible pride when giving Cluny this reconquered land pro amore filii nostri dilecti ibi dormienti. ⁽³⁶²⁾ Here we move beyond the intercessional motive to observe that the gift is coincident with an abrupt reversal of Arnau's earlier hope of associating the Gregorian papacy with his drive against the infidel, and in its stead a rapprochement with the Burgundian congregation. As early as ca. 1030 Arnau Mir can be found advancing against the Moors from his base at Tost and the environs of the Seo d'Urgell into the conca de Tremp and across the Montsech to the Vall d'Ager. ⁽³⁶³⁾ After conquering the latter in 1034, briefly losing it and then in 1047-1048 definitively regaining it, he proceeds in subsequent years to acquire numerous castles and lands all the way from Tost to the basin of the lower Noguera Pallaresa. In part this expansion paralleled the efforts of the two foremost Catalan princes, Counts Ramón Berenguer I of Barcelona and Ermengol III of Urgel, to extract both territories and parias from the Taifa kingdom of Lérida and, before long, from that of Zaragoza as well. ⁽³⁶⁴⁾ In September 1050 Arnau received from the Barcelonese count the castle of Camarasa, which lay just below the junction of the Pallaresa with the Segre, on the route to Balaguer, along with a promise of one-half of that city when it should be captured from the infidel. In the act of infeudation for this castle, of 5 November, the lord of Tost and Ager promised that if his military obligations to Count Ermengol prevented his accompanying Ramón Berenguer on a campaign in Hispania , he would instead send his son Guillem once the latter had come of age. ⁽³⁶⁵⁾ Arnau's participation in the Catalan Reconquista, it is to be noted, took him not only southwards towards Balaguer and Lérida, but also west across the Noguera Ribagorzana in the direction of the valley of the Cinca River and the Muslim fortresses of Graus and Barbastro in Baja Ribagorza. In 1057, two years after Ramiro I's initial vain attack upon Graus, Ermengol in entrusted to Arnau Mir the highly strategic Ribagorzan castle of Caserres, one of the most advanced spearheads of the Catalan offensive pressing towards Barbastro and the Cinca. ⁽³⁶⁶⁾ The following year Arnau's name appears among those of the lieutenants of the two counts who subscribe to the military pact that was negotiated in preparation or a new Ribagorzan campaign in the vicinity of Purroy and Caserres. ⁽³⁶⁷⁾

It is precisely at this point in his career that Arnau Mir de Tost begins to display a consciousness -- perhaps more acute than that any other magnate of the Catalan West -- of the religious implications of his long decades of struggle against the infidel in the marches of Urgel, and undertakes to place the most important church of all his domains, San Pedro de Ager, in tutelam et proprietatem sancti Petri . This gives the original grant to Pope Nicholas II a distinctly Ribagorzan, not Bajo Urgelian, setting, and places the [60] papal encomium of this stalwart warden of the Catalan frontier as nobilissimum et religiosissimum uirum, inimicorum Dei Agarenorum aduersarium et debellatorem , and as liberator of terram illam de potestate et errore Agarenorum , ⁽³⁶⁸⁾ just at the moment when the Catalan drive was pushing through the territory across the Moguera Ribagorzana. Thus the bull of 1060 proves that papal involvement in the region of Graus and Barbastro precedes Ramiro's death in 1063 by at least three years, which means that Alexander II's concern with this area was inherited from his predecessor and dates from the very commencement of his pontificate. Arnau Mir does not subscribe the pact of 25 July 1063, which the Counts of Barcelona and Urgel drew up on the eve of the Barbastro War and which embodies their plan to strike westward past Estopinan, Purroy and Canelles towards the Cinca; ⁽³⁶⁹⁾ and there is no proof that, like his ill-fated son Guillem, the old veteran of the Catalan marches personally participated in the fighting at Barbastro the next year. What we can, however, accept is that between 1059 and Alexander's bull of April 1063 Arnau Mir's plans at Ager included the papacy, but that at the time of the Barbastro War he substituted a new, broader alliance with Abbot Hugh. Why this shift from Rome to Burgundy?

To be sure, well before 1066, the master of Tost and Ager must have been aware of Cluny's fame. As far back as ca. 1020 he appears in Catalan circles known to have been sympathetic to propagation of the abbey's reformism below the Pyrenees. Here can be noted his ties with the venerable Oliba, bishop-abbot of Vich-Ripoll-Cuixá, on whose counsel Sancho el Mayor had brought the Hispani of Paternus to San Juan de la Peña;⁽³⁷⁰⁾ with Poncio, el Mayor's agent in the pro-Navarrese faction at Alfonso V's court and later bishop of Oviedo and Palencia, who was himself a Catalan, indeed an Urgelian, and former abbot of San Pedro de Tabèrnoles, a house with which both Arnau Mir and Arsendis can be associated; ⁽³⁷¹⁾ and with leading figures of Seo d'Urgell, Gerona and Barcelona, in whose diplomas his subscription occurs.⁽³⁷²⁾ It is also demonstrable that Arnau's devotion to Cluny was peculiarly his own and did not originate with his wife. Arsendis' extant will of 22 May 1068, in its long list of legacies to numerous Catalan monasteries, makes no mention of the Burgundian monks, although it does provide for a bequest to another French house, Sainte-Foi de Conques of  -- in Sanahuja's translation -- "mi espejo de Indias con mis hebillas mayores de oro". ⁽³⁷³⁾ But when her husband acted to execute this testament as her albacea, he revised its terms to include "una capa de aztor encarnada" for Cluny, ⁽³⁷⁴⁾ and in his own will of 11 August 1071 he gave the abbey a third -- reduced in a codicil to a quarter -- of all his movable wealth. ⁽³⁷⁵⁾ Whatever were the factors adverse to the cession of 1066, these provisions make it certain that the wealthy noble's affection for the Burgundians remained strong down to his death between 11 August 1071 and 1 January 1072. ⁽³⁷⁶⁾ This record might be used to explain why in 1066 Arnau Mir de Tost looked to Cluny on intercessional grounds after losing his one surviving son; it does not satisfactorily account for the demotion of Rome or the investiture of Abbot Hugh with Ager, especially since the daughter Letgarda had successoral rights under the strong protection provided women as heiresses and castellans under Catalan law.

[61] Furthermore, this is not the only case we encounter of a Catalan switch from the papacy to Cluny in the epoch of Barbastro. On 25 January 1074, after the legally required interlude of six months, the seven executors of the last will and testament of the baron Girbert Mir of Olérdola (in the Panadés below Barcelona) transferred to Cluny, in accordance with his wishes, certain castles and lands that had belonged to him.⁽³⁷⁷⁾ These include the castella of Berano and Roda, one half of Bleda and various lands at Foix, 'Mager', Cabrera, Bañolas, Vilademuls; houses at Barcelona and 'Provincials'; and, lastly, the promise of a third castle, 'de Durius', which was to be reunited with Berano on the death of its holder, Girbert's sister Dalmidana. The terms of the bequest resemble those of Arnau Mir's benefaction of 1066: like San Pedro de Ager, the castles had previously been placed under the protection of the Roman Church; ⁽³⁷⁸⁾ they were likewise located in the dangerous frontier country of Catalonia, although in this case in the eastern sector, near Vendrell; their castellans were to pay feudal obedience to Abbot Hugh; and a permanent monastic center, i.e., a priory, is again envisaged, with four monks, who will presumably reside at Berano and solicit divine help for the kavallarii qui defendant ipsam terram contra Sarracenos.

There is no allusion in this text to the Barbastro campaign of ten years previous, but the chances are good that Girbert Mir's generosity to Cluny traces back to his years of combat in western Catalonia. He was the son of an old rebel against Barcelona, Mir Girbert, but Girbert himself became a trusted lieutenant of Ramón Berenguer el Viejo, not least in the letter's push through Urgel and Ribagorza against the Hudids. ⁽³⁷⁹⁾ This confidence reached a peak when, on 5 February 1064, just before the attack upon Barbastro, the count placed Girbert in command of his recently captured Ribagorzan stronghold of Estopiñán, a principal key to further Catalan movement towards the Cinca valley. ⁽³⁸⁰⁾ Thus, in 1064, both Girbert Mir and Arnau Mir de Tost were holding major castellanies of Ramón Berenguer in Catalan Ribagorza; both had earlier sought papal suzerainty over certain of their possessions; both after Barbastro turned to the Cluniacs and sought their partnership, religious and political, in their frontier lands confronting the Moor. It is not necessary to insist that Girbert was directly inspired by Arnau Mir's gift eight years before, although this is not unlikely; it is enough to perceive that both donations reflect a Catalan reaction against Rome and swing to Cluny set in motion in the years immediately after the 'crusade' of 1064.

There is a third Catalan cession to Cluny. This one occurs in 1079, fifteen years after Barbastro, when Ramón Folch, viscount of Cardona, donates his Church of San Pedro de Caserres which successfully avoided the fate of San Pedro de Ager and Berano to become the earliest authentic Burgundian priory in western Catalonia. ⁽³⁸¹⁾ This is not the Caserres across the Noguera Ribagorzana where Arnau Mir held the castle of Count Ermengol III, but the one in Urgel near Berga. Ramón Folch's close connections with Arnau over many years may however well explain the vicecomital esteem for the abbey. For under the will of Eriballo, the bishop of Seo d'Urgell (1036-1040) who was also viscount of Cardona, Arnau Mir was to act not only as executor [62] but as guardian of the bishop's nephew Ramón Folch until the latter reached the age of fourteen and came into possession of Cardona and other castillos of the district. ⁽³⁸²⁾ This personal relationship alone makes it unlikely to have been accidental that Ramón's donation of Caserres to Abbot Hugh bears the identical day and month -- 25 November -- as the grant of Ager in 1066, which may have been the anniversary of the death of Arnau's son Guillem. The donation act says nothing of Barbastro or the war against the Muslims, or of any assignment of castles, lands or feudal authority, simply providing that Cluniac religious are to dwell at Caserres according to the Benedictine Rule and their abbey's customs, and to offer intercession for Ramón, and his co-donors, his brother, the archdeacon Folch, and his wife Ermesinda. But there are the old ties with Arnau Mir, and at the very least the gift confirms Cluny's Catalan orientation in the epoch of Barbastro.

Finally, it is in the light of these three Catalan cessions to Cluny that we must seek the meaning of two pieces of evidence that indicate that knights from Burgundy actually participated in the War of Barbastro among the foreign troops forming part of the Christian army. In his L'ystoire de li Normant, Amato di Monte Cassino records that the forces attacking the Muslim citadel consisted of grant chevalerie de Francoiz et de Borguegnons et d'autre gent . ⁽³⁸³⁾ An inscription from a tomb at Tolosa in Guipúzcoa preserves the name of the probable commander of this Burgundian contingent: Thibaut de Chalon, the count of Semur and a half-uncle of Abbot Hugh of Cluny, who appears to have died, presumably of his wounds, while returning home from Spain in or soon after 1065, following the loss of Barbastro to the enemy. ⁽³⁸⁴⁾ Between them these two notices show that if the abbot of Cluny did not actively promote recruitment of crusaders for Spain in 1063-1064, he at any rate permitted this to be done by others within the duchy and on a familial level where one would suppose his veto decisive. But this is only to assert that when in 1063 Alexander II, as we believe, preached the crusade in Spain -- a summons couched in general terms of Christian emergency, surely, rather than calling for a rescue operation on behalf of an Aragonese state still almost unknown north of the Garonne -- Hugh complied with the papal will. Indeed, like the Catalans themselves, he may well have felt it unwise, by adopting a purely negative stance, to deny himself all influence over the conduct of the proposed expedition.

This remains conjectural; but the larger question is why in the last years of Fernando I's reign Cluniacs and Catalans should have drawn together and the abbey been given dependencies and political functions below the Pyrenees. The answer must be sought in the framework of mutual Catalan and Leonese-Castilian hostility towards Aragon in the frontier zone of Ribagorza.

(ii) The Ribagorzan Question and the Fernandine-Cluniac Alliance. The realization that Cluny's documentable connections with the Barbastro War run not through Aragon but Catalonia, and specifically Urgel, underscores what is fully apparent once the episode of 1064 is approached from the peninsular and not so exclusively as is customary from the ultramontane side. Its character is twofold: in [63] part a Franco-papal crusade in the Aragonese interest; in no less significant part, a Catalan operation of basically secular, anti-Aragonese aims constituting a climactic phase in the long-standing Barcelonese and Urgelian thrust against the Hudid Taifas of Lérida and Zaragoza. Because this latter movement took the Catalans in the third quarter of the 11th century not only into Bajo Urgel but also deep into Muslim Ribagorza, it inevitably brought Ramón Berenguer I, Ermengol III and other powerful lords of the Hispanic March into a struggle with Ramiro I and the Jacan kingdom for the mastery of the valley of the Cinca River. ⁽³⁸⁵⁾ The extent of the consequent Catalan-Aragonese hostility on the approaches to Graus and Barbastro before 1064 still awaits proper study, but enough is known to make it possible to correct the distortions imposed upon the Barbastro War by the Boissonnade school of interpretation and to perceive how throughout these years Cluny was consistently friendly with the two anti-Aragonese Christian powers in Ribagorza, the Catalans and the Leonese-Castilians.

From the Catalan side the key to understanding the situation is Ramón Berenguer's well-conceived plan to create a long, narrow corridor of Barcelonese-controlled territory conquered from the Muslims that would stretch due west from the Camp de Tarragona across the foot of Cerdaña, Pallars jussa and Bajo Urgel into Baja Ribagorza and the comarca of Barbastro.⁽³⁸⁶⁾ Such a salient, defensible through -- in Miquel's words -- "una barrera infranqueable de castillos"⁽³⁸⁷⁾ -- promised to maintain el Viejo's suzerainty over the frontier conquests of Ermengol III of Urgel, Ramón of Cerdaña, Ramón of Pallars jussà, Arnau Mir de Tost, and other ambitious Catalan magnates, while simultaneously permitting the Count of Barcelona to extend his power and tributary-territorial gains beyond the repeatedly subdued Léridan kingdom into the Ribagorzan reaches of al-Muqtadir's Zaragoza. This program was already well under way when in 1050 Count Ermengol III entered into the compact with Ramón Berenguer in which he agreed to assist Ramón in war against the Saracens and become his vassal for any lands Ermengol himself conquered; in return the lord of Urgel was to receive the castle of Cubells (below the junction of the Noguera Pallaresa with the Segre, on the road to Balaguer) and one third of any parias secured in the conflict. ⁽³⁸⁸⁾ This alliance comes at a time when Ramón, already in possession of Camarasa, was enfoeffing this Ribagorzan castle not far from Cubells to Arnau Mir de Tost, along with the promise of one-half of Balaguer whenever that city was taken.⁽³⁸⁹⁾ This is the era also when the coalition of Catalans, Navarrese and Léridans forces al-Muqtadir to agree to the tribute, out of his share of which King García of Pamplona in 1052 was to give a tenth to his newly founded monastery of Santa María de Nájera. ⁽³⁹⁰⁾

It may well have been a combination of Aragonese dismay over such successful Catalan-Navarrese capture of the Zaragozan paria and of fears aroused by Fernando I's decisive victory over Navarre at Atapuerca two years later, that led Ramiro I to take a hand in the game by assaulting Graus. At any rate, his intervention here, despite its dismal failure, can be seen as related to a marked shift in the direction and primary focus of the Catalan Reconquista away from the push towards Lérida and to concentration upon Ribagorza. ⁽³⁹¹⁾ [64] On 5 September 1058 Ramón Berenguer and Ermengol agree to a new treaty of alliance which envisages their combined operations in a war against al-Muqtadir, with the Count of Urgel again being promised one third of all tribute gained. ⁽³⁹²⁾ It is this campaign, presumably, that results in the successful Catalan conquest of a notable group of Ribagorzan castles pointing towards Barbastro -- Pilzán, Benabarre, Purroy, Estopiñán and Canelles, the last three of which el Viejo turned over to his wife Almodis. ⁽³⁹³⁾

As for Aragon's Ribagorzan policy during the years before the siege of Barbastro, despite obscurity in detail its main lines are discernible. Whether there were serious armed clashes between the Catalan and Aragonese vanguards in the contested frontier zone it is now difficult to be sure, although, with the Jacan kingdom's advance by ca. 1060 into the Isábena valley at Laguarrés and beyond Lascuarre and Viacamp to near Benabarre and the Guart River, all lying to the east and southeast of Graus, the rival armies were in dangerously close proximity. ⁽³⁹⁴⁾ Ramiro's defeat at Graus in 1055, a severe blow to the Aragonese fiscal and territorial hopes, helps explain what in succeeding years looks like a deliberate attempt to win over the Urgelians to alignment with Jaca against Barcelona. One sign of this is Ramiro's marriage of his daughter Sancha to Ermengol III, an event whose precise date is unknown, but which justifiably can be placed between 1058 and 1062. ⁽³⁹⁵⁾ This marriage saw the Count of Urgel bestow upon his Aragonese spouse, presumably as arras, certain of his Ribagorzan holdings -- the important Castillo of Pilzán and his third of that of Purroy -- in the Catalan salient near Caserres and Estopiñán in extremis finibus marchiarum iuxta Hispaniam.⁽³⁹⁶⁾ It is her right in these citadels that in 1067, after her husband's death at Barbastro, the countess willingly or unwillingly returns to Ramón Berenguer, from whom Ermengol had originally received them. ⁽³⁹⁷⁾ Still another transparently anti-Barcelonese move of Ramiro I in Urgel appears in the king's surprising readiness to accept an Urgelian rather than an Aragonese orientation of the Ribagorzan see of Roda in the Isábena valley, with a bishop Arnulf of known predilection for the Seo d'Urgell.This policy was to be drastically reversed in 1064 or 1065 by his son Sancho Ramírez who deposes Arnulfo and installs a prelate loyal to Jaca; and although the ablest student of this question, Ramón d'Abadal, does not link the royal shift to the Barbastro War, it can surely be viewed as a by-product of the breakdown of Aragonese-Urgelian raprochement resulting from that conflict.

By 1063 however Aragonese expansionist ambitions in Ribagorza confronted an even more formidable adversary than the Count of Barcelona and the Catalans -- Fernando I of Leon-Castile. The king-emperor was not seeking territory along the Cinca, but he was bound to protect the welfare of his prized tributary, King al-Muqtadir of Zaragoza under an alliance which from 1059 injected a complicating factor into the Ribagorzan question. The years 1058-1059 had seen 
the rise of Fernandine interest in the kingdom of Zaragoza -- at first, hostile, as when the king-emperor attacked al-Muqtadir's domains along the Upper Duero, devastated the valley of Bordecorex, raided as far as Caracena and Medinaceli, and annexed the fortresses of San Esteban de Gormaz and Berlanga; ⁽³⁹⁹⁾ very soon thereafter, friendly. [65] as a consequence of the Zaragozan monarch's severance of his ties with Sancho Garcés of Pamplona during the Navarro-Castilian war of 1058-1060 and his entry into tributary alliance with Fernando, presumably in 1059. ⁽⁴⁰⁰⁾ This immediately changed the rules of the game, although Ramiro I was slow to realize this: an attack upon Zaragoza was now an attack upon the king-emperor. In the spring of 1063, therefore, when Ramiro once again marched against Graus, he found himself confronting in addition to the Muslims, the strong force of 300 Castilian knights under the Infante Sancho which Fernando I had sent to the support of his Hudid ally. Thus the Aragonese king's death and the crushing defeat of his army on 8 May was as much a Leonese-Castilian as a Zaragozan victory, portending possible triumph of the dynastico-imperial program that ever since Atapuerca in 1054 Fernando had pursued with the purpose of imposing his authority over all the former domains of Sancho el Mayor. Gratis thus posed a double threat to the Pyrenean kingdom: its independence and the future expansion fundamental to its viability as an Hispanic great power were both now in jeopardy.

There can be little doubt that not only for Aragon but for all the Christian states involved in the Ribagorzan Reconquista, the second battle of Graus is of capital importance, leading to early preparations for an international expedition in the Cinca Valley and bringing both the papacy and Cluny into new, more intimate and geographically more extended relations with Iberian Christendom. As far as Aragon is concerned, what happened is reasonably clear, unless we follow David in ruling out Alexander II's intervention and thus leave the 'crusade' without an organizer. The recent Catalan successes at Estopiñán, Purroy and Canelles, the open collaboration of al-Muqtadir and Fernando I in the defence of Graus, the grim spectacle of a slain king and beaten army, must all have stirred in Sancho Ramírez' counsellors painful memories of that other Fernandine triumph, when García of Nájera similarly fell at Atapuerca and Navarre was compelled to acknowledge the imperial hegemony. One alternative, totally unprecedented in the Iberian lands west of the Hispanic Mark, must have presented itself, whether or not we can allow for the presence at Jaca in this crucial summer of Cardinal Hugo Cándidus: a plea to the dynamic leadership of the Gregorian papacy for immediate ideological and military support. To be successful north of the Pyrenees, such an appeal had necessarily to be couched not in terms of Realpolitik and the grave dangers posed by Catalan and Leonese-Castilian success in Ribagorza; but as a Christian emergency in the face of the far from menacing Saracens of al-Muqtadir. At this moment the Spanish crusade was born. In 1063, for the first but by no means the last time, ⁽⁴⁰¹⁾ an isolated, relatively weak, Aragon was to have recourse to large-scale foreign intervention in the Iberian civil war of the Reconquista, ostensibly to strike against the Hudid Kingdoms to the south and to extend her frontier to the indispensable Ebro, but in rea1ity to checkmate the Ribagorzan moves of her Leonese-Castilian and Catalan rivals and to insure her continued independence before the growing might of the Hispanic Empire founded by Fernando I.

Things moved very differently in Catalonia, where acute concern must have been felt both over Ramiro I's renewal of military activity [66] in Ribagorza in the spring of 1063, and now, after Graus, at the prospect of Franco-papal intervention in the strategic region. Through the summer of this year we can discern certain preparations for a new Ribagorzan war. On 5 July Ramón Berenguer infeudates the castles of Balcereny and Gaià and the honor of Oristà, in the Llobregat valley south of Berga, to the brothers Bernat and Mir Riculf, who were obliged to discharge a service of 50 knights. ⁽⁴⁰²⁾ On the 25th of the same month, Ramón negotiates his third military pact with the Count of Urgel, in a document bristling with references to the defense of Barcelonese citadels well back from the Moorish frontier, such as Barcelona itself, Olérdola, Gerona, Manresa, Vich and Cardona, and looking forward to a Ribagorzan war in which Christian as well as Muslim foes may be encountered: inimici eius christiani uel sarraceni, super christianos uel sarracenos . ⁽⁴⁰³⁾ This document is of exceptional value for its vivid revelation of the attitude of the Catalans towards the preparations that must have been already in progress for the expedition of 1064. Clearly they saw the latter as a serious menace to their long-term interests and recent successes across the Noguera Ribagorzana, yet one which because of its papal sanction they could not openly oppose and from which they dared not abstain. It is evident from the text that Ramón Berenguer, despite the frequent allusions to his supreme political authority, has no intention of taking an active part in the campaign. This contrasts with the contrary assumption in the pacts of 1050 and 1058, as does also el Viejo's willingness to allow Ermengol the entirety, and not merely a third, of any parias he can exact from al-Muqtadir, although the Count of Urgel is still barred from retaining more than a third of the conquered lands. Here is the basis for Ermengol's command of the Catalan forces the following year, and the allotment to him after victory of one-third of Barbastro's territory; ⁽⁴⁰⁴⁾ but the emphasis throughout is thoroughly secular, lacking all overtones of holy war. Repeated references to the key Ribagorzan fortresses of Purroy, Estopiñán and Canelles reflect Ramón's deep concern for the safety of the forward bases of the Barcelonese corridor, now halfway to the Cinca and Barbastro; and the Christian enemies against whom Ermengol swears to defend the count's possessions can only be the Aragonese, if not also their anticipated transmontane allies, the French crusaders. Much has been made of the supposed relevance to Barbastro of a text containing ten articles presumably issued in 1064 by Ramón Berenguer and Almodis in an assembly of bishops and magnates at Barcelona. ⁽⁴⁰⁵⁾ This has been taken to establish the Peace of God throughout Catalonia in order to further the 'crusade';⁽⁴⁰⁶⁾ but in fact its only mention of the Moors is in a single article (no. X) relating to peruersi homines who capture Christians in order to sell them to the infidel. More significant is the fact that a few months after his pact with Ermengol, the Count of Barcelona takes another decisive step to bolster his position across the Noguera Ribagorzana, when on 5 February 1064 he entrusts the fortress of Estopiñán to his mesnadero Girbert Mir.⁽⁴⁰⁷⁾ Thus on the eve of the Barbastro campaign two future benefactors of Cluny commanded castles in what was for the Catalans the most sensitive theatre of the whole operation: Girbert Mir at Estopiñán, Arnau Mir de Tost at Caserres.

[67] It is then, in 1063 and 1064, at the time when Sancho Ramírez was forging the alliance with Alexander which a few years later was to take him to Rome and lead to the conversion of his kingdom of Aragon into a papal fief, that we should place the estrangement of the Catalan barons from papal association with the Reconquista and the beginnings of their attraction to Cluny as the ecclesiastical shield of their Ribagorzan interests -- an attraction felt most strongly, it would seem, not by the great counts and viscounts of the Spanish March, but by the active reconquistadores just below them. The campaign of 1064 was to bypass Graus, its most natural objective since this twice had been the scene of Aragonese defeat and was the northernmost Muslim outpost blocking Jaca's passage down the Cinca, in order to attack Barbastro, which belonged to al-Muzaffar of Lérida, although that prince, possibly under Catalan restraint, failed to defend the city. ⁽⁴⁰⁸⁾ It would be interesting to know whether Cluny played any part in deciding upon this strategy, which avoided Ramiro I's fatal mistake in 1063 of bringing on confrontation with Fernando I, and at the same time transferred the whole operation into a zone of Catalan tributary and territorial expansion where, in contrast with Graus, Barcelonese and Urgelian pressure could be far more effective.

This brings us finally to the crucial question for the evolution of Fernandine-Cluniac relations, that of the impact of the Barbastro War upon them. Had Cluny 's friendship with the king-emperor already attained the level of confraternal-censive conjunctio prior to the second battle of Graus, and was this a prime factor in driving the Aragonese towards the papacy? or was it rather the success of Jaca in securing papal and French intervention in Ribagorza, with its twin threats to the Leonese-Castilian vassal at Zaragoza and to Aragonese subordination to imperial suzerainty, that inspires Fernando I's conversion of his original purely devotional ties with the Burgundian congregation into a standing alliance with implicitly political connotations? It is a choice between hypothesizing a pious association gradually maturing into societas and subsidy, which would have been possible from 1059 with the Zaragozan paria as Fernando's first regular supplier of Muslim gold pieces; or looking for a political conjuncture of such broad dynastic and national import as to have led the king-emperor into formal spiritual subordination to Abbot Hugh in order to defend the gravely challenged Hispanic program.

On the whole, the probabilities favor the second solution. So little is actually known of Fernando's reaction to the campaign that most historians altogether ignore the problem, impressed no doubt by the fact that he took no part in the siege or countersiege but on the contrary in January of that year led the Leonese-Castilian army off to the opposite side of the Peninsula, to the six-months invasion of the Kingdom of Badajoz and the successful siege of Coimbra. ⁽⁴⁰⁹⁾ But the monarch who in the previous spring had sent troops to stop Ramiro I at Graus can hardly have remained indifferent to this latest Aragonese adventure in the Cinca Valley. The Pyrenean kingdom's successful turn to Rome in defiance of the lesson administered at Graus, the prospect of a Franco-papal army fighting in Jaca's interest, above all the possibility that Aragon had discovered a sure [68] formula in foreign ecclesiastical and military aid for overcoming her own material limitations and blocking for good her subjection to Leonese-Castilian domination, must have thoroughly alarmed the imperialists at León and imposed upon them the necessity of seeking a similar partnership with Cluny. The Leonese-Castilian western campaign of spring 1064, which removes the imperial army far from Ribagorza at a most critical moment, implies certain premises behind the king-emperor's action: a determination to avoid all possibility of an armed clash with the international expedition; information received well before January 1064 of the plan to attack Barbastro, not Graus, which eliminated the contingency that violation of Zaragozan territory might force Leonese-Castilian intervention against the crusaders; conviction that a close, dependable, and already operative union with Cluny assured full protection of imperial interests at Rome and among the French as well as probably among the Catalans also. ⁽⁴¹⁰⁾ On these grounds the forging of the conjunctio belongs in the summer of 1063 when, with the Aragonese-papal entente already moving towards the enterprise of Barbastro, and with widespread preparations for the war under way in Catalonia and France, Fernando I would have entered into the negotiations with Abbot Hugh out of which were to come the lost real privilegio, the munificent subsidy of the thousand metcales , the admission into confraternity -- in short, the forging of the Fernandine-Cluniac alliance in reaction to that of Jaca with the Holy See.

This conclusion -- that Fernando I's entry into a quasi-dependent, parapolitical relationship with a foreign ecclesiastical power subsumes a grave dynastico-political crisis of the Hispanic Empire -- finds support also in the record of Alfonso VI's better documented relations with the Burgundian abbey. Here too we have, between late 1072 and July 1077, an initial phase of pious esteem, friendship and extensive generosities which was transformed through papal intervention in Spain -- in this case Gregory VII's claims to suzerainty over the Peninsula, and specifically over the Leonese-Castilian kingdom, his attack upon the Leonese-Castilian Church in the demand for abolition of the Hispanic Rite -- into a revival on 10 July 1077 of Fernando's old conjunctio, with the re-institution of the census, now in doubled amount, and entry into societas. ⁽⁴¹¹⁾ The cases of father and son seem to run parallel: both initially approach the abbey solely in terms of devotion and generosity, eschewing all suspicion of political corollary; both at a certain point abandon this caution and enter into regular financial and confraternal obligations as a consequence of papal pressures directly adverse to dynastic and imperial interests.

Can we then believe that the relationship Fernando I and the Leonese-Castilian state thus formed with the Burgundian abbey was in truth a vassalic one, much like that which Sancho Ramírez was to formalize in 1068 between Aragon and the Gregorian papacy? and that the Hispanic Empire was in effect a Cluniac vassal state? It is to this final problem of the Fernandine-Cluniac alliance that we now proceed.

 
No aspect of the Leonese-Castilian alliance with Cluny in the 11th and 12th centuries is more striking than the fact that the relationship consistently operates in a context not merely royal but pan-Hispanic and imperial. Students of Leonese imperialism have often tended to minimize this doctrine as a highly theoretical neogothicist utopianism lacking juridical or institutional counterpart; ⁽⁴¹²⁾ but however valid such judgment may be for the Asturoleonese epoch, Sánchez Albornoz is surely right in maintaining that with the Navarro-Basque dynasty the imperial idea, with its implicit promise of future national unity, enters upon a new, more meaningful stage of its historical significance. ⁽⁴¹³⁾ It is nevertheless indispensable to perceive that in this evolution the decisive turningpoint is not the reign of Sancho el Mayor or, as frequently suggested, that of Alfonso VI, but the terminal decade of Fernando I, commencing on the morrow of Atapuerca, when El Magno  -- and this is patently one of his principal claims to that appellation -- undertakes to fuse the ancient Oviedan belief in the peninsular primacy of the Hispanic king-emperor of León with his own newly acquired primacy of the Navarro-Basque house. ⁽⁴¹⁴⁾

The fratricidal conflict of September 1054 constitutes indeed one of the great watersheds of medieval Iberian history, insufficiently as this has been recognized. Before it, as first king of a Castile truncated by Navarrese annexations, a king overshadowed by his stronger older brother García at Pamplona whose familial seniority he acknowledges and upon whom he depends for military assistance in the war of 1037 against Vermudo III,⁽⁴¹⁵⁾ Fernando appears as primarily a Castilian sovereign, ruling from Burgos, who happens also to be, by reason of Tamarón and the earlier marriage with Vermudo's sister and rightful heiress D. Sancha, monarch of . ⁽⁴¹⁶⁾ García's defeat and death completely transforms the situation. Fernando now ranks as Sancho el Mayor's oldest surviving son and therefore Navarro-Basque dynastic leader, possessed of hegemonic rights over all subdivisions of his father's old nuclear and conquered domains. ⁽⁴¹⁷⁾ At the same time, no doubt in part because it serves to reinforce his position with regard to Navarre and Aragon, he suddenly embraces with enthusiasm, as he had not previously done, the neogothicist ideology of an imperial regnum Hispaniae under the rex-imperator of León. There commences now an intensive Leonicization of the entire kingdom, the imposition upon Castile of an unwelcome subjection to her erstwhile conquered western territory -- a reversal of regional status against which Castilian particularists like the Cid would subsequently react -- and the vigorous promotion of the old Leonese ideal of war against the infidel. The results after 1054 are dramatic: a shift of the whole political and cultural fulcrum of western and central Spain from the banks of the Arlanzón and the Arga to those of the Bernesga; and, after twenty years of unexciting Castilianism, a truly revolutionary decade in which tne Leonicized Fernandine monarchy creates that henceforth pivotal center of peninsular power and ultimate domination that [70] history, somewhat ironically, knows as the Kingdom of Castile.

In this post-Atapuercan age of dynastico-imperial synthesis everything confirms the Leonicizing, imperialist basis of Fernando's alliance with Cluny as an integral element in the new Hispanic program. Any familial motivation can be discounted: despite the politically inspired reburial of Sancho el Mayor at San Isidro de León in 1064, ⁽⁴¹⁸⁾ neither Fernando nor his successors ever seem to have regarded their ties with the abbey as in any way stemming from their progenitor's charities to, and societas with, the monks or from Abbot Odilo's faith in indissolubilis amicitia. On the contrary, the Leonese-Castilian chronicles, and even more decisively the extant Hispano-Cluniac reales privilegios, including especially those of that strong family loyalist Alfonso VI, fail to mention Sancho in their intercessional stipulations for parentes , but commence these solely with Fernando I. ⁽⁴¹⁹⁾ The same is true of the notices occurring in the Burgundian custumals, abbatial epistles, Hugonic Vitae, etc.: however long the monks continued to observe the anniversary of el Mayor's death, ⁽⁴²⁰⁾ they never recall that monarch as the fons et origo of the lucrative Hispanic connection or include him in their daily prayers for the reges Hispaniarum.⁽⁴²¹⁾ Modern historians make much of el Mayor's introduction of the Cluniacs into Spain; for the men of the 11th and 12th centuries, on the contrary, it is Fernando -- no mere Pamplonese king but authentic Hispanic emperor -- who rightly stands as the venerated founder of the abbey's fortunes below the Pyrenees.

This confirms what we have repeatedly observed: Fernando I's compertio and tightening bonds with Cluny were a distinctly Leonese phenomenon. It is true that Frater Galindus' visit to the Tierra de Campos precedes Atapuerca by a year; but this may well be an indication of the already rising tide of Leonese influence in the kingdom that must have been a prominent factor in the challenging of García. It has been emphasized above how non-Castilian, how deeply Leonese, was the praeparatio Cluniacensis in western Spain: the trans-Pyrenean and Catalan connections attested in the diffusion of the San Antolín cult; the planting of Catalan ecclesiastical outposts in the Tierra de Campos at Palencia and Dueñas; the prominence of religious Europeanizers in the highest levels of the Leonese aristocracy, including the Queen-Empress Sancha, whose influence over her husband's policies after 1054 was exceptionally strong. ⁽⁴²²⁾ Evidence also abounds that Fernando's turn to Cluny was but one of a whole series of decisions in the religious sphere taken between 1054 and 1065 to further the Leonese orientation of the transformed state. There is the convocation on Leonese soil of the imperial reform councils of Coyanza (1055) and Compostela (1056); ⁽⁴²³⁾ the ceremonious translation from Seville to the Hispanic capital in 1063 of the body of St Isidore, to be deposited there in the newly reconstructed monastery of San Juan Bautista, henceforth famous as San Isidro de León; ⁽⁴²⁴⁾ Fernando's abandonment -- which the Silense attributes to D. Sancha's persuasion⁽⁴²⁵⁾ -- of Oña or San Pedro de Arlanza in Castile, for his own entombment, and the choice of San Isidro de León as the official pantheon of the Navarro-Basque dynasty, and the already mentioned removal thither from Oña of the remains of Sancho el [71] Mayor ⁽⁴²⁶⁾ The impression persists, although this still awaits proper study, that Fernandine donation acts to churches and monasteries disclose a decided preference in this decade for Leonese over Castilian beneficiaries. ⁽⁴²⁷⁾ Bishop Bernard of Palencia's intention to establish his see in the Leonese Tierra de Campos as the metropolitical capital of the Leonese-Castilian Church, long before the recovery of Visigothic Toledo, points in the same direction.⁽⁴²⁸⁾

As for the specifically imperial character of the Cluniac connection under both Fernando and Alfonso VI, this finds repeated illustration on both sides of the Pyrenees. In Spain, where we possess, as we do not for Fernando, various privilegios on the census and for monastic transfers to the Burgundians, the Leonese-Castilian chancery regularly depicts the ruler as acting in his Leonese or imperial Hispanic capacity, never as rex Castellae, even in the cession of Santa Coloma de Burgos. When giving Cluny on 29 December 1073, the eighth anniversary of his father's death, the real abadía of San Isidro de Dueñas, Alfonso styles himself in a pergamino filled with Fernandine reminiscences, rex Ispaniarum atque Leonensis. ⁽⁴²⁹⁾ In 1077 it is as serenissimus princeps -- a distinctly imperial honorific -- that he gives the Burgundians his house of Hérmedes de Cerrato, while also describing himself as prolis magni et gloriosi inperatoris Fredenandi.⁽⁴³⁰⁾ As diuina gratia imperator totius Hispanie or in similar words he transfers Santa María de Nájera in 1079, Santa Columa de Burgos in 1081.⁽⁴³¹⁾ Imperial Leonese is the connotation also of the two diplomas on the census duplicatus: that of 1077 granted by him as rex Le(gi)onum , the confirmation of 1090 as Hispaniarum rex . ⁽⁴³²⁾ In subsequent years the Hispano-Cluniac charters of Urraca, Alfonso VII, and Fernando II illustrate the consistently imperial character of the Leonese-Castilian monarchy's approach to Cluny. ⁽⁴³³⁾

As for the abbey, here too the trumpets of empire sound forth. The liturgical evidence is particularly important as bearing directly upon Fernando I. Bernard of Cluny's Consuetudines , mirroring life at the abbey within twenty years of Fernando's death, records meticulously at various points the extraordinary intercessional privileges assigned to the memory of Fredelanus Hispaniarum rex and makes patent their more than merely regal character by regularly placing them alongside the comparable honors conferred upon Cluny's other exalted reigning patrons, the Romano-Germanic Emperors Henry II and Henry III and the Empresses Adelaide and Agnes. ⁽⁴³⁴⁾ According to Bernard, of the three prebendae the abbey possessed for subsidizing the daily feeding of the poor, one was dedicated to Fernando's memory.⁽⁴³⁵⁾ The Hispanic sovereign's obit on 29 December was classed among the rare magna anniuersaria reserved for monarchs qui magnum quid contulerunt ecclesiae;⁽⁴³⁶⁾ and was one of the only five such occasions (the others commemorated the German quartet) when the monks feasted upon a full repast of fish and wine mulled with honey and spices (pigmentum ). ⁽⁴³⁷⁾ Furthermore, although the Fernandine anniversary fell within the Nativity octave, when the Burgundian monks rigidly omitted their customary celebration of the Office of the Dead, by unique exception the Spanish ruler's passing was given in this season, at Abbot Hugh's own express command full liturgical solemnization such as only Cluny's own abbots enjoyed. ⁽⁴³⁸⁾ To this exceptional privilege, which Bernard [72] states Hugh ordered pro Fredelano Hispaniarum rege qui multa bona loco Cluniacensi contulit -- the words Bernard also uses to define the magnum anniuersarium -- should be added also Fernando's inclusion in the special collect pro regibus Hispaniarum sung so frequently during the liturgical year in the daily major and matutinal masses of the abbey-church. ⁽⁴³⁹⁾ To be sure, something of all this is due to Alfonso VI's later devotion to his father's memory; but much more is involved: the explicit use of the title rex Hispaniarum and the juxtaposition of Fernando I alongside the members of the German imperial dynasties prove that the Cluniacs rated Fernando I as a ruler comparable with those of the sacrum imperium.
 

Nor is it in the liturgy alone that such recognition is to be found. It has been denied that the Burgundians ever actually applied the imperial title as such to the reges Hispaniarum, but however true this may be of Hugh and the Hugonic Vitae, by Peter the Venerable's time, at least, there was no hesitation or fear of injuring Germanic sensibilities, for the latter abbot in the preface to his Contra sectam nefandam Saracenorum calls Alfonso VII uictoriosus Hispaniarum imperator, and in a letter of 1143 to Pope Innocent II imperator Hyspanus, magnus Christiani populi princeps.⁽⁴⁴⁰⁾ Whether or not we should fall in with Professor Conant's idea that the heraldic lion carved in stone on the hospice-stable Hugh built ca. 1077-1079 as the first edifice in the vast new building program launched on the sea of Spanish censive dinars symbolizes the abbey's gratitude to its Leonese -- and, we may add, therefore imperial -- donor Alfonso VI, ⁽⁴⁴¹⁾ we do have a sure reflection of Burgundian sentiments towards the king-emperor in the magnificent late 11th-century codex of the Biblioteca Palatina of Parma, which contains San Ildefonso de Toledo's De uirginitate beatae Mariae.⁽⁴⁴²⁾ Prepared in all likelihood at Cluny her self as a fitting thank-offering to Alfonso (perhaps soon after the resumption in 1089 of the census duplicatus , with the payment of 10,000 talenta , and Hugh's visit to Burgos, or in 1090-1093 at the time of the Statuta sancti Hugonis abbatis Cluniacensis pro Alphonso rege Hispaniarum tanquam insigni benefactore ), ⁽⁴⁴³⁾ this truly imperial manuscript is copied with gold capitals on purple-dyed parchment, and contains 35 miniatures. One of these depicts St Peter's liberation from prison, possibly in allusion to Cluny's successful solicitation of the Apostle's aid in the freeing of Alfonso VI when in Sancho's prison at Burgos he too lay in uinculis. ⁽⁴⁴⁴⁾ Schapiro identifies the style of these miniatures as 11th-century German imperial, considering them painted in the abbey's own Burgundian-Romanesque rather than in its equally available Italo-Byzantine manner, ⁽⁴⁴⁵⁾ so that the whole work, in such a sumptuously imperial format as only the German rulers were accustomed to be honored with, stands as an official Cluniac acclamation of the Hispanic rex-imperator .

With so much assurance, then, that Cluniacs and Leonese-Castilian sovereigns consciously dealt with one another on the exalted imperial plane, we can advance upon two closely related and much more difficult questions. First, did Fernando and his emulator Alfonso VI enter into a genuinely vassalic or quasi-vassalic dependency upon the abbot and in effect accept the suzerainty of a foreign ecclesiastical power over the Hispanic Empire, as Sancho Ramírez did in 1068 when he swore [73] fealty to St Peter and converted the Kingdom of Aragon into a papal fief? Secondly, whether or not we can believe this is true, why should Fernando, after some eight years of pious amicitia and mutuus amor marked by doubtless substantial but still purely voluntary charities, have moved into the para-political, conceivably demeaning, and in any event more demanding, obligations of societas and census, i.e., into the second phase of his alliance with the Burgundians?

A fairly persuasive case can be made for resolving affirmatively the possible existence of some sort of quasi-feudal bond between the abbey and the Hispanic crown; but this requires to be presented at some length and with due qualification and caution. Certainly Fernando' s admission to Cluniac confraternity, by 1063 at the latest, did not in itself necessarily create anything juridically different from the normal spiritual tie between medieval laymen and a monastic corporation. Bernard of Cluny 's Consuetudines contains a chapter rubricked De societate nostra danda extraneis , which describes the procedure for receiving laymen and others into the abbey' s familia, in much the same form as it must have taken during Fernando I's own lifetime? ⁽⁴⁴⁶⁾ The applicant first requested this boon outside the chapter from the abbot or prior, either in person or through the hospitarius ; then escorted into the chapter, he entered into fraternity, which was bestowed by inscribing his name in the Liber uitae. Likewise he was granted his share in the merits of all prayers, alms and other good works performed both at the mother abbey and at all her dependencies; throughout his life he would be included in the daily prayers for the socii and benefactores at each hour of the Office and in the major and minor Masses, and in especially solemn supplications at various fixed points during the liturgical year. After his death he would be remembered in the collect Omipotens sempiterne Deus cui of the Mass and of the Office of the Dead, and on various other occasions, especially at the beginning of Lent and just after the Feasts of SS Peter and Paul and All Saints, when the great Burgundian congregation interceded for its faithful departed.

Neither Fernando I nor Alfonso VI ever crossed the Pyrenees to participate personally in such a confraternal receptio or create that special occasion mentioned by Bernard when by exception Cluny's novices were allowed to attend the Chapter: cum aliquando aliquis magnus homo propter societatem uenit in capitulo . For both king-emperors, letters and intermediaries of the type of Frater Galindus or (under Alfonso) the monk Robert must have served. That Fernando I submitted the customary request for admission is plain from the passage in his son's Burgalese diploma of 1090: societatem fratrum Deo et sancto Petro ibidem militantium humiliter expetit. ⁽⁴⁴⁸⁾ This petitio must have been followed by a reception in absentia by the Chapter, which was then reported in Spain, where, as the same text continues, Fernando (societatem) devotius accepit. Then, and only then, is it likely that the king-emperor granted the mille aurei of the census in his now lost real privilegio; such as least is the conclusion to be drawn from the two extant official texts relating to his son Alfonso VI's institution in 1077 of the census duplicatus . In one of these, the diploma issued on 10 July we have the juridical act providing for payment of the census as an hereditary obligation of the imperial Navarro-Basque house. ⁽⁴⁴⁹⁾ In the other the epistle of doubtless the same date addressed to Abbot Hugh, Alfonso acknowledges the role of the abbatial envoy Robert in winning him over to restoration of the census in doubled amount, pledges perpetual payment henceforth of this sum, requests that Robert be allowed to remain in Spain, and [74] reports royal execution of the abbot's order regarding the Roman Rite.⁽⁴⁵⁰⁾ Neither document requests or even mentions societas , while the letter's reference to Hugh's iussio on the liturgical change implies the abbot was already exercising authority over his imperial socius . That the Alfonsine census was in fact preceded by the establishment of confraternity is also shown by the king-emperor's statement in the pragmatica of 1090, where in speaking of his original concession of the 2000 metcales in 1077, he places this action after his initiation and formal conclusion of the pactum fraternae societatis cum meis Cluniacensibus. Furthermore, the emphasis he places here upon his close emulation of his father's manner of negotiation with Cluny -- sicut heres paternae dignitatis ita quoque bonae successor uoluntatis -- makes it certain that a similar formal treaty or pactum of societas existed in Fernando I's time. ⁽⁴⁵¹⁾

Cluny's socii in the 11th and 12th centuries included other important rulers than the Hispanic -- among them the Empresses Adelaide and Agnes and the emperors Henry II and Henry III of Germany, Kings Henry I of England and Pedro I of Aragon-Navarre, various dukes, counts and lesser nobles, and an unknown number of ordinary folk whom Bernard's custumal classes as pauperes in contrast to the diuites ⁽⁴⁵²⁾ As far as the princely confrater is concerned, in addition to enjoying spiritual guidance by the abbot, he might obviously seek the letter's advice on familial and other troubling questions, and even receive from him motu proprio what in 1089 Alfonso VI calls salutaria monita . But so far as we now know this is as far as the relationship normally extended. Neither side in such imperial or royal unions with the abbey can ever have believed that St Benedict's prescription of the monk's duty of absolute obedience to the abbot bound the lay socius in his temporal concerns as such. Nevertheless, in the case of the Leonese-Castilian king-emperors, confraternity unmistakably involves something more than pastoral solicitude and paternal counsel on the part of the abbot, an additional, discernibly authoritative power, which is never described for us in the documentation, and definition of which has to be sought by indirection and inference in the mutual activities of the two contracting parties. Most of the clues inevitably come from the post-Fernandine years, which introduces an element of uncertainty, even so, it seems possible to arrive at conclusions guardedly applicable to the precedent-setting reign of Fernando I along three lines of inquiry: papal and Cluniac attitudes towards temporal suzerainty over lay princes; the historical record of the abbey's interventions in political crises under Alfonso VI and Urraca; and the possible vassalic implications of stipendiary contractual ism as embodied in the census.
 

(i) The Temporal Suzerainty of St Peter at Rome and Cluny. Fernando's later years and the long reign of his son Alfonso VI coincide with the epoch in which Popes Nicholas II, Alexander II and Gregory VII were actively promoting the extension of papal feudal suzerainty over lay rulers and their kingdoms. ⁽⁴⁵³⁾ It is difficult to believe that this policy was without influence upon Cluny. Fabre has shown how Gregorian practice in this field combines reformist ecclesiological theory with various older institutions of the Roman Church [75] rooted in its history and patrimonial administration: the collection of rents and census from the emphyteutic tenures and terrae censuales of the Patrimonium s. Petri; the subordination of churches and monasteries, with or without the exemptive privilege of the libertas Romana, to the tutela or tuitio of the Holy See, commonly under obligation of a relatively nominal annual fee; the feudal ization of the meaning and terminology of such practices after ca. 1000 under the influence of trans-Alpine feudalism. ⁽⁴⁵⁴⁾ As applied from the 11th century to whole kingdoms, a vassalic relationship with Rome could develop out of a ruler's own request, as a means of escaping subjection to a neighboring state (e.g., with Hungary and Poland as against the Romano-Germanic Empire), or as legitimizing conquests (e.g., those of the Normans in Southern Italy and Sicily); but it also springs increasingly from Alexander II's time on out of active papal claims to territory based upon such grounds as previous regular payments to the Holy See (cf. the English denarius s. Petri), the Donation of Constantine, the recovery of land from Islam, and the like.

Such papal vassalage normally entailed the swearing of an oath of fealty to St Peter and formally placing one's kingdom in ius et proprietatem s. Petri; payment to Rome of an annual census, pensio, seruitium or tributum of some kind; and acceptance of some measure of papal oversight in affairs of state. The status of apostolic vassal was not considered demeaning or seriously affecting the royal prerogative; to the contrary, it gained the ruler, as the cases of Poland, Hungary, Aragon, Scotland, Portugal and others illustrate, a powerful defender of his kingdom's territorial integrity and independence. ⁽⁴⁵⁵⁾

The Aragonese enfeoffment to St Peter is particularly pertinent as the earliest example of papal suzerainty over a peninsular kingdom and as an explicitly feudal parallel to the Leonese-Castilian alliance with Cluny.⁽⁴⁵⁶⁾ The climax of a friendship dating from the time of the Barbastro War, almost certainly a counter-measure to the Fernandine conjunctio with Cluny, this rival coalition on the peninsular side embraced both the king's person and his kingdom. Alexander II's privilegio of 1071 to San Juan de la Peña says Sancho protinus semetipsum apostolicae dignitati comisit ac subdidit; by 1034-1085, in Gregory VII's bull Apostolica sedes , with its confusion of the king with his father Ramiro I, this has become: beato clauigero Petro se et regnum suum prius in Ispaniam tributarium fecit; Sancho himself, in the royal letter of 1088-1039 partially preserved in the Collectio Britannica, declares: cum annis essem uiginti quinque et iam tunc Deo uolente in honore beati Petri limina libens adii meque regnumque meum in Dei et eius potestate tradidi . ⁽⁴⁵⁷⁾ We have suggested that the confusion in Kehr and Erdmann over the customary pensio derives from their failure to allow for Sancho Ramírez' transfer to Rome of various reales monasterios, until in 1088-1089 he pledges Urban II the census of 1000 mancusos, in apparent emulation of the Leonese-Castilian stipend to Cluny. ⁽⁴⁵⁸⁾ What we unfortunately lack, despite Kehr's meticulous survey of papal relations with the Aragonese church to 1150, ⁽⁴⁵⁹⁾ is an investigation of the political functioning of St. Peter's potestas, both within the kingdom and in external questions. That it served to shield Aragon from dangerous [76] dynastic and imperial pressures from León, and advanced the progress of the Aragonese Reconquista, can probably be safely assumed. Of most immediate relevancy, however, is the impression that in internal matters affecting the Aragonese dynasty after 1068 the Roman pontiffs rarely exercised their suzerain authority, and this phenomenon stands in marked contrast with the frequency and extended range of the Cluniac abbot's involvements, also possibly on feudal grounds, in the kingdom of Alfonso VI and Urraca.

That in this 11th-century world of expanding papal suzerainty over a growing number of principalities Cluny also might seek in St Peter's name to institutionalize her partnership with Fernando I and Alfonso VI, is a tenable enough hypothesis. It is important to remember that the abbey's penetration of the Leonese-Castilian state precedes by some twenty years the letter's establishment of communication with the reform papacy, and that in this interval Cluny represented the Prince of the Apostles in the Iberian West and her abbot possessed for Fernando and his subjects something of the prestige and veneration elsewhere accorded the bishop of Rome. We know that Abbot Hugh was seeking to expand Cluniac societas among secular churchmen and laymen; ⁽⁴⁶⁰⁾ he was, of course, from his Burgundian surroundings and the abbey's own local lay dependents, thoroughly familiar with feudal contractual practices. ⁽⁴⁶¹⁾ The signs that such an outlook could extend across the Pyrenees can be found in the two Catalan grants we have already discussed: Arnau Mir de Tost's cession of not alone his Church of San Pedro but all the comarca of Ager, in 1066.;⁽⁴⁶²⁾ and Girbert Mir's bequest of 1073 (executed in 1074) of Berano and Roda.⁽⁴⁶³⁾ These texts envisage the Cluniac abbot in the exercise of typically feudal suzerainty over lands held of him as senior by fideles or homines comandi who owe him adfidamentum ,  fidelitas , securitas castrorum; and they assign him such secular power in dangerous frontier zones of the Reconquista. It has been argued that the failure of both these benefactions to be realized proves Hugh's adamant refusal to undertake feudal responsibilities in Spain; ⁽⁴⁶⁴⁾ but with equal or even greater plausibility it may be contended that the real reason why neither cession succeeded was the opposition of, first, the immediate overlords of Arnau Mir and Girbert, and secondly, of the Roman Church, which had previously been given rights in both cases. ⁽⁴⁶⁵⁾ Cluny's negotiation of the act of donation of 1066, the appearance at Ager in this connection of so eminent and presumably well instructed an envoy as Frater Henricus, the careful preservation in the Burgundian archive of both the Catalan privilegios of donation, all point, in the context of Cluny's general expansionism at the time, to Hugh's probable willingness to accept for St Peter and his abbey a feudalized temporal jurisdiction within the Iberian Peninsula. And if in Catalonia, why not also in the West, in Leon-Castile?
 

(ii) The Imperial-Papal Crisis of 1077 and its Fernandine Counterpar t . The interesting fact that there can be discerned in Fernando's and Alfonso VI 's connections with Cluny two successive phases, first of pious amicitia and then binding conjunctio, makes it imperative to try to obtain from the son's better documented reign [77] a clue to the nature of the alliance through analysis of the circumstances and motives that might persuade a Leonese-Castilian king-emperor to form such a partnership, Alfonso, from his accession at the commencement of 1066 down through the deposition and restoration of 1072 and all the way to the spring of 1077, was at most the abbey's amicus, certainly not its socius ; and in this period the Fernandine census was in complete abeyance. For the earlier troubled years 1066-1071 no Cluniac signs appear. These commence only in the spring or early summer of 1072 when Abbot Hugh for the first time, so far as we can tell, intervenes in the fratricidal struggles among Fernando's three sons by responding to an appeal from the Leonese clergy and nobility. At this point, in the aftermath of Alfonso VI's defeat at Golpejera in January and imprisonment in chains at Burgos, the abbot orders his monks to pray for the deposed king-emperor, sends as his envoy to Sancho the former bishop Ximeno of Burgos, then resident at Cluny, and, through what tradition came to regard as the direct assistance of the Apostle St Peter, induces the understandably recalcitrant Castilian monarch to allow his brother to go free into exile at al-Ma'mun's Toledo. ⁽⁴⁶⁶⁾ Since Sancho's assassination in October before the walls of Zamora by a Leonese magnate paved the way for Alfonso to return north and assume the rule of all three subdivisions of his father's now reunited kingdom, it is not hard to understand how deeply grateful he must have felt to the monks for the release which made possible his miraculous restoration, or how eagerly the Burgundians might seek to have this feeling take the material form of a revival of the Fernandine census.

Then in 1072-1073 we encounter proof of Cluny's at least partial responsibility for the blocking of a second projected Aragonese-papal crusade, without doubt directed towards Ribagorza. David has rightly, if by no means at various points convincingly, called renewed attention to this poorly known re-play of Barbastro, sponsored by Alexander II, inherited at the very start of his pontificate by Gregory VII, which for obscure reasons failed to take the field as planned under its designated commander Count Ebles de Roucy.⁽⁴⁶⁷⁾ Two aspects of this enterprise are of immediate pertinence as disclosing Hugh's continuing defence of Leonese-Castilian interests in the Cinca valley. One is Alexander's appointment of Bishop Gerald of Ostia as his principal legate (along with the archdeacon Raimbald) for launching this military operation, with jurisdiction for this purpose in both France and Spain. ⁽⁴⁶⁸⁾ Gerald was a Cluniac, the former grand-prior of the abbey and thus an intimate of the abbot. David sees in this choice an attempted appeasement of the Cluniacs but fails to explain the reason; actually it must have been intended to provide assurance, perhaps originally to Sancho II and at any rate after October 1072 to Alfonso VI that the Cluniac presence would minimize any threat to the political and financial stake in Zaragoza, the Taifa kingdom Fernando I's partition plan had assigned to control by Burgos. ⁽⁴⁶⁹⁾ The other suggestive fact is that on 30 April 1073, just eight days after his elevation to the chair of St Peter, Gregory VII relieved Gerald all authority in Spain and entrusted this responsibility to that old counsellor of Sancho Ramírez and forger of the Jacan-Roman coalition, Cardinal Hugh Cándidus, although the Cluniac abbot was permitted to [78] supply advisers to the new peninsular legate. ⁽⁴⁷⁰⁾ This looks like a victory of the pro-Aragonese forces in Rome; and while we have no way of telling what followed, the collapse of the whole expeditionary scheme points to a struggle in which the Cluniacs successfully defended the imperial Hispanic cause by helping to wreck the whole international project.

These twin Burgundian involvements in the civil war of 1071-1072 and the abortive crusade of 1072-1073 set the stage for Alfonso VI's henceforth warmly friendly relations with the abbey between 1073 and 1077. This is the era in which for the first time an abbatial representative is stationed in Spain, the highly adept monk Robert, ⁽⁴⁷¹⁾ and the grateful monarch cedes Cluny four reales monasterios : San Isidro de Dueñas (29 December 1073), San Salvador de Palaz del Rey (27 August 1075-1076), San Torcuato de Astudillo (31 January 1077), San Juan de Hérmedes de Cerrato (22 May 1077). ⁽⁴⁷²⁾ Nevertheless, as late as the last of these transfers, just as with Fernando I prior to 1063, there still existed neither societas nor census, although Robert was plainly pressing vigorously for both. But by July of this same year, less than two months after the gift of Hérmedes de Cerrato, Alfonso VI suddenly terminates his policy of monastic cessions in León, enters into confraternity with Cluny, and revives his father's annual subsidy in doubled amount. How can this basic shift from amicitia to a confraternal-censive union be accounted for?

The explanation is unquestionably linked to the formidable attack by Pope Gregory VII upon the ideological premises and objectives of the imperial Hispanic program, an attack that reaches its climax in 1077 in the form of papal claims to sovereign political authority and tributary rights over the entire Iberian Peninsula. ⁽⁴⁷³⁾ The origins of these demands can be traced back to Alexander II's extension to Spain of Rome's interest in bringing lay rulers under the temporal stipendiary lordship of the Holy See. Whether or not this goes back as far as the War of Barbastro -- which is quite likely and warns against too ready acceptance of Menéndez Pidal's hypothesis that the crusaders in 1064 placed their conquests under Aragonese suzerainty -- it is certain that by the time of the second Ribagorzan crusade in 1072-1073 this pope was insisting upon papal suzerainty over at least all territory conquered in Spain from the Muslims as a stipulation in the pactio or pactiones being negotiated between him and the French barons recruited for the expedition. ⁽⁴⁷⁴⁾ With Gregory VII this doctrine, which of course had since 1068 a powerful basis in the actual subordination of the Aragonese kingdom to Petrine suzerainty, becomes a sweeping assertion of similar authority over all other peninsular states and inevitably leads to bitter conflict with the most prominent ruler whose independence and sovereignty were threatened, the Leonese-Castilian king-emperor -- just as in trans-Pyrenean Europe it was another emperor not unfriendly to Cluny, Henry IV, whose political position was most critically endangered by Gregory VII's decrees of 1075 against lay investiture. ⁽⁴⁷⁵⁾

The first visible sign of the gathering storm is the adoption on the part of the Leonese-Castilian chancery by at least 27 March of a new defensive form of the imperial title: imperator totius Hispaniae a formula Menéndez Pidal and David agree should be viewed as a [79] determined rejection of all papal claims to suzerainty over Spain.⁽⁴⁷⁶⁾ Then on 28 June comes the thunderbolt from Rome: Gregory VII 's long, closely argued epistle addressed to the kings, counts and other princes of Hispania, in which he expounds the legal foundation, comprehensive scope and divine sanction of Petrine temporal supremacy in the Peninsula. ⁽⁴⁷⁷⁾ Under ancient constitutiones (presumably the Donation of Constantine), the pope declares, the regnum Hispanie belongs to St Peter and the Holy Roman Church in ius et proprietatem; and despite prolonged occupation by the Muslims and the failure of earlier pontiffs to press their rights, it still owes the tribute (seruitium ) it formerly paid the Holy See. Since the progress of the Reconquista makes early recognition of such papal proprietorship indispensable, two envoys, Bishop Amatus of Oloron and Abbot Frotardus of Saint-Pons de Thomières, are being sent to Spain to secure this; and all peninsular rulers are reminded that their salvation is bound up with acknowledgement of the apostolic authority. The formidable character of this text, so threatening, as it must have seemed, to the imperial court, has not always been sufficiently appreciated. With its forceful juridical and historical justification, its implicit resolution to extend to all Spanish princes, including notably the rex Hispaniae whose regnum is the object of the demands, the vassalic tributary bond subordinating the kingdom and territory of Aragon to St Peter, its scarcely veiled threat of excommunication in the event of defiance, the document is in effect an ultimatum: a powerful direct assault upon the ideological tenets, political position, future ambitions, and prized tributary resources of the Leonese-Castilian state in the imperial form given it by Fernando I's Leonicization and continued by Alfonso VI. It is not hard to understand why three years of bitter strife between pope and king-emperor followed.

On 10 July 1077, two weeks after the promulgation at Rome of Gregory VII's epistle to the Spanish rulers, and so close in date to that ultimatum as to prove that the Leonese-Castilian court anticipated the early reception of such a papal missive, Alfonso VI issued the real privilegio pledging Cluny the census duplicatus. ⁽⁴⁷⁸⁾ It is at this point, as we have said, not in 1066 or 1072, that he first becomes the abbey's socius; in his own words in the pragmatica of 1090 describing his action of 1077; pactum fraternae societatis cum meis Cluniacensibus inii, statui, firmaui, censum quoque largitatis paternae duplicaui.⁽⁴⁷⁹⁾ Writing in this same summer of 1077 to Abbot Hugh, the king-emperor ascribes the addition of a second thousand gold pieces beyond the old Fernandine stipend to Robert's indefatigable urgings: mihi die hac nocte addere suasit ;⁽⁴⁸⁰⁾ but the revival of the census as such and the entry into Burgundian confraternity were much more the achievement, however involuntary, of Gregory VII. Alfonso's move on of 10 July has to be assessed in the context of the imperial-papal crisis and of the time with, its most critical phase falling between 22 May 1077 and 10 July. Down to the former date Alfonso VI, for all his generous friendship towards the Cluniacs, sedulously avoided following his father into societas and censualitas , preferring over the four previous years that followed his restoration to reward the abbey not in dinars but with the four Leonese monasteries. As late as 22 May, with the gift of Hérmedes de Cerrato, this was still the policy. Then [80] comes on 10 July the doubled census; and henceforth for the remaining thirty-two years of the reign -- with the possible exception of a projected transfer of Sahagún ca. 1080 -- no further such cessions in León proper occur. ⁽⁴⁸¹⁾

In the shift from monastic to monetary benefactions within the Leonese kingdom, the entry into societas, the doubling of the census, thus stands embodied Alfonso VI's decision, taken in the face of Gregory VII's critical challenge to the viability of the whole imperial Hispanic program, to form the tighter, legally constituted union with Cluny, the formal conjunctio which, while still entailing some degree of undesirable personal and perhaps national subordination to a foreign ecclesiastical entity, was plainly far more tolerable than the Petrine dependency the papacy sought to impose. Unlike the Roman pontiff's fragmentizing concept of individual Iberian rulers, each acknowledging like Sancho Ramírez (who in 1076 had just added to the crown of Navarre that of Aragon) ⁽⁴⁸²⁾ immediate fealty to the Holy See, attachment to Cluny left inviolate the doctrine of imperial hegemony over tota Hispania . for the abbey would be content with one trans-Pyrenean ruler as its chief benefactor and ally, the rex-imperator of the Spains. Alone of ecclesiastical centers in Western Europe, the Burgundian congregation could effectively shield the Hispanic Empire against the extreme claims and pressures of papal temporal policy, on the grounds of its own prior protectorate over Iberia. The new census duplicatus paid to St Peter at Cluny negated demands for a papal seruitium; and Alfonso could be confident that at Rome Abbot Hugh and his influential spokesmen would act to halt further unwelcome schemes of the Aragonese-papal coalition aimed at Ribagorza or other areas in which León-Castile was interested.

Much of the foregoing unfortunately remains surmise; what is certain is that the new conjunctio must have proved an effective bulwark for the Hispanic Empire in the crisis of 1077. Imperial relations with Rome remain painfully strained for several years thereafter, reaching a new peak of acute tension in 1080; but this time it is the liturgical issue of the Roman Rite, not the anti-imperial demand for Petrine suzerainty, that is the cause of dispute. ⁽⁴⁸³⁾ As far as the confrontation of 1077 is concerned, that threat faded, perhaps before the end of the year, as alliance with Cluny blunted the papal thrust. To change the metaphor, the camino real to Burgundy made unnecessary a trip to Canossa.

As Gregory VII drove Alfonso VI into Cluny's confraternity, so in all probability his predecessor Alexander II inspired Fernando I's original institution of such a connection across the Pyrenees. Between the crises of 1063 and 1077 lie many differences but one overriding common factor: papal intervention in Iberian inter-state relations and in the Reconquista, which was sharply hostile to the policies of the Leonese-Castilian monarchy. It suffices to recall the situation on the eve of the War of Barbastro: Ramiro I's defiant assault at Graus upon Fernando's ally and tributary King al-Muqtadir of Zaragoza and his Castilian reinforcements; the newly acceded Sane Ramírez' turn to Alexander II for the recruitment of a French baronial army to launch a crusade in Ribagorza; the grave peril of Fernandine involvement against the Christian forces and the possible loss of the [81] imperial position in the Cinca Valley and on the Middle Ebro. In this conjuncture the monarch takes the step from which he had hitherto abstained, converting himself from amicus into socius of Cluny and pledging the perpetual subsidy of the 1000 metcales that would give the abbey the strongest possible material motives for protecting the tributary revenues of the imperial erarium against all encroachers, including the Roman Church. How vital a contribution in consequence the Burgundian monks made either in the diversion of the expedition from Graus to the rebellious Léridan stronghold of Barbastro or to the eventual ruin of the whole enterprise we have no way of determining; at any rate, Aragonese-papal hopes were shattered, the imperial interests preserved. Revealing above all is the fact that in 1063 as in 1077 the imperial formula is identical: resort to confraternal-censive conjunctio with Cluny as the checkmate to Rome.
 

(iii) Census and Clientage: The Para-Vassalic Relationship. If both Fernando I and Alfonso VI, in spite of many earlier charities, thus deferred entering into Burgundian confraternity until compelled to do so for reasons of state under the pressure of major crises, the factors explaining such reluctance, whether constitutional or fiscal or both, require to be isolated. That much more is involved in the Hispanic link with Cluny than in the conventional societas the abbey maintained with its other princely and noble confratres , becomes clear immediately from the record under Alfonso after 1077 of the frequency and suggestively authoritative nature of the abbot's interventions in the familial affairs of the monarchy. The signs of this commence in 1079 when Alfonso wed as his third wife Constance of Burgundy, Hugh's own niece, a marriage that has all the appearance of those so frequently arranged between royal families to seal a treaty of alliance.Then comes, during what David calls 'la crise de 1080', the anti-papal collaboration of the new queen-empress and the monk Robert of Cluny, for a long time highly successful, in defence of the king-emperor and the Leonese-Castilian church against Gregory VII's insistence upon abolition of the Hispanic Rite. ⁽⁴⁸⁵⁾ As David has shown, the two Burgundians ruined the success of the legatine mission of Cardinal Richard of Saint-Victor de Marseilles, evoked the wrathful Gregory's condemnation of Constance's marriage as canonically invalid and of Robert's inspiration as diabolic, and only surrendered when Hugh himself found it advisable to yield and gave Alfonso VI the command ( iussio) to accept the Roman Rite in what the ruler called his terra admodum desolata.⁽⁴⁸⁶⁾ The Cluniac alliance can also safely be credited with the advent, in 1087, after the Almoravid victory at Zallaca, of the first Burgundian knights to serve under Leon-Castile as active participants in the struggle against the infidel.⁽⁴⁸⁷⁾

Above all, however, it is in connection with the dynastic succsession crisis created by Constance's failure to produce a male heir, and Alfonso's several attempts to solve this problem, that Cluniac participation in domestic affairs is prominent from ca. 1090 on. Behind the marriages of Alfonso VI's daughters to two Burgundian counts-- Urraca's  in ca.1087-1091 to Raymond, Teresa's in ca. 1094-1095 to Henry -- would seem to lie, as least in part, moves by the abbot to reinforce the Burgundian orientation of the dynasty and safeguard [82] the future of the conjunctio. ⁽⁴⁸⁸⁾ Indicative of Hugh's exercise of political power in the kingdom is the use of his deputy in Spain, the abbatial chamberlain Dalmace Geret, to bring about the celebrated successoral pact of 1105-1106 between the two Burgundian sons-in-law of the king-emperor. ⁽⁴⁸⁹⁾ This agreement aimed to assure Raymond of the throne while partitioning the Empire and the treasure of Toledo between him and County Henry. Here the abbot is actually undertaking to impose a Cluniac solution of the succession question, backing Raymond against Alfonso VI himself, who had by this time discarded his original preference for a condominium of Urraca and Raymond in favor of the rule of Sanchuelo, his son by the conversa Zaida, quondam daughter-in-law of King al-Mu'tamid of Seville. ⁽⁴⁹⁰⁾ This divergence of opinion discloses that Alfonso by no means always submitted meekly to abbatial commands. But it is to be noted that when Raymond's death in 1107 and Sanchuelo's in 1108 made Urraca once more the heiress apparent, the aged monarch not only moved swiftly to arrange his daughter's marriage with King Alfonso I of Aragon, but made certain of Cluny's acceptance of her succession by having the Infanta, on 22 February 1109, cede the Burgundian monks the monastery of San Vicente de Pombeiro, in token of her intention to maintain the conjunctio in the next reign. ⁽⁴⁹¹⁾ This was no idle gesture: it secured Cluniac approval of D. Urraca in the dark, troubled years of matrimonial incompatibility and civil strife that followed Alfonso VI's death; and at a time when Archbishop Bernard of Toledo and most of the Leonese-Castilian hierarchy refused to recognize the canonicity of the marriage, Abbot Ponce de Melgeuil loyally supported the queen-empress. Only in 1113, when he came to Spain as the legate of Paschal II, did the governor of Cluny abandon this fruitless policy and join the ultimately victorious Alfonsoraimundist party of Count Pedro Froílaz of Galicia and Bishop Gelmírez of Compostela, which was willing to accept an Urraca separated from her Aragonese spouse on condition that she be succeeded by her half-Burgundian son, the future Alfonso VII.We need only mention, to conclude this hasty review of the Cluniac factor in imperial politics in the first half of the 12th century, Urraca's later tribute, patently for aid in secular matters, to the chamberlain Stephen as fidelissimus aimcus meus;⁽⁴⁹³⁾ and the abbey's involvement during the critical years of 1126-1127 in strengthening Alfonso VII and the cause of Leonese-Castilian unity against Queen Teresa of Portugal's attempted consolidation of a secessionist Luso-Gallegan kingdom spanning both banks of the River Miño. ⁽⁴⁹⁴⁾

Still another clue to the position of the Burgundian abbot vis-à-vis his imperial socius is the potentially feudal connotation of Alfonso VI's language in the epistles and privilegios addressed to Hugh, even if we allow for some of this as possibly due to Cluniac scribes, and avoid too literal an exegesis of the deferential, not to say at times obsequious, terminology. ⁽⁴⁹⁵⁾ For the king-emperor in 1077 to address his revered correspondent as uenerabilis et excellentissimus abbas, gloriosissimus pater, egregius pater, is innocuous enough; but what are we to make of the approach to him as a senior whose filius regards him cum omni deuotione mentis et corporis, terminology strongly reminiscent of feudal contracts even if the effect of senior is lessened by joining to it the phrase cuncte dulcedinis mellifluus? ⁽⁴⁹⁶⁾ In 1089, at a time, to be sure, of self-exculpation, the royal expressions are even stronger: uenerandus pater et dilectissimus dominus suus dominus abbas, pater atque patrunus. maiestatis uestre preclara benignitas , and the like.⁽⁴⁹⁷⁾ In referring to the imposition of the Roman Rite in 1077, Alfonso speaks of obeying Hugh's command (iussio); ⁽⁴⁹⁸⁾ in 1089 he mentions not only the salutaria monita but the merited castigation he has received from Burgundy. ⁽⁴⁹⁹⁾ Senior, patronus, iussio: these expressions may well incorporate more than merely spiritual or metaphorically filial submission.

Finally, there stands, as the most distinctive element of the conjunctio in contrast with the abbey's normal confraternal arrangements, the substantial annual subsidy in gold, whether this be the Fernandine census of 1000 or the Alfonsine of 2000 metcales . Medieval usage associated payment of any regular levy with dependency upon a seigneurial or domanial superior; and as a contractual obligation, even if self-imposed, the stipend to Cluny carried this implication, just as Alexander II and Gregory VII interpret the denarius s. Petri as visible proof of England's submission to papal suzerainty. ⁽⁵⁰⁰⁾ Certainly by the mid-12th century Cluniac tradition held that Alfonso VI had been in this same sense the abbey's vassal. Peter the Venerable, writing ca. 1142 in his Liber de miraculis, can say of Alfonso: magnificentissimus et famosus rex censualem se regnumque suum Christi pauperibus eiusdem Christi amore fecerat; ⁽⁵⁰¹⁾ and these words have numerous parallels in papal documents relating to fealty of lay rulers.⁽⁵⁰²⁾ So too the anonymous Alia miraculorum quorundam s. Hugonis abbatis relatio calls the census Alfonso's tributum. ⁽⁵⁰³⁾ The king-emperor himself, when threatening with damnation and deposition any of his successors who should deny Cluny the census duplicatus or reduce it in amount, declares that any arrears must be compounded for by suitable pensiones, the favorite term employed by the chanceries of Alexander II and Gregory VII for payments due the Holy See from subject principalities. ⁽⁵⁰⁴⁾ Very likely the fact that the Hispanic gold pieces were drawn, even if by way of the imperial erarium, out of the parias of the Taifas, helped give the census the look of a vassalic tributary offering. Suggestive also is the attitude of Alfonso VII, when, in August 1142, he negotiated with Abbot Peter the drastic reduction in amount of his grandfather's census. The royal diploma incorporating this agreement admits that Fernando and Alfonso VI had paid the subsidy de redditibus sui regni censualiter , but at we same time defines it as elemosina , i.e., free alms, a voluntary charity. ⁽⁵⁰⁵⁾ since Alfonso VII was seeking to escape from what had become, with the disappearance of the parias, an intolerable financial burden, his view of the census, therefore, carries no more weight than does the emphatically feudal meaning assigned it by Peter, whose own interest naturally lay in emphasizing the imperial obligation to provide a revenue that the bankrupt abbey then needed far more than ever before.

Yet there can be little doubt that the Venerable's conception of an authentic subordination of some sort to Cluny, whether affecting only the king-emperor's person alone or his kingdom as well, accords far more accurately with the historical record than does el Emperador's contrary tenet of an eleemosynary voluntarism. It would be [84] fruitless to try for a precise definition of a relationship which the contracting parties themselves never couched in formal, judicial terms; nor should we lose sight of the possibility that the Cluniacs, the product of a pervasively feudal society, and well informed of the contemporary papal resort to the feudal formula for institutionalizing relations with lay rulers, would naturally incline towards giving the conjunctio a vassalatico-beneficial interpretation such as the Hispanic monarchs neither intended nor acknowledged. Whatever the nature of the bond, furthermore, it can safely be taken as much looser, more vaguely envisaged, less suggestive of abbatial interventionist rights in Fernando's time than with his personally weaker and more harassed son. Within these cautionary limits one conclusion at least seems firm enough: neither under Fernando nor Alfonso VI is the Leonese-Castilian state validly classifiable as a Cluniac fief, as can be said of Aragon's subjection to the papacy. Fernando's promise of the census of 1063, Alfonso's of the census duplicatus in 1077, created a binding obligation; and the ratification of this in 1090 by the primate and hierarchy of the Leonese-Castilian Church and the magnates of the realm -- assuming such formal assent was actually given, which is almost but not altogether certain -- further established the annual stipend as a veritable national commitment. This is indeed what Abbot Peter's phrase censualem se regnumque suum . . . fecerat must mean. But the gold subsidy was not accompanied even in rudimentary form by any prestimonial or beneficial submission of the kingdom to the abbot; with reference to Cluny the Fernandine-Alfonsine regnum Hispaniae was zinspflichtig but never lehnsabhängig .

The personal connection of king-emperor with abbot and monks within the framework of societas and censualitas is however another matter, and plainly the true cement of the imperial alliance. No act of homage, no oath of fealty, no formal commendatio such as characterize the recognition of papal lordship by Sancho Ramírez and his successors, can be found, although something of this might be implicit in the petitio of confraternity and what Alfonso VI calls the pactum societatis. Yet feudal or para-feudal overtones are impossible to ignore in this vitalicial nexus which, as with that between señior and vassal, required to be renewed with each new king-emperor, even while the censive obligation, like the fief, passed automatically to the heir. No other royal or noble socius of Cluny save King Henry I of England, who at a later date pledged a much smaller amount in different circumstances, ever found himself permanently liable to the abbey for a substantial pensio like the Hispanic census. Nowhere does the Cluniac counterpart of the Gregorian legate, the abbatial chamberlain, appear so early, with such frequency and with such decisive importance in a kingdom's intimate affairs as in Leon-Castile. ⁽⁵⁰⁶⁾ In no other medieval kingdom do we witness the many temporal interventions of the Burgundian abbot. Since neither Fernando I nor Alfonso VI entered into the second, tighter phase of their already existing friendship until driven to do so by papal anti-imperial moves, they were fully conscious of the personal and political as well as financial connotations of a dependency resting upon societas-censualitas but came to accept these in order to obtain a juridically impregnable shield against Rome. Thus everything subsumes a status of monarch to [85] abbot unmistakably ultra-confraternal in nature, although we must also