THE break-up of the Carolingian empire gave birth to new national groupings and a new political structure in western Europe: it was the occasion also of the rise of a new kind of society: feudal society. The years 888 to 911 were years of political insecurity and military chaos: not only were the claims of new aspirants to rule, of Carolingian descent or otherwise, liable to provoke civil war, but, even more, the attacks of outside enemies, raiders by sea and by land, rendered life and property insecure. The northmen had already by 888 penetrated and plundered the land of the west Franks and to a less extent that of the east Franks: the Saracens had raided Italy; but as great a danger was now to threaten the land of the east Franks, the raids of the nomad horsemen from the steppes, the Magyars who were later to be called Hungarians. Pannonia, Bavaria, Suabia, Franconia and even Saxony in the north of Germany were to suffer from raids as cruel as those of the Vikings; in the raids on Saxony in 906, women were driven away naked from their homes, tied to one another by their hair. The Carolingian counts and dukes of Germany made a better showing in defence than the Carolingian emperors and kings had done earlier in the west against the north-men: they fought grim battles, but met with heavy defeats. Military insecurity in Merovingian times had made bands of fighting retainers necessary to the Frankish kings; but the wars of the Carolingians, and above all those of the latter part of the ninth century, made the need of such armed retainers even more urgent. The origins of feudal society, and ‘feudalism’ as a legal system, have been the subject of prolonged research, and the weightiest modern historical opinion now sees in this need of armed local defence, this need by kings, dukes and counts for armed retainers, for vassals, the factor that produced the new social order. Professor F. L. Ganshof in his Qu’est-ce-que la féodalité? (translated by Professor P. Grierson under the title Feudalism), has set forth the rise and consequences of the institution of vassalage so clearly and convincingly, that the matter can here be treated briefly.
In the period between the collapse of the Roman empire and the rise of a feudal society in the ninth century, certain institutions and terms were in use that can properly be called ‘pre-feudal’: they foreshadowed the personal relationship between a freeman and his dominus that was to be the heart of vassalage. There was in the period a shortage of coined money such as precluded the payment of high officials by a cash salary, such as had been possible under the Roman empire: such officials must needs be rewarded by maintenance at the lord’s court, or some kind of a lease or grant of land. There was also a need for order and local peace: for the opportunity for peasants and townsmen to till the land and pursue their affairs with safety to life and harvest. There was need for better justice, for better and more modern courts than the old mallus, presided over by the counts, or the rare circuits of the Carolingian missi. The penal clause in an old Roman will of land had begun: Si quis, and laid down that If anyone tried to prove this testament ineffective, let him pay so much to the Roman fisc and lose his case; but when there was no longer a Roman fisc to protect testaments, the notaries tried to substitute a Christian penalty: Si quis (they wrote): If anyone try to upset this my last will, or prevent these lands being given to my wife, or to such and such an abbey for the remedy of my soul: let him be cut off from the communion of the faithful in this life: and let him hear the words of the Lord when he comes to judge the world by fire:
Depart, ye accursed.
A terrible tribunal this, and a terrible penalty: but, as all notaries knew, less efficacious than the certainty of having to render so much to the fisc. The need of secure justice, which in the weakness of central power had to be local justice, reinforced the more pressing need of bands of armed warriors for local and personal defence.
The pre-feudal institutions of Merovingian times had an ancestry going back both to the Roman and Germanic past. By the Roman institution of patronage (patrocinium) the head of a patrician family had absolute rights over his own family and descendants, and also over a train of outsiders, clientes; the Roman senator, too, exercised a patrocinium. He was bound to protect these clientes in the law courts, and they gave him social deference; by the end of the fourth century they had come to include also small rural proprietors, who afforded protection to their villagers from the extortions of the tax collector. The emperor had his own clientes, and they, as his companions and friends (comites and amici) formed his entourage (comitiva). These clients, however, incurred no military obligation.
The beneficium again, so important an institution in feudal society, had a Roman origin. It was the free gift (either of land or office) by a donor not bound by any contract to make such a gift, and as such it could be recalled by the donor. Roman law enforced all contracts freely made, as it enforced all promises made before legal witnesses in the courts; but in some cases, and especially after the fall of the empire, it was desirable to afford protection, or to promise services, without the making of a legal contract: that is, to enter into a personal relationship where the circumstances were unsuitable for the making of a legal contract. Similarly, a revocable grant of land, a precarium, was extra-legal. Ulpian in the Digest defined it as ‘that which is granted at the prayer of a petitioner to his use’. The holder of a precarium got, not the ownership, but the use of the land. Neither the beneficium nor the precarium was hereditary. The grantor of the precarium profited by his right to take part of the produce of the land, and, by joining these precarial tenancies to his own estate, increase his lands and defend them more advantageously from the tax-collector. No military service was involved. Both beneficium and precarium were in use in the Frankish kingdoms in Merovingian times.
The Germanic elements of feudalism came with the invaders of the west, and had their new development in the kingdoms of the Merovingian Franks: they should be looked for (as Professor Ganshof points out) ‘more particularly in the heart of the Merovingian kingdom between the Loire and the Rhine’. Gaul was rarely united under a single descendant of Clovis and family partitions brought frequent wars. Even when a civil war was not raging, the Frankish officials were too few to secure public peace and safety: magnates and kings sought retainers for protection as well as prestige, retainers who should fulfil the function of the old Germanic comitatus. The Frankish trustis appears to have corresponded to the comitatus, and the freemen who thus served the king were his antrustiones and had a wergeld threefold that of the ordinary freeman. They were also termed fideles, leudes and, in Italy, gasindi: and they were essentially picked fighting men. Only the (Frankish) kings and queens had antrustions: they had also servants of lesser social rank to whom the Latinized Celtic word vassus was applied: it came from gwas, a boy, servant or slave. Those under the king’s protection, his patrocinium, were now said to be under his mundeburdis or tnundium: a freeman might place himself voluntarily under the mundium of king or another, and such an act was made by a specific formula and had legal force. The petitioner asked for protection and maintenance, and entered into a lifelong relationship: he ‘commended’ himself to his new dominus, but he remained a freeman. Clerks sometimes sought the initio or patronage of a layman, which made the bishops fear they would be withdrawn from their own authority; and, on the other hand, bishops often maintained a household whom they supported, including laymen, and these formed, as it were, an episcopal comitiva. Commendation among laymen was in fact a contract made by two partners: the earliest extant formulae of commendation do not specify the services required from the vassus in any precise terms: they might be military or domestic or agricultural, according to the social status of the petitioner. The social rank of the vassus in early Merovingian times was low, for the word implied slavery: but the status of the vassus was to rise and to include men of high rank in the king’s service.
To the end of Merovingian times there was no necessary and normal connexion between the benefice, the precaria (Merovingian for precarium) and the vassus. The king might bestow a benefice, secular or spiritual office or land, without endowing the recipient with precarial land: he might and often did maintain him in his own entourage till he had performed many years of service, and then make him a grant of land, absolutely or as a precaria. The churches and abbeys often granted precarial land, expecting in return a part of the produce. Benefice and the precaria were still separate. Nor did the grant of either necessarily accompany commendation and the entry into the state of vassalage: the would-be vassus might commend himself for protection only. Professor Ganshof holds that beneficial tenements were sometimes granted in this period to vassals to ensure their maintenance: but that there is not enough evidence to show that the practice was very widespread before the middle of the eighth century.
Under the Carolingians, however, vassalage and benefice are commonly found united and in extended use. The century from c. 750 to c. 850 was one of almost continuous wars, first aggressive wars against the Alemanni, Bavarians, Aquitanians, Lombards and Saxons, and then wars of defence against northmen, Saracens, and later, the Hungarians. The early Carolingians increased the number of their vassals and maintained them by landed grants, in some cases from their own estates but mainly from lands once the property of churches and abbeys. Complaints of injustice and of clerical irregularity consequent on these confiscations followed (see p. 290), and Pepin the Short and Boniface had to reconcile military needs with the demands for ecclesiastical reform: their solution marked a great increase in beneficial tenure. It was arranged that all secularized lands should be held by the mayor of the palace (744) and later by the king, and that he should grant them as benefices to their present holders; at the same time, the holders, the king’s vassals, should pay a precarial rent to the church who originally owned the land. Further, in the mid-century Pepin the Short ordained a divisio of the land of all the west Frankish churches: a part remained in their effective possession, but the larger part remained to the mayor of the palace or the king to grant as benefices to his vassals. It was as a sort of compensation to the west Frankish church that Pepin ordered that tithe should be paid to the church by all his subjects. Even this treatment of the secularized lands was insufficient to secure Pepin enough military support, and he and the great magnates, dukes, counts, margraves, etc., granted many benefices to their vassals from their own or conquered lands. By the accession of Charlemagne, the union of benefice and vassalage had become common: and there was a marked rise in the social status of the vassal.
Under Charlemagne and his successors there was a further extension of vassalage, normally now beneficed vassalage. The institution spread from the region between Loire and Rhine into Franconia, Aquitaine, Alemannia and Bavaria; in Lombardy it was influenced by the position of the Lombard gasindus. Great estates bestowed by the king (villae: maneria) were sublet to the grantee’s own sub-vassals, and cultivated by peasants rendering food and labour services. The Carolingian rulers encouraged this great spread of vassalage, for it gave them military support: the vassi were horsemen, cavalry, the spearhead of the Frankish army; the Frankish kings now, moreover, had vassals planted out in all the countryside forming a local governing class whom they could have paid in no other way. They could correspond with counts and other vassals and get their orders carried out locally; the importance of the vassals’ work and their increase in numbers is attested by the frequent reference to them in capitularies, charters and annals. Even in the creation of sub-vassals, owing fidelity to their immediate lords, the Frankish rulers saw no danger.
The word vassal was now in general use, and as covering dukes, counts and the lesser vassals and sub-vassals: it embraced a large class, including all those who had put their joined hands between their lord’s hands in the rite of homage (the inmixtio manuum). The word homo was often used in the technical sense of vassal, as was the term miles (soldier and knight) from the second half of the ninth century; the vassal might owe his lord administrative and other services, but the military obligation had become paramount. The wars of Charlemagne increased not only the number of the vassals, but their rise in status, for the heads of the subjugated states themselves signified their submission, and the incorporation of their land in Charles’s kingdom, by becoming his vassals. Tassilo III, duke of Bavaria, had become the vassal of Pepin the Short in 757, commending himself into his hands and swearing oaths on holy relics; a Danish king became the vassal of Louis the Pious, etc.
The act of commendation was accompanied, as is shown in a contemporary reference to that of Tassilo III, by the swearing of the vassal’s oath of fidelity or fealty. The magnates of the provinces between Seine and Loire are said to have commended themselves to the future Charles the Bald by the giving of hands and binding themselves with the oath of fidelity: other references show that the inmixtio manuum and the taking of the oath of fidelity constituted the legal bond of vassalage. The words of the vassal’s oath after A.D. 800 ran:
By this oath I promise to be faithful to the lord Charles, the most pious emperor, as a vassal should rightfully be to his lord, for the preservation of his kingdom and his rights.
Vassals to the magnates took similar oaths, and Professor Ganshof explains that, in theory, they owed military service to their immediate lord only so long as the lord was serving the emperor or king; but from the time of Louis the Pious magnates in rebellion against the emperor were in fact able to take the field at the head of their vassals (and see supra, p. 443). The swearing of the oath upon the relics gave the vassal’s oath a religious character and a peculiar solemnity: to the old Germanic obligation to serve the chieftain to the death in battle was added the sacred-ness of the Christian oath, taken in public before witnesses, before the saints in heaven, and under the protection (patrocinium) in particular of the saint on whose enshrined relics the vassal laid his hand.
The size of the benefice with which the vassal was endowed now varied greatly, from a duchy or even sub-kingdom to the small estate with which churches and abbeys were forced to endow their milites from their own lands; but by the mid-ninth century the legal union of vassalage and benefice was assumed. Vassalage was not yet hereditary: the grant of the benefice (later called fief) lapsed at the death of either lord or vassal: but from the year 877, when Charles the Bald made certain rulings at Quierzy-sur-Oise about the fate of benefices if their holders died during the expedition he was planning to make to Italy, the hereditary nature of benefices was recognized as normal (see p. 458). Some benefices consisted of offices, supported normally by landed estates, such as duchies and counties, and in such cases the complex of rights, obligations and territories were known as honores; by 877 such ‘honours’ had been assimilated to benefices and were covered by the rulings of Quierzy-sur-Oise. The son of a count, like the son of a landed vassal, was presumed to succeed to his father’s honour: the all-important Carolingian count became a hereditary official, and the lands of the county became a family inheritance. This assimilation of the honour to the benefice was more complete among the west Franks than the east Franks; but here too assimilation was often accepted. When the old pre-Carolingian tribal areas in Germany regained their local independence after the death of Louis the Child in 911, they were not headed by some descendant of the old Germanic kingly families, but by some strong Carolingian count in the area who had amassed lands and rights within the two generations since 877, and was strong enough to get himself accepted as duke of a ‘national’ duchy.
It is of interest that the vassal’s oath of fidelity was also used by Charlemagne as a model for the oath of fidelity that he demanded that all his male subjects should take to him after he had become emperor. A subject’s loyalty before Charlemagne’s day was implicit, and not attested by any oath-taking, but even before Charlemagne became emperor his conquests had occasioned his demand that certain of his newly incorporated subjects should swear fidelity to him individually.
In a recent essay,1 Professor Ganshof has explained the relation between the general oath of fidelity, finally imposed by Charlemagne on all his subjects, and the oath of the vassal. An oath was so sacred that Charlemagne not only came to demand one from all his subjects, as a safeguard against local rebellion, but even distrusted the practice of oath-taking by the members of any association, such as the gilds: the members of an association might plead justification in following their alderman or leader in civil or military disobedience to the king’s orders. An oath-bound subject sharing in a rebellion incurred, however, the penalties not only of infidelity, but perjury; it was clearly expedient that newly conquered subjects should taie such an oath. Moreover, it enabled Charlemagne to require from cathedral chapters petitioning leave to make a free election of a bishop, the insertion in the diploma issued of a clause requiring the candidate to be in all points fidelis to the king.
Charles was influenced in his policy of requiring a general oath of fidelity to be taken by all his subjects by the conspiracy of count Hardre and other Thuringian nobles in 785, where the rebels swore each other a mutual oath; Charles quelled the rebels, made its leaders take a solemn oath of fidelity to him, and then, curiously, condemned them to death for oath-breach. In 789 he enjoined a general oath-taking on all his subjects by capitulary, the subject promising to be faithful to king Charles and his sons all the days of his life, sine fraude et malo ingenio. The missi were commissioned to exact the oath: but were not numerous enough or strong enough to compel the oath to be taken in all cases. In 792 some counts made another rebellion led by Philip le Bossu, a bastard son of Charles: they denied having taken the oath of fidelity, and Charles accordingly enjoined more effective measures. The Frankish and Lombard kingdoms were divided into circuits, and missi appointed to take the oath themselves from all officials, lay and ecclesiastical, and to see that the local counts made all the remaining males of twelve years old and over to take it. The oath-taking of 793 was generally enforced, and a written record made; the sub-vassals of other lords took it as well as the king’s vassals and the general population.
In 802, when Charles was emperor, a general oath-taking was again required, and made both more detailed and more solemn by direct allusion to the vassal’s oath: a man should be fidelis to Charles ‘as a vassal (homo) should be of right to his lord’. The earlier oath-forms had been negative, an oath not to do certain things disloyal in a subject: now it was explicitly laid down that he should not usurp royal lands or serfs, or aid usurpers, or do wrong to churches, widows or orphans or despoil a benefice granted by the king; not to neglect a summons to the army or an imperial order; not to hinder justice or tolerate its hindrance. Charles was seeking to use the oath now not merely to prevent the giving of aid to rebels, but for the elimination of abuses. Three kinds of oath-taking and three only were legal, as was laid down in 805: this oath of fidelity to the king: the vassal’s oath: and the judicial oath. All other oath-taking was conspiracy, and forbidden.
The political submergence of the Carolingian empire after 888 resulted from there being no single survivor of Carolingian stock who could command sufficient respect from east and west Franks and the papacy to attain the imperial title, and at least nominal rule. The urgent needs of defence made local rulers in command of their own army and of what had been imperial resources a first necessity, even though the tradition of Carolingian rule still had great weight with the secular magnates and very strong support from the church. The ninth-century papacy, scholars and episcopate never ceased to deplore what they regarded as the usurpations of local rulers of non-Carolingian stock: they desired unitary rule as consonant with St Augustine’s teaching about the City of God: they anticipated Dante in believing that the empire exists for peace. They had doubtless also approved the Carolingian willingness to support the church and use her rulers and scholars: there was self-interested regret at the lessening of their own political influence and the confiscation of church lands in the wars; but this was not all. A pale Byzantine splendour had suffused the Carolingian background, and the Byzantine concept of the emperor as another Constantine had largely been accepted by the Frankish episcopate. After 888 the bishops were left with the unacceptable situation that Carolingian and non-Carolingian rulers, often at odds with each other and helpless as defenders of the Christian empire from the pagans, yet exercised the old imperial, Byzantine control of episcopal elections and the right to dispose of church lands; the ninth-century bishops and abbots had all the disadvantages of the Byzantine system without its advantages. By their fostering of vassalage, moreover, the new rulers had passed on to local kings, dukes, barons, counts and margraves their own rights of patronage and control over the local churches and abbeys. Lay control, which had arisen from a passing on of Byzantine and then Frankish imperial powers to local magnates, was to lead to many abuses in the tenth and eleventh centuries, and it was suspect to the Frankish bishops from the ninth. The Carolingian tradition had naturally no stronger supporters than the episcopate.
Charles the Fat was deposed by the diet of Tribur in November 887, and died on 13 January 888. The abbot Regino of Prüm, a famous annalist, recognized the change his death brought, and wrote
After his death the kingdoms that obeyed his sway, as if deprived of a lawful heir, dissolved into their component parts: and now they waited not upon their natural lord, but were disposed to create for themselves a king from their own stock.
This, he continued, led to great wars, not because these regions lacked Frankish princes fit to rule a kingdom through their nobility, fortitude and wisdom, but because their equality in generosity (noble birth), dignity and power fomented discord: ‘for no one so much excelled the rest that these others would deign to submit themselves to his dominion’. The land of the Franks gave birth to many princes fit for the government of the kingdom; but chance, through natural jealousy, armed them to their mutual destruction.
There was, in fact, only one lawful Carolingian claimant to rule in 888: he was the seven-year-old Charles, the posthumous son of Louis II, the Stammerer (see p. 459). Louis the Stammerer’s cousin, the Bavarian Carloman, had only an illegitimate son, Arnulf, now duke of Carinthia; Arnulf’s mother was a noble lady, and he himself was energetic, able, and a good fighter, but his illegitimacy was against him. When he was only twenty-five he had been put in charge of the marches of Pannonia and Carinthia, and the German nobles knew his personal qualities: he had even administered the duchy of Bavaria during the illness that preceded his father, Carloman’s, death. His uncle, Louis III, succeeded to the duchy of Bavaria, and Arnulf found Charles the Fat unfriendly and suspicious; but on his deposition in 887 the German magnates elected Arnulf king of Germany. Only the stain on his birth prevented his election to the empire, for he inherited the loyalty due to the Carolingians: but, as it was, the west Franks, who had just undergone the dreadful siege of Paris in 885-6, and who could not hope for leadership from a seven-year-old Carolingian boy, elected as king Odo, count of Paris. Count Robert the Strong, his father, had defended Paris, and Odo had distinguished himself fighting against the northmen. He and Arnulf, however, had not been the only claimants to the west Frankish throne. Fulk, archbishop of Reims had intrigued to crown Guy II of Spoleto king of the west Franks: Guy came of a Frankish family which had settled in Italy and had pretensions on the Italian crown and even the empire. But the Frankish nobles preferred Odo. Guy II, conspiring with some north French bishops and Baldwin II, count of Flanders (grandson of Charles the Bald), then went to Arnulf at Worms, asking him to claim the kingdom of the west Franks and the empire, denying any part of it to Odo, who was unconnected with the Carolingian stock. Odo, however, had just defeated the northmen and his prestige was high: Arnulf in 888 met him at Worms, recognized him as king of the west Franks and received from him some kind of profession of loyalty: not, however, the oath of vassalage. The west and east Franks now had their separate kings, the line of Arnulf holding Germany till the death of his son, Louis the Child, in 911, and that of Odo alternating with a Carolingian descendant and others till his own line, the line of the counts of Paris, gained the permanent rule of ‘France’ in the person of Hugh Capet in 987.
But the Carolingian empire in 888 was split into more than two kingdoms: the smaller kingdoms of Provence, Burgundy and Lorraine now attained their independence without serious dispute, and the crown of Italy was claimed by two candidates, who fought a civil war for some years.
Provence, long separate in tradition and language from north France, had tried more than once since the emperor Lothar I’s death in 855 to gain her independence. Count Boso had ruled as king from 863 till his death in 887, and left a young son, Louis, and an ambitious wife, Ermingarde, who fought for her son’s rights to succeed eventually. Arnulf had few supporters in this distant province and allowed Ermingarde to summon an assembly at Valence, which early in 890 elected Louis king and had him anointed. Ermingarde ruled as regent till her death in 897. She had hoped Arnulf would support her son’s claim to the whole of the old corridor kingdom of Lotharingia: but this Arnulf was unwilling to do. He installed his illegitimate son Zwentibold as king of Lorraine (in the north of the old Lotharingia), and allowed count Rudolf I to make himself king of Burgundy in the middle portion of the old corridor.
The Jura mountains lie between the Lake of Geneva and the Saône, and count Rudolf had ruled this region for some years as marquess when the magnates of the old province of Transjurania met and elected him king of Burgundy, in 888. He also was ambitious, and desired to become king of the old Lotharingia: he at once occupied Alsace and part of Lorraine, but the magnates held aloof, and he had to be content with Arnulf’s recognition of him as king of Transjurane Burgundy only. His small kingdom embraced the archdiocese of Besançon and the towns of Basle and Geneva.
The northern part of the old corridor, now Lorraine, accepted the royal rule of Arnulf from 888 till 895, losing its independence. When Arnulf’s queen in that year gave birth to a son Louis (the Child), he wished to provide for his bastard son, Zwentibold, and at an assembly at Worms persuaded the magnates of Lorraine to elect him their king. Like his father, Zwentibold was able and energetic, but he was arrogant and no statesman. Zwentibold’s unpopularity was shown when certain Lotharingian counts stole church lands from the bishops of Toul and Trier, and Zwentibold forced them to restore them, rousing such bitter opposition from the magnates that Arnulf had to intervene to bring about a reconciliation. Zwentibold roused fresh opposition when in 898 he exiled count Renier (Renard) of the Long Neck, on which Renier appealed to the then king of France, Charles the Simple. Charles invaded Lorraine, but received little support; he negotiated a peace near Prüm and retired. The final struggle came for Zwentibold on Arnulf’s death in 899: Louis the Child was elected king of Germany, and Zwentibold’s own subjects elected Louis king of Lorraine also. The armies of the two half-brothers, Louis and Zwentibold, fought near the Meuse in August, 900, and Zwentibold was killed. Louis the Child was recognized as king by all Lorraine, to which he granted a local autonomy; he appointed Gebhard, count of Franconia, duke of Lorraine, and the land had no more kings.
The succession to the kingdom of Italy was long disputed after 888 between representatives of the families of two Frankish counts associated with the Carolingian conquest of Italy. In the north, the margraves (marquesses) of Friuli had long defended a Frankish march against the Slavs, an assembly at Pavia, now, with its palace the capital of northern Italy, in 888 elected Berengar, marquess of Friuli and also a grandson of Louis the Pious on his mother’s side, to be king of Italy. He had no supporters however in central or southern Italy.
Guy II, duke of Spoleto, whose old Lombard duchy straddled the Alps and who also ruled Naples, was better placed strategically to command central Italy: the south, in any case, was still held by Byzantines and Saracens. The Frankish dukes of Spoleto came from a family of the Moselle region, but by 888 they had Italianized themselves by long settlement and prudent intermarriage. Guy’s father, the first duke, had gained the duchy in 879, and passed it on to his eldest son Lambert, while Guy held the county of Camerino. In 882 Lambert died, and Guy II succeeded; he was an able intriguer, ready to ally with Saracens or Byzantines when it suited him. Both pope and emperor distrusted him and he was declared a traitor; Charles the Fat commanded Berengar to seize both his person and his lands, but he failed. In 885 Guy made his peace with Charles.
In 888, his first reaction to Charles’s death was to intrigue with archbishop Fulk to gain the French crown, and when this scheme failed he fought again with Berengar near Brescia, but indecisively. Berengar vainly hoped for help from Arnulf, whom he had recognized as king of Germany, but no sufficient help came; at a fierce battle on the Trebbia in 889 he was completely defeated by Guy’s forces. The Italian bishops, before the battle, had just concluded an agreement with Guy to elect him king in return for promises to respect their rights and those of the papacy: but they were ready to accept his possible defeat as the judgment of God between the two contestants: Guy, however, maintained his predominance.
In south Italy the Byzantines were anxious to regain their old territories. They held Bari from 876, and in 888–889 repelled a Saracen attack from Sicily. The Saracen admiral had actually landed on the continent and been forced to return with his troops to Sicily. The Byzantines might have regained all south Italy: they temporarily gained Benevento: but no help came from Constantinople and the Lombard princes were as hostile to them as to the Saracens. They could hold only the coastal region from Siponto to Bari.
Between 888 and 890, therefore, six kings had established themselves in regions covering the old Carolingian empire: but there had been no revival of the imperial title. The strongest possible claimant, in lands and ability and Carolingian parentage, was illegitimate; but pope Stephen V was in difficulties with the Roman nobles, and had noted how Arnulf had maintained the Carolingian tradition in his expedition of May 889 against the pagan Obodrites. In 890, therefore, he invited Arnulf to come to Rome and deliver Italy from ‘bad Christians and menacing pagans’: it was clear that he would be ready to crown Arnulf as emperor. But Arnulf in the years 890–894 was withheld from any Italian expedition by rebellions in Germany and what he deemed to be the menace of the great Slav empire of Moravia on his eastern frontier; the northmen, moreover, were still riding through Lorraine and plundering even Aix-la-Chapelle. In spite of the papal invitation, any Italian expedition was deferred. In 891 Arnulf won a victory over the northmen near Louvain, and the year following he marched into Bavaria to interview the Moravian prince, Svatopluk, who failed, however, to meet him. Though Arnulf led an army into Moravia, he could take no towns or strongholds, and, after plundering the countryside, returned to Germany leaving his relations with the Slavs unsettled by treaty. Had either the Slavs or Arnulf known of the imminent menace from the Hungarians, their true policy would have been one of alliance.
Stephen V meanwhile, in default of Arnulf, had sought another protector in Guy II, whom he had long distrusted both for the fear of encirclement between Spoleto and Naples, and Guy’s willingness to ally with the Saracens. In February 891 Stephen broke with the long tradition of Franco-papal friendship and crowned Guy emperor in St Peter’s. Guy henceforward claimed the imperial title and issued capitularies on the Carolingian model: in 892 he effected the coronation of his son Lambert by the new pope, Formosus. But his power did not even extend over all Italy; Berengar in the north was hostile; Odo of the west Franks and the three small kingdoms of the old Lotharingia, all disregarded him; in Germany, Arnulf bided his time to reckon with him.
In 893, at pope Formosus’ invitation, Arnulf sent an army under king Zwentibold to help Berengar in a campaign against Guy. Zwentibold reached Pavia, where Guy had taken refuge, and then unaccountably retired to Germany. Very early in 894 Arnulf himself led an expedition into Italy and with Berengar marched on Bergamo, held against them by its count and bishop. They took Bergamo, hanged the count, and sent off the bishop to Mainz; most of the Lombard towns, including Milan and Pavia, made their submission. All seemed favourable for an advance on Rome, for the Tuscan princes had submitted to Arnulf and the pope was ready to welcome him: yet he hesitated and then for some unknown reason, returned to Germany. Possibly he returned because the plague had ravaged his army: possibly because the marquess of Tuscany was unreliable, or because Rudolf, king of Burgundy, was hostile to his acquiring the imperial title: both rulers could endanger his communications with Germany.
Arnulf chose indeed to return by the Brenner, and encountered Rudolf’s open hostility in the form of a contingent sent to aid the marquess Anchiar, a relation of Guy of Spoleto: they were defending a town besieged by Arnulf. Arnulf’s power was strong in Germany: but he could not prevail in Italy or prevent Rudolf of Burgundy from reigning there till his death in 912.
Arnulfs second expedition to Italy was no more successful: Guy of Spoleto died in 894, his son Lambert was too young to be dangerous, and the pope Formosus again invited Arnulf to come to Rome. Accordingly, he marched for Rome in December 895, through country foodless and wasted by the Spoletan army, which retreated before him. He reached Rome with an exhausted army and found the gates closed, and Rome held against them by Guy’s widow, Agiltrude. Arnulf and his Germans forced the gates, and he was crowned emperor in St Peter’s by Formosus on 22 February 896: the Romans gave him the oath of fidelity. Even now, however, the situation was dangerous, for neither Agiltrude nor the young king Lambert would submit. Arnulf marched northward to take Spoleto, held against him by Agiltrude, but was struck with paralysis on the way: he was carried back to Germany and lived there helpless till 899: his army had evacuated Italy.
When Arnulf died, in December 899, his son Louis, to whom the German magnates had already sworn fidelity, was only six years old; the descriptive title, Louis the Child, was used of him till his death in 911. He was Arnulf’s only legitimate heir, and an assembly at Forchheim in February 900 proclaimed him king of Germany. The crown of Lorraine was disputed between Zwentibold and Charles the Simple of France, but both accepted the succession of the young Louis III, though Zwentibold did so unwillingly. Louis came to Thionville to receive their oath of fidelity, but Zwentibold rebelled, was defeated, and killed. Lorraine was merged in the German kingdom.
The Forchheim assembly had appointed a council of regency, on which two bishops, Arnulf’s closest friends and counsellors, had the chief place; they were assisted by some lay barons and the marquess of Bavaria. There was no general disposition to upset this regency government inaugurated by respect for the Carolingian tradition, but real Carolingian rule was threatened by two great dangers, external and internal: attack from the Slavs and Magyars on the eastern frontier, and the drift to decentralization of power in order to deal with swiftly moving land and sea raiders. The danger from the northmen was not over: and in this reign it was to be matched by the swift, terrible raids of the Hungarian horsemen from the south-east.
The Magyar horsemen from the steppes, later to be called Hungarians, had settled around 860 between the Don and the Dnieper. Pushed by the arrival of new eastern tribes, they had crossed the Dnieper and then the Dniester, remained a short time in Wallachia, and in 895 crossed the Carpathians and settled on the middle Danube, attacking both the Slovenes and the Moravians. The plain of the middle Danube was inviting to these half-nomad horsemen, and while Arnulf intrigued with them against the Moravians, divided leadership among the Moravians themselves made defence impossible. The Magyars took the plain but were not ready to settle and practise agriculture: they used the rich plain for horse-breeding, to support their plundering raids on Italy and Germany. Away in Germany the paralysed Arnulf could do nothing against them. Moravia, which should have formed a buttress protecting Germany, was now weak and divided; when her strong duke, Svatopluk, died in 894, the jealousy of his two sons, Moimir and Svatopluk II, was fomented by the intrigues of Arnulf, and led to open war between them in 898. Arnulf was helpless and the Magyars overran Moravia; in 899 they raided the Lombard plain, and in 900 Pannonia. In November, 900, bishop Richer of Passau and count Liutpold raised an army and fought the invaders in a bloody battle; the Magyars retreated, but their offensive strength remained.
The German regents in 901 now made formal peace with the Moravians, but too late to stop the Magyar invasion. In 901 the Magyars invaded the German province of Carinthia: in the years 905–906 they bore down all Moravian resistance, and the Moravian state collapsed.
Germany was now directly threatened: in 906 and 907 strong Magyar forces ravaged the valleys of the Danube and the Elbe. The Bohemian Czechs had granted the raiders passage through their lands and in the high summer of 906 they passed by the Elbe valley through the northern mountain frontier of Bohemia and barbarously ravaged Saxony. In 907 they raided an almost undefended Bavaria: only Innsbrück was strong enough to withstand them. The next spring an assembly of German magnates sitting under the young king Louis, tried to raise an army that should protect Bavaria, a province with access both to Germany and Italy: but in July 907 this army was almost destroyed and its leader, duke Liutpold, killed in battle. In 908 the raiders ravaged Saxony again, defeating an army of Franconians and Thuringians, and killing the bishop of Würzburg and the marquess of Thuringia in battle. The next year the Magyar horsemen again raided south Germany, penetrating Suabia, a province hitherto unravaged: they took much plunder. On their way back, the invaders were surprised near the river Inn by Arnulf, the young marquess of Bavaria and the son-in-law of count Liutpold: in the ensuing battle, July 909, he avenged Liutpold’s death. In 910 Louis the Child collected with difficulty an army of Suabians, Franconians and Bavarians: but the invaders won a victory over them near Augsburg; the defeat of so much effort was a catastrophe. When Louis the Child died in 911, Germany was exhausted and the situation desperate. The Magyar Hungarians had established themselves as a permanent danger to the German frontier in the south east, and the possible plunderers of almost any part of Germany. The burning of villages, the pillage and profanation of churches were as common a sight now in the south and east of Germany as they long had been in the north and east from the Viking raids. Defence had to be local: well-timbered burgs in the Alfredian manner, or mound-and-bailey castles, were the only places of refuge.
The need of a strong local power in the general anarchy explains the second outstanding feature of the reign of Louis the Child: the rise of the so-called ‘national duchies’. Louis the Child’s efforts to halt the Magyars, or to deal with confusion amounting to anarchy, drove each region to set about its own defence: the old tribal groups, with a common historical provenance and a common dialectal form of the old Germanic language, regained their pre-Carolingian independence. Their leaders, however, were in no case the descendants of the old pre-Carolingian tribal kings or dukes, but the descendants of the Frankish officers appointed by Charlemagne or his successors to rule the region. Out of the families of the many Frankish counts and margraves, and the comparatively few (military) dukes, the most able, fortunate, or diplomatic made themselves rulers of territorial duchies. The fluid political conditions of the reign of Louis the Child made possible the emergence of the four national duchies, Saxony, Franconia, Suabia and Bavaria, and shortly afterwards, the duchy of Lorraine, which was not an old, pre-Carolingian entity, a ‘tribal duchy’, but the remains of the old corridor kingdom of Lotharingia. The emergence of ruling families of the national duchies had indeed preceded the reign of Louis the Child: in Saxony, the Liudolfing Bruno had headed the wars against the Danes and Wends and gained for his family such a position as early as 880: count Liutpold of Bavaria attained the local rule and defence of his ‘duchy’ by about 900: duke Henry, afterwards king Henry the Fowler, was ruling Saxony by 906. These old tribal duchies were the easier to rule in that they had their own laws and customs, and were old entities submerged for some generations in the Carolingian empire, often with their own sharply defined frontiers. Their new local rulers, charged with the duty of defence, found it the easier to command their local counts and margraves, to usurp revenues and powers properly belonging to the monarchy, and to hand on their own position, estates and privileges from father to son. The great opponents of the centrifugal drift were the bishops, so long attached to the old concept of unitary rule: but the urgent necessities of the times overbore their resistance to the new ducal usurpations.
As to the five great duchies: the ducal house of the Liudolfings in Saxony was descended from a duke Liudolf of the east Saxons, who, in the days of Louis the Pious defended Germany from the Danes and Swedes; his son Bruno was killed fighting the Danes, and his second son, Otto, added to Saxony the march of Thuringia. This Otto succeeded in appropriating many regalian rights, and at the same time in keeping the goodwill of the church; his predecessors had given great endowments to local churches and abbeys. The Liudolfings became unquestioned masters of Saxony without civil war.
The duchy of Bavaria was also a frontier march of Germany, in this case against the Slavs, and it too emerged as an independent duchy without violence. The margraves of Bavaria often bore the title of duke, and were closely attached to the Carolingians. For a short time, under Carloman, son of Louis the German, Bavaria even became a kingdom. When Arnulf, son of Carloman, became king of the Germans, he gave Bavaria to his friend, count Liutpold, and treated it as a march against the Slav Moravians. Liutpold prospered, gained Pannonia, Carinthia and other fiefs, and married Kunigund, sister of the Suabian counts Erchanger and Berthold. When he was killed fighting the Magyars, his son Arnulf proclaimed himself duke ‘by the grace of God’ and ruled Bavaria as an independent state.
The duchy of Franconia, centring in the valley of the Main, the old motherland of the Franks, proved naturally less easy to weld into an entity. It was strongly attached to Frankish tradition and therefore to Carolingian rule: it was not a march, and its defence needs were therefore slightly less pressing. The formation of a duchy was hindered by the contesting claims to power of two great families, the Conradins of western Franconia and the Babenbergers, whose interests lay round Bamberg and eastern Franconia. The representatives of the Babenbergers at the turn of the century were three sons of the count Henry who had been killed at the siege of Paris by the northmen in 886: that of the rival family, Conrad the Old, father of the future king of the Germans, Conrad I. King Arnulf had long favoured the Conradins, and after battles between the two factions in 902, when one of the Bamberg brothers was killed and another taken prisoner and beheaded, royal sentence was pronounced (903) condemning the Bambergs and confiscating their lands. The remaining Bamberg brother, Adalbert, refused to submit, and raised rebellions in 903, 904 and 905, finally defeating Conrad the Old and killing him in battle. The regency council was, however, strong enough to deal with him: he was accused of treason and executed in September 909. The Conradins were left in power in Franconia: when Louis the Child died at the age of eighteen in 911, the German magnates elected Conrad, duke of Franconia, king of the Germans. The situation in Germany was so desperate that the Carolingian tradition went by the board, and Conrad was elected on his merits. He had been chosen in the secular interest, and saw no reason to strengthen his position by receiving anointing and coronation from the church.
Suabia also became a duchy in the reign of Louis the Child, though precariously. Burchard, count of Rhaetia, with his brother Adalbert were the great magnates in the region, once the old territory of the Alemanni; but their position (and especially Burchard’s use of the ducal title) was opposed by Solomon III, bishop of Constance. In the resultant civil strife, Burchard was killed in 911, and the enmity of Solomon was able to procure the exile of his son, which was followed by the assassination of Adalbert. The newly-founded duchy of Suabia was, however, to recover its independence later.
The duchy of Lorraine had been merged in the German kingdom from the beginning of the reign of Louis the Child. The regency council appointed as duke, not a local Frankish magnate, but the Conradin Gebhard, who ruled Lorraine till he was killed in battle with the Magyars in 910. Subjection to a Franconian duke was, however, resented in Lorraine, and two counts of Lorraine, Gebhard and Matfrid, raised rebellions; they were finally beaten by Conrad the Young, and exiled. The magnate Renier then rebelled, witnessing to Lotharingian discontent; and when Louis the Child died in 911, and there was no longer a German Carolingian, Lorraine offered its allegiance to the last Carolingian ruler, Charles the Simple. Lorraine had been the Carolingian homeland, and Charles the Simple was not only received willingly as ruler and defender against Conrad, king of the Germans, but thereafter showed a preference for residing in Lorraine. Only in 925, after conquest by Henry the Fowler, king of the Germans, did Lorraine reappear as a duchy.
The new German duchies were an outward sign of the weakness of the new German kingdom: they menaced also the spiritual power, the Carolingian church in Germany. The German bishops, supported by great demesnes and rich revenues, had been in the Carolingian period great personages in the state: but the new dukes, with their usurped regalian rights, could not tolerate a position of mere equality in the counsels of the king. They feared the episcopate, not only as richly endowed but as well organized under a metropolitan, obeying, that is, a power often outside the boundaries of their own duchies. The bishops could not readily be submerged in the local duchies, for they met in the synods of the ecclesiastical province, or in even wider synods summoned by the metropolitan: they issued canons binding on all Germany. In Suabia it was the bishop of Constance who had opposed the setting up of the duchy, and episcopal influence was hostile elsewhere. The bishops needed supra-ducal protection, and could only find a protector in the person of a strong king: they were the natural supporters of the monarchy, poor substitute as it was for the Carolingian empire. The desired entente between the bishops and the ruler was finally achieved in the reign of Otto the Great.
Meanwhile, among the west Franks, Carolingian rule survived only intermittently between 888 and the beginning of permanent Capetian rule with Hugh Capet in 987. France did not disintegrate into tribal duchies under a single weak monarchy, as in Germany: but the fiefs of the magnates were semi-independent under a king elected now by one territorial interest, now another.
Military valour in fighting the northmen had influenced the election of Odo, count of Paris, as king, in 888: he was strong in Neustria, being not only count of Paris, but of Anjou, Blois and Tours; he was moreover lay abbot of Saint-Martin of Tours and Marmoutier, as well as other Neustrian abbeys. He was crowned at Compiègne by the archbishop of Sens: but he was unacceptable to Fulk, archbishop of Reims and a considerable party of Franks who objected to being ruled by one not of the royal race, and by Neustrians. Opposition to Odo, however, gradually gave way: Baldwin, count of Flanders, submitted, until Odo’s fresh defeats by the northmen encouraged him to revolt in 892. Fulk, meanwhile, had been working in the Carolingian interest: in January 893 while Odo was absent in Aquitaine, he had the thirteen-year-old Carolingian Charles the Simple crowned at Reims. In spite of requests for recognition to the pope and king Arnulf, Fulk could not establish Charles as king; but when Odo was dying he begged the Frankish lords to accept the Carolingian as his successor. He died in January, 898.
Charles was therefore acclaimed king in 898, even Odo’s brother, Robert, count of Paris, greatest of the Neustrian magnates, supporting him. But Charles lacked the ability to restore Carolingian rule in his own line: he was rash and presumptuous, and the most important act of his reign (898–922) was his concession of a Danelaw to the northmen in Normandy. They had already settled in some numbers on the lower Seine and he could not eject them: but their leader Rollo tried and failed to take, successively, Paris and Chartres, and both sides were willing to negotiate. At a meeting in 911 between Charles and Rollo at St Claire-sur-Epte, Rollo agreed to accept Christianity and vassalage to Charles, together with the fiefs of Rouen, Lisieux and Evreux and the country lying between them and the sea. The treaty was realistic and seemed to interpose a coastal barrier between the northern raiders and Neustria. Charles’s other achievement was his acknowledgment by the Lorrainers; but he never obtained the loyalty of all the west Frankish magnates, and had always to contend with an opposition led by the church of Reims. He was captured by his enemies and Raoul (Radulf), duke of Burgundy, was crowned king at Soissons in 923: Charles died in prison in 929. His son, Louis IV, called d’Outremer because he was long in exile in England, reigned from 936’954, his grandson Lothair from 954 to 986, and his great-grandson Louis V from 986 to 987: Carolingian power had long been shadowy, and supported by lands and revenues insufficient to give the king a more than titular authority over the great magnates.
Meanwhile, in the high noon, the twilight and the sunset of Carolingian power, two other foundations were being laid in south-west Europe: that of an Islamic civilization within Europe itself, in Spain (and Sicily) and that of medieval Christian Spain which grew out of the resistance groups of the Asturias and the western slopes of the Pyrenees. In the Iberian peninsula and Sicily the Arabs and the Berbers were bringing an eastern culture and an Arabian religion over to the west: the finest period of the schools of Cordoba and the court of Palermo were to fall in the tenth and eleventh centuries, but Islamic power prevailed in the Iberian peninsula from 711, and in Sicily from the capture of Palermo in 831, and the whole island in 917. By the end of the Carolingian period Islam held both ends of the Mediterranean, and Sicily in the centre. Eventually, the heirs of the Christian Roman empire, in east and west, were to drive her back, but not before the schools of Cordoba had transmitted their version of the old Greek learning, and the science of Baghdad, to the west.
In Spain, the Visigothic king Roderick had been defeated in 711 (see p. 106), and the Saracens passed northward, in pursuit of Visigothic nobles still resisting: by 718 they held nearly the whole peninsula. The rivers of Spain and her mountain ranges run east and west, forming natural boundaries: and the invaders, used to a warmer climate, had no wish to settle north of the region where at least the olive would grow. But apart from small groups of rebels in the Asturias mountains of the north, and large numbers of Basques, still unconquered, at the western end of the Pyrenees, the Muslim tribal chiefs now ruled Spain. They were not, however, themselves united: the more numerous Berbers from north Africa settled in separate groups from the Arabs and Syrians, who, however, received the best of the conquered territories. The whole of Moorish Spain, called by the invaders Andalusia, ranked as a province of the Umayyad caliphate of Baghdad: the caliph, however, had small enough control over it. About a fifth of the conquered land was retained for the state and allotted to the conquered Visigothic peasantry, who paid a heavy food rent. The rest of the land went to the Muslim conquerors and was also cultivated by Christian peasants; they paid a capitation tax as well as the food rent.
Islam in north Africa had worked out a manner of living with and using a conquered Christian population, and the same method was used in Spain. The Christians, the Mozarabs, were allowed their own religion and laws, and conversion to Islam was not encouraged, for that involved the loss of the capitation tax on the new convert. Nevertheless, in the ninth century religious troubles provoked persecution, and some Christians, called by their fellows renegados, embraced the religion of the Prophet: others fled to the Christians in the north. The Jews, on the other hand, profited from the Muslim invasion: they had earlier been persecuted, but were now tolerated. From the days of Muhammad they had always been nearer than the Christians to Islam and had indeed once been her teachers; like the Arabs and Syrians they traded in the Mediterranean, and their trade now increased.
The economic prosperity of Spain was affected both by the continuous civil war with the Christians in the north, and by the improved methods of agriculture, new industries and enlarged export trade brought about by the invaders. There was always, in fact, a wide band of ravaged country between Andalusia and the Christians of the Asturias, and, later, Asturias-Leon; for it was the practice of the Moors, when forced to retreat southwards, to lay waste the countryside they evacuated. The practice bred as much bitterness as the long frontier war itself, and apparently more than the rapid original conquest.
In Moorish Spain, however, a brilliant civilization developed, partly conditioned by economic prosperity. The great estates of the Visigothic nobles were broken up: there was more irrigation, and new crops were introduced, such as rice, the sugar cane, the pomegranate and many more fruit trees. The black plums of Damascus were now grown in Spain, together with new cereals, the olive and, in much greater areas than before, the vine. Both gold and silver were worked in the mines.
Industry again flourished under the double stimulus of sale to a rich home market, for the tribal chiefs, their officers and above all the court of the amir of Cordoba lived with a splendour and comfort then known only in Constantinople and the east: and to an export trade borne in Arab shipping to a relatively backward Europe. Woollen cloths were woven from the fine Merino sheep on the hills of central Spain, silks from the silkworms and mulberries of the south, while metal work, glass ware and paper were all produced. Arab pottery had always been of peculiar beauty, with its borders of Arabic lettering, fine bold ornament and beautiful glaze: the skill of finer artists than elsewhere found an outlet in the decoration of this relatively cheap and coarse material (see p. 206). Toledo became famous for its sword-blades and weapons, Cordoba for its leather goods (cf. the English term cordwainer), and Moors, Syrians and Jews grew rich on an overseas trade handled chiefly through Seville and Malaga. Arab shipping took out cargoes of raw silk, wine, sugar, olives, dried plums and figs and the weapons, textiles and pottery made in Moorish workshops. They took also Christian slaves to the markets of the east: they sailed to Africa, Constantinople, Italy and the east Mediterranean, and brought back oriental silks and purple textiles. With a relatively large merchant class, Andalusia attained a more extended literacy than the rest of western Europe.
Even after the Muslims had reached the limit of their conquest in 718, they suffered long from tribal rivalries, such as that between the Quais and the Kalb. Better leadership and a measure of unity were, however, afforded to Moorish Spain when in 755 ‘Abd ar-Rahman, a member of the Umayyad family who had been driven from Damascus by the Abbasids, arrived in Spain. In 758 he set himself up as an independent amir in Cordoba, a city largely inhabited by Christian renegades who had no tribal loyalties to render his rule unacceptable; his descendants ruled there till 1031. But while Cordoba was, for a time, a semi-independent republic, its hold on the other Moorish provinces was slight: Saragossa, on the river Ebro, not far from the Pyrenees, tended to break away, and the amirate seemed likely to lapse. ‘Abd ar-Rahman I’s son, Hisham I, however, ruling from 788 to 796 with military skill, ruthless cruelty and eastern craftiness, maintained his power: his faqihs, theologians, thronged Cordoba, and drove the Christian renegades to discontent and sedition. The turbulent faqihs and other rebels were dealt with in the reign following, and the rebels of Cordoba and Toledo subdued by massacres and exile: but it was not till the reign of ‘Abd ar-Rahman III (912–962) that Moorish Spain attained a peak of security and prosperity. ‘Abd ar-Rahman III beat back attacks from the Christian provinces of the north, largely by substituting a single mercenary army for the earlier, tribal army, which had been fighting under its own Arab or Berber chiefs. The provincial rulers themselves were subdued, among them Umar ibn Hafsun who had made himself independent for some years in southern Andalusia: by 930 all the independent Arab and Berber chieftains were subdued. Not only internal disunion and civil war had been dangerous: but the establishment of the Fatimite caliphate in Egypt in 909 had led to religious disunion in Spain. The Fatimite caliphate was heretical and Shi’ite, and Shi’ite propaganda spread to Spain among the Berbers.
‘Abd ar-Rahman III, however, coped with all these dangers, and asserted his supremacy in Andalusia and his equality both with the Abbasids of Baghdad and the Fatimites of Africa. In 929 he assumed the title of caliph and commander of the faithful. The caliphate of Cordoba, his capital, was to last from 929 to 1061 and to see the flowering of Muslim civilization in Spain, a civilization which had grown up under the amirate. The amirs and caliphs of Cordoba were hereditary and absolute rulers, using viziers (ministers) and a council of state. The wali’s governed the cities, the cadi judged offences against Muslim law; Arabic was so widely used that nearly all the inhabitants of Spain were bilingual and the Romance tongue of the conquered adopted a very large number of Arabic terms.
Though the rulers of Spain were Umayyad, their attitude to learning was liberal and Abbasid rather than narrowly Meccan: the schools of Cordoba, like those of Abbasid Baghdad, became famous. Cordoba in the tenth century became the most cultured city in Europe: later, it was to reintroduce Aristotelian logic and Aristotelian science to western scholars. Arab astronomy had been transported by the Muslims into Spain and was studied at Cordoba and Toledo; in 1081 the Toledan Tables, mainly based on Ptolemy and al-Khwarizmi (see p. 209) were drawn by al-Zarqali, to be translated into Latin by Gerard of Cremona (d. 1187). Arab medicine was studied in Spain, though the first beginnings of the famous school of Arabic medicine at Salerno sprang from contact with Arab Africa rather than Spain; Constantine the African who had long wandered in the east, fled from his recent home in Tunis to Salerno just before the coming of Robert Guiscard. The Englishman, Adelard of Bath, visited both Cordoba and Salerno and wrote his treatise, De Sphaero, on Arabic cosmology (built upon that of Ptolemy): he was wandering in Arabic Spain in the early years of the twelfth century.
The small Christian states, meanwhile, maintained themselves with difficulty: compared with Moorish Spain they were very poor. They fell into two groups: those based upon the hills that border the southern shore of the Bay of Biscay, and those based upon the Pyrenees. Between the Roman Asturica Augusta and the Bay, the Asturian mountains sheltered the Visigothic nobles and their followers who retreated from the Moors: the Cantabrian mountains, the eastern end of the range, had also their bands of refugees: the whole range was to form the earliest kingdom of Christian resistance, that of the Asturias. The Pyrenean Christian states were to become kingdoms later.
The history of the kingdom of the Asturias is associated with the victory of a fugitive Visigothic noble, Pelayo, at Covadonga (718) over his pursuers: the battle is reckoned the beginning of the long Spanish reconquista. Alfonzo I (739–757) erected the small Christian province into the kingdom of the Asturias, and for a century its existence was precarious and its southern frontier fluid. The confusion and decline of the amirate of Cordoba, however, gave opportunity for Asturian advance: the frontier reached the Douro valley and a line of towns, including Oporto and Burgos, were fortified. The splendid reigns of ‘Abd ar-Rahman III and his successors, the unity and military strength of the new caliphate, checked Christian advance for a time, but the capital was moved from Oviedo to Leon, and the kingdom of the Asturias became the wider kingdom of Leon, covering all north-western Spain. The union of the small Christian states of the Iberian peninsula was long ahead: but the eventual winning of the Atlantic coast of southern Europe, and eventually the New World, for the old Greco-Roman-Christian civilization was opened.
BIBLIOGRAPHICAL NOTE
See A. Fliehe, L’Europe occidetttale de 888 à 1125, 1930: G. Barraclough. ‘The Problem of the Duchies’, in Medieval Germany, 911–1250, 1938. For the rise of a feudal society, see F. L. Ganshof, Qu’est-ce que la féodalité?, 1947 ed., translated by P. Grierson as Feudalism, 1952: M. Bloch, La société féodale: la formation des liens de dépendance, 1939: J. Calmette, Le monde féodale, ed. 1951: F. L. Ganshof, ‘Charlemagne et le serment’, in Mélanges d’hist. du moyen âge dédiés à la mém. de Louis Halphen, p. 259 ff. See also J. Calmette, L’effondrement d’un empire; E. Lévi-Provencal, Histoire de l’Espagne musulmane, tom, i, 1950; Sanchez-Albornoz y Meduina, En torno a los origenes del feudalismo, 3 tomes, 1942, and La Espãna musulmana ségun los autores islamitas y cristianos médiévales, 2 vols., 1946
GENERAL REFERENCE BOOKS
Chambers Encyclopedia, published the most recently of the general encyclopedias, in 1950. is most useful, both as regards articles and bibliographies. The last ed. of the Encyclopedia Britannica was published in 1929: its articles are sometimes longer. The Catholic Encyclopedia, though published in 1907. etc., is still useful as including ecclesiastical personages and subjects not found in the general encyclopedias, as does the Herzog-Hauck Realencyclopaedie, and its English version, the Schaff-Herzog.
There are two great collections of medieval Latin texts, of importance to historical study: the most recent and best edited is the Monumenta Germaniae Historiae, comprising series of ancient authors, early medieval writers, letters, laws, verse writers, annals, diplomas and Merovingian and Carolingian capitularies (royal edicts). The other great collection is Migne’s Patrologia Latina and Graeca: there is also, for Byzantine history, the Corpus scriptorum historiae byzantinae, ed. B. G. Niebuhr, 50 vols., 1828–97. For lists of rulers, bishops and other officials, with dates of their office-holding and much useful information, see J. M. J. L. de Mas Latrie, Trésor de Chronologie, 1889, and in a forthcoming ed.; for bibliographies: L. J. Paetow, A Guide to the Study of Medieval History, ed. 1931: A. Molinier, Les sources de l’histoire de France, part i, 1901: Wattenbach-Levison, Deutschlands geschichtsquellen im Mittelalter, part i. The following, not mentioned in earlier notes, are useful in connexion with special aspects of European history: H. Breslau, Handbuch der Urkundenlehre fur Deutschland und Italien, vol. i (for lists of referendaries and other officers): F. Cabrol and H. Leclercq, Dictionnaire d’archéologie chrétienne, 1903 etc.: R. Naz, Dictionnaire de droit canonique, 1924, and in progress: L. Thomassin, Vetus et nova ecclesiae disciplina (useful for ecclesiastical officers in this early period): C. Eubel, Hierarchia catholica medii aevi, 1913: the Encyclopaedia of Islam, 1908–1938: C. H. Philips, 1951, Handbook of Oriental History: The European Inheritance, ed. E. Barker, G. Clark and P. Vaucher, 1954.