The Development of Southern French and Catalan Society, 718-1050

Archibald R. Lewis


The Carolingian Conquest

[20] In 718 Moslems from Spain crossed the Pyrenees into Southern France. In so doing they set in motion a series of events which was to end sixty years later with most of the Midi in Carolingian hands. By 778 Provence, the Narbonnaise, and Aquitaine had been forcibly added to the Carolingian Empire, while Gascony had accepted its nominal overlordship. The Moslems had been driven back across the Pyrenees, and Carolingian armies had begun the task of liberating a band of Spanish territory where Christian Catalonia, Aragón, and Navarre were to arise.

As we examine these years one fact stands out -- the slowness of the Carolingian advance to the south. Not until 732 do we find Carolingian armed forces in the Midi. Even after this date their advance seems to have been a spasmodic one. Periods of activity seem to have alternated with periods of quiescence. Thus we find them active between 732 and 739, between 752 and 759, from 760 to 769, and again in 778. During the remaining years of these six decades they did little.

It was Moslems from Spain, as a matter of fact, who began outside intervention in Southern France, not the Carolingians. Charles Martel, who was busy consolidating his authority north of the Loire and in Germany, seems to have had little inclination to follow their lead initially. True, Aquitaine and Provence, which were menaced by this Moslem invasion, were nominally part of the Frankish domains he ruled as mayor of the palace; and true also, Eudes, duke of Aquitaine and a partisan of some of the later Merovingians, was hostile to him and his pretensions. But Charles was in no very strong position at the time, and he seems to have been more concerned with the menace of the Frisians than with events in the Midi.(1) Thus he did not react in any effective way to the news of the Moslem invasion of the Narbonnaise.

[21] So it was that when the forces of Islam reached Narbonne in 720 and established themselves there and turned it into a permanent base for future operations, they did so with only local, ineffective opposition.(2) After all, the Narbonnaise had always been Visigothic rather than Frankish, and so the Moslem conquest could easily be considered an extension of their activities in Spain -- wiping out the last traces of Gothic opposition to their rule -- rather than any immediate threat to Frankish territory.

If such thoughts existed in the minds of the Franks, Aquitanians, and Provençals, however, they were soon rudely dispelled. Almost at once the Moors advanced from Narbonne into Aquitaine where, late in 720 or early in 721, they were defeated before the walls of Toulouse by an Aquitanian-Frankish force led by Duke Eudes.(3) Checked in this direction, the Moslems appear to have contented themselves with a consolidation of their position in the Narbonnaise, occupying Carcassonne in 725 and perhaps some of the other civitates(4) and launching razzias into Rouergue and the Albigeois, one of which seems to have reached as far north as Autun.(5) Such raids, it would seem, did not concern Duke Eudes too much, or perhaps his fear of his Carolingian rival, Charles Martel, was greater still, for we learn that some time before 731 he allied himself to the Moslem ruler of Cerdanya, by giving him his daughter in marriage -- an alliance which he may have felt would secure his southern frontiers.(6)

[22] If Eudes hoped to secure himself from Moslem attack by this alliance with a ruler who controlled the eastern passes from Spain into Septimania, he was doomed to disappointment. For in 731 the new governor of Spain, Abd-ar-Rahman, who was not a party to this alliance, entered the lists against him. In doing so, Abd-ar-Rahman avoided the passes of the Pyrenees controlled by Eudes' Moslem ally and instead, with a large army, crossed into France by way of Pampeluna and the western Basque country. Moving north he met and destroyed an army led by Eudes outside Bordeaux and sacked the city. Eudes fled north to ask Charles, his Carolingian rival, for assistance, while the victorious Moslems followed, spreading destruction in their wake and sacking abbeys like Saint-Hilaire of Poitiers.(7)

News of this Moslem advance north seems to have at last alerted Charles Mattel to his danger and so, with an army of heavily armed vassi whom he had enriched with Church lands, the Carolingian mayor of the palace moved south to meet the Moors. Outside Poitiers in 732 the Franks won a victory, and the Moslems retreated south leaving Abd-ar-Rahman, their commander, dead on the field of battle.(8)

There is little evidence that Charles Martel moved to follow up this victory, and it seems clear that its chief beneficiary was Duke Eudes, who probably did homage to Charles and then regained control of his duchy of Aquitaine, dying undisturbed a few years later in 735. A contemporary source gives us an explanation of Charles' inactivity in Aquitaine at the time, for it tells us how in 734 he was busy pacifying Burgundy, which he turned over to trusted followers or fideles after its conquest.(9) Only upon Eudes' death in 735 do we find him active in Aquitaine, this time leading an expedition as f at south as Bordeaux which he captured despite its formidable fortifications. It seems probable, though, that this expedition was less an attempt to conquer Aquitaine than a show of force to make sure that Hunald, son and successor of Eudes, would acknowledge [23] Carolingian overlordship.(10) Whatever its cause, Charles does not appear to have kept possession of Bordeaux and other parts of Aquitaine. Instead he returned north to his domains beyond the Loire.

In 736, however, he was ready to move south. But this time he concerned himself not with Aquitaine but with the Rhone Valley south of newly conquered Burgundy. The reasons for this are obscure. A passage from the Annals of Aniane gives us a possible explanation. This tells us that some time prior to 736 Moslems from the Narbonnaise had occupied Arles, perhaps on the invitation of Maurontius, patrician of Provence, and for four years had been raiding widely in the Rhone Valley.(11) Perhaps such activities had begun to menace Charles in Burgundy, or perhaps they furnished him with a needed excuse. At any rate in 736, after occupying the Lyonnais and the Middle Rhone region and placing his judices or supporters in the principal civitates,(12) Charles and his brother Childebrand advanced down the Rhone with a Frankish and Burgundian force. They besieged and captured Arles and Avignon and then moved west into Septimania. The civitates of the eastern Narbonnaise surrendered to the Frankish forces, and in 739 these forces defeated a Moslem army at Buerre not far from the Moslem base of Narbonne.(13)

Charles did not follow up this victory with a seizure of Narbonne and an expulsion of the Moors from the Midi. Perhaps, as has been suggested, he did not have sufficient siege equipment to take this stronghold which could be supplied from the sea. More probably, however, he was deterred by other happenings. Hunald, princeps of Aquitaine, was hostile and threatened his long line of communications to the north.(14) More important, [24] Maurontius, patrician of Provence, from his unconquered city of Marseille, raised a revolt against him from the rear. Charles was forced to withdraw from Septimania to the Lower Rhone, destroying many of the castella of the region and the fortifications of such civitates as Agde, Béziers, and Mauguio as he retreated. Once in Provence with Lombard assistance he appears to have crushed all opposition. Patrician Maurontius had to flee for refuge to the Alps. Though the Narbonnaise had escaped him, by 739 Provence and the Valley of the Rhone were Carolingian.(15)

Charles Mattel returned to the north after this campaign, and a little later in 742 we learn of a revolt by Hunald, duke of Aquitaine. Charles had to invade Auvergne and capture the castle of Castelluc before Hunald would return to his allegiance -- a peace, however, which left Aquitaine still effectively in the latter's hands.(16) Hunald was succeeded peacefully in 745 by his son Waiffre, and Charles' heirs who succeeded him in the same year made no attempt to disturb the status quo. We can only guess at the reasons for this peaceful decade in the Midi. One reason may have been the fact that Charles Mattel left his domains to his two sons, which meant that until his brother had retired to a monastery, Pepin, Charles' able son, did not have a free hand. A second reason may have been the fact that Pepin was more concerned with events in Germany, his intervention against the Lombards in Italy, and his efforts to replace the Merovingians as kings of the Franks with his own Carolingian family than he was with Aquitaine or Septimania. Not until 752 was Pepin, who now securely held the position as Frankish king, free to concern himself with the Midi.

Once the decision was made in 752 Pepin turned his attention to Septimania, taking up where his father had left off in 739. He led a strong Frankish army south. In the Narbonnaise a local Gothic count, Ansemundus, went over to him and delivered into his hands the cities of [25] Nîmes, Agde, and Béziers, while still another local magnate did the same for Mauguio.(17) All but unopposed, Pepin and his forces advanced to the outskirts of Narbonne, the principal Moslem base.

What followed is somewhat confused in our sources, but they agree that for some seven years Frankish forces invested Narbonne without taking the city. What caused this delay? There seem to have been several reasons for it. In the first place we learn that Count Ansemundus, Pepin's Gothic ally, was killed outside Narbonne, thus limiting local support.(18) This seems to have been followed by a revolt at Nîmes in the Frankish rear, a rising which ended in the placing of this region under a Frankish governor, a certain Count Radulf.(19) Finally we hear of an attack on the Narbonnaise by forces under the control of Duke Waiffre of Aquitaine.(20) All these developments must have made the Frankish position before Narbonne difficult to maintain. So it was not until 759 that Pepin secured the city, and then only after promising the Gothic inhabitants of this region their own laws and probably their own government. Thereupon, we are told, they killed the Moslem garrison and delivered the city over to him.(21)

Pepin's conquest of Narbonne eliminated the last effective stronghold which the Moslems held north of the Pyrenees and was probably followed by an advance south in which he or his Gothic allies secured Roussillon.(22) [26] Soon afterwards we are told of a more important conquest, for one local chronicle tells us that after the fall of Narbonne Pepin's forces moved north and west to occupy Toulouse, Rouergue, and the Albigeois.(23) By 760 both Septimania and Southeastern Aquitaine were Carolingian.

It was this last conquest which led to a Carolingian assault on Aquitaine proper, probably the most difficult operation which had been attempted so far. True, Pepin, the Frankish monarch, had certain advantages in this struggle. Thanks to his Gothic allies, he appears to have had a firm hold on Septimania, while the Moslems driven south of the Pyrenees and involved in the civil wars which attended the arrival of Abd-ar-Rabman I in Spain, were no longer a menace. Waiffre's domains lay exposed to Pepin on three sides, the north, the east, and the south. On the other hand the rugged terrain of the Massif Central which Waiffre controlled and the many fortified cities and castella of the region favored the defense. In addition to this, Waiffre appears to have countered Pepin's alliance with the Goths of the Narbonnaise by making a similar arrangement with Duke Lupo and the Gascons south of the Garonne, which assured him of Gascon troops and assistance when needed.(24) He also appears to have begun to confiscate Church lands in Aquitaine -- probably binding his followers to him by generously distributing such property among them, as Charles Martel had done some decades earlier in Northern France.(25)

In 760 our sources reveal that Pepin, after denouncing Duke Waiffre's spoilation of the Church, sent an army into Berry and Auvergne which ravaged a large part of Aquitaine, or as the chronicler puts it, "maximam partem Aquitaniae," with fire and sword.(26) The next year two of Waiffre's [27] border counts, those who controlled Bourges and Auvergne, replied in kind by ravaging Burgundy.(27) Pepin's answer to this was an expedition into Auvergne which captured the key fortresses of Bourbon and Clermont, killed a number of Gascon auxiliaries who were fighting for the Aquitanians, and forced Count Bladino of Auvergne to submit.(28) With Auvergne in his hands, the next year he attacked Berry and Poitiers. Bourges surrendered with its Count Humbert, and so did the fortress of Thouars and the count of Poitou. Gascons fighting for these counts were deported across the Loire into Northern France.(29)

Then in 763 Pepin advanced further into the heart of Duke Waiffre's domains with a raid which carried his forces through the Limousin as far south as Quercy. For the next three years a seesaw struggle took place. Pepin captured such fortified civitates as Poitiers, Limoges, Saintes, Angoulême, and Périgord and a number of castella and destroyed their fortifications if he did not feel he could garrison them. At the same time he occupied and rebuilt certain other fortresses with his Frankish levies. Duke Waiffre seems to have followed a policy of destroying such Frankish-held castles if he could capture them and raising revolts in Pepin's rear like that led by Count Bladino in Auvergne. Under the circumstances the war became a bitter affair, and Pepin began to wage it with deliberate frightfulness, for we are told he burnt villas, destroyed vineyards, and depopulated monasteries.(30)

By the end of 765 it seems clear that such brutal tactics had destroyed resistance in central Aquitaine and in 766-767 it was the turn of Périgord, Angouléme, and the Bordelais, close to the Gascon allies of Duke Waiffre. In a series of campaigns the castella and civitates of these regions were reduced and the countryside so devastated that "nullus colonus terram ad laborandam" escaped unscathed. Waiffre's Gascon allies were badly defeated, and in 768 the pro-Frankish peace party of nobles led by Count Humbert of Bourges and other magnates submitted. Most of Waiffre's family were captured, and he was killed in a forest in Périgord [28] as his son Hunald escaped south to Gascony.(31) Pepin, who had led this war to a successful conclusion, died soon afterwards.

One last flurry of resistance followed. Pepin's death appears to have raised hopes again among the Aquitanians and their Gascon allies, and Hunald, son of Duke Waiffre, raised the standard of revolt. In vain. Pepin's heirs, Charles and Carloman, immediately moved south to the Garonne where they put down the revolt. Then we are told they built a fortress at Fronsac on the Dordogne to control this sensitive border region and crossed the Garonne into Basque territory. Unable to resist, Lupo, duke of Gascony, surrendered Hunald to them and submitted himself.(32) The ten-year campaign in Aquitaine was over at last and even the Gascons had been forced to accept a nominal Carolingian suzerainty. What had begun with the Moslem invasion of the Narbonnaise in 718 had ended half a century later with the Carolingians supreme in the Midi.

For the next nine years our sources contain no information on events in Southern France. Probably they were peaceful years, as the Carolingians concerned themselves with other problems and other parts of their Empire. During this period Carloman, Charlemagne's brother, retired to a monastery leaving the latter supreme in the Frankish Empire. Charles himself was fully occupied with problems arising in Saxony and Bavaria, and with his first campaigns in Italy. Under the circumstances the Midi could claim little of his attention.

All this changed, however, in 778. In this year dissident Moslem nobles of Saragossa and Northeastern Spain, in revolt against the emir of Cordova, Abd-ar-Rahman I, invited Charlemagne to intervene in Northern Spain, promising him suzerainty of the region north of the Ebro. Charlemagne seems to have been attracted by this prospect and gathered together a large army which included Gascon and Aquitanian contingents and invaded Spain. He appears to have used western Pyrenean passes and to have entered Spain by way of Pampeluna. This campaign was a failure, despite efforts of Carolingian chroniclers to conceal the fact, and Charlemagne had to withdraw across the mountains without having accomplished anything. Emboldened by his failure, Basques in the Pyrenees rose against him and massacred his rear guard. Charles had to return and punish them for their treachery. But he also had to do more. He seems [29] to have been forced to a realization that the ties which bound Gascon and Aquitanian to him were very fragile ones indeed.(33) This realization probably explains why he almost immediately reorganized Aquitaine and inaugurated new policies in the Midi -- policies which will be examined in later chapters.

With the year 778, then, we reach the end of one era and the beginning of a new one. The conquest of the Midi by the Carolingians had been completed, but it was already apparent that something needed to be done to provide for a better organization of areas south of the Loire which had been added to the Carolingian Empire. Equally important, the events of 778 had shown the nominal and superficial nature of Gascony's submission, and this presented a problem for the future. As for Spain, despite the failure of Charlemagne's intervention, it continued to beckon as an area in which something might be accomplished in the future.

Such then is the story of the Carolingian conquest of the Midi and their first campaign across the Pyrenees on Spanish soil. But it raises some important questions. Of these the one that most needs answering is why the conquest of the Midi proceeded so slowly and was attended by so many difficulties. Why did it take sixty years, and why at the end of the period was Carolingian control so fragile that it demanded a drastic reorganization soon after 778?(34)

One answer certainly lies in the field of geography. To conquer Provence, Gascony, and Aquitaine, the Carolingians had to operate very frequently in a mountainous terrain near the Alps, Cévennes, and Pyrenees quite different from the flat plains of Northern France and Germany to which they were accustomed. This terrain gave every advantage to its native defenders. While such terrain does not appear to have hindered the conquest of Provence, it undoubtedly prolonged the resistance of Aquitaine and kept Gascony from ever being completely Carolingian.

But more than geography lies behind Carolingian difficulties. What of the role of Moslem opposition to them? Can we blame it for the slowness of their advance? This seems unlikely. Moslem forces never proved very worthy opponents to the Carolingians, except in Spain. Down to 778 they lost every battle they fought with Frankish contingents. This was true at Poitiers in 732 where they were badly mauled, in 736-737 when [30] they failed to hold cities like Arles and Avignon, and in 739 when they were crushed at the Battle of Buerre. Outside Narbonne between 752 and 759 they did not dare meet Carolingian troops in the open field. It seems clear then that though Moslem forces represented the principal excuse for Carolingian intervention in the Midi, they were not a real obstacle to a successful Carolingian advance to the Pyrenees.

This leads one to the conclusion that neither geography nor the Moslems furnished the Carolingians with real difficulties but rather something else -- the resistance of the local inhabitants and society of the Midi, whether it be in Provence and the Valley of the Rhone, in Septimania, in Aquitaine, or in Gascony. And we can discern a general pattern which this resistance seems to have taken. Upon the arrival of Frankish forces in each region, the inhabitants would almost always submit initially, but such submission was always temporary and nominal. It was almost always followed by a revolt when a Charles Martel, a Pepin, or a Charlemagne was in difficulties or when they tried to transform a nominal suzerainty into something more real and enduring.

Thus in Aquitaine the submission of Eudes to Charles Mattel in 732 did not mean a Carolingian Aquitaine, but a tie, fragile in nature, made necessary by Eudes' defeat at the hands of the Moslems.(35) Equally formal were the submissions of Hunald in 735 and 742, each time the result of a Carolingian invasion of his principality.(36) Waiffre's attempted intervention in the Narbonnaise during the years 752-759 shows the inherent opposition of his family to the Carolingians(37) -- an opposition which did not end until the brutal campaigns of Pepin and his sons during the years 760 to 769 finally made Aquitaine Carolingian.

In much the same way the inhabitants of Provence and the Lower Rhone Valley submitted in 736-737 only to stage a revolt in Charles Martel's rear when the opportunity arose -- a revolt which only ended when Maurontius, patrician of Provence, was driven out and the rebellion crushed by Lombard and Frankish troops.(38) During the same period Charles Martel's campaign in Septimania again proved how little he could rely upon native Gothic counts and the local population for support against the Moslems. And this was repeated during the years 752-759, when, after the death of Count Ansemundus, the inhabitants of Nimes revolted and native rule had to be done away with and the region put [31] under a Frankish count.(39) Narbonne, as a matter of fact, only became Carolingian in 759 after Pepin had specifically guaranteed its inhabitants their own laws and perhaps their own native rulers.(40) Charlemagne's experience in Gascony in 778 seems to have shown him the same unreliability as far as the native population and its rulers were concerned.(41)

In the face of this apparently deep-seated unyielding attitude of local populations and their leaders in every part of Southern France, the Carolingians, during this period, seem to have followed two policies. On the one hand, wherever possible, they made alliance with local magnates in the regions they conquered, like Count Ansemundus in Septimania, the Goths of Narbonne, or a Count Humbert of Berry. On the other hand they found it wiser not to put too much trust in such allies to control these regions. So they early began to introduce into Southern France their own Northern Frankish vassi to serve as rulers and garrisons of the areas which they had conquered. Thus in 734 we hear how Charles Martel, having conquered Burgundy, turned over authority to his fideles.(42)A little later our sources tell us that in the Lower Rhone Valley he placed the civitates under the control of his own officials or judices.(43)In the Narbonnaise Pepin placed Frankish counts in control of Nîmes, Uzès, and Rouergue after revolts developed in these areas.(44)

In Aquitaine during the fierce campaigns of 760-779 we again and again hear of Franks whom Pepin stationed in the castles and fortresses of the region.(45) Later on, the Astronomus, writing of Aquitaine in 778, correctly explains this basic Carolingian policy by saying that Charlemagne made sure that all abbots and counts of this region were Franks, and that he also introduced into the land large numbers of Frankish [32] warriors who were called vassi.(46) In fact only in the region about Narbonne and in Gascony was such a policy not followed prior to 778.

To this policy, which has been called Frankish colonization by churchmen, officials, and warrior garrisons, one should add one more feature of Carolingian conquest and control -- the ruthless crushing of opposition. Historians have often commented on Charles Mattel's confiscation of land in the Rhone Valley and noted how he replaced a native clergy, after 739, with one recruited from the north.(47) Contemporary chronicles have left us a picture of his destruction of civitates in Septimania, like Mauguio for instance.(48) In Aquitaine Pepin's campaigns present a picture of deliberate frightfulness in which villas were burnt, crops and vineyards laid waste, and monastic establishments depopulated,(49) and the end of such campaigns seems to have been followed by confiscations of land which passed from the hands of native landowners into that of Carolingian vassi or swelled the royal fisc.

It thus seems fair to say that the Carolingian conquest of Southern France, at least down to 778, was more than a simple taking over of certain regions by a distant monarch. It resulted in conditions which represented, in many areas of Southern France, a deliberate, sustained, and fundamental assault upon the pre-existing society and institutions. Local magnates and the local Church down to 778 suffered serious damage as a result of Carolingian policies, which deliberately introduced alien officials, churchmen, and others into the Narbonnaise, Provence, and Aquitaine, along with a Frankish proto-feudal system quite different from that familiar to these areas. Resistance was general, but ineffective, except in some parts of the Narbonnaise and Gascony. Unlike the Midi's earlier conquerors, the Visigoths and the Merovingians, we must regard the Carolingians as bringing with them into Southern France certain fundamental [33] mental changes. These triumphed over local opposition, at least initially, and brought a new system and pattern of development to the lands which lay south of Poitou and Burgundy.

One final word concerning these decades of war and change -- the economic results which followed. We have already noted that neither Carolingian conquest nor Moslem depredations suffice to explain the extent of vacant and unused land found in the Midi after 778. This state of affairs was the result of longer-range and more fundamental causes. This is true. But nevertheless one cannot escape the fact that these sixty years did see much destruction, particularly of churches and abbeys which were unable to resist, and of villa life also. It is significant that so much reconstruction was necessary in the Midi after 780. Nor can we escape feeling that the disappearance of every mint south of Poitou and Burgundy was in part the result of these disorderly patterns of life during these years,(50) as was the almost total disappearance of the Midi's external commerce. Thus it was that the Midi not only faced the future in 778 under the alien and different rule of the Carolingians, it also faced it with its churches in ruins, its villa life disorganized, and its economic life at a level lower than had been the case in 718 -- all this was to provide a challenge which its new overlords had to face in planning its future.

Notes for Chapter 2

1. On the hostility which existed between Duke Eudes of Aquitaine and Charles Martel, as well as the latter's difficulties with the Frisians prior to 732 see Fredegarii cont., ch. 10, p. 174; Chron. of Moissac, pp. 291-292; Ademar of Chabannes, I, 51, pp. 51-52; and Annals of Aniane, cols. 2-3.

2. On the Moslem conquest of Spain and advance toward Septimania the best account is still F. Codera y Zaidín, "Límites probables de la dominación árabe en la cordillera pirenaica," in Collección de Estudios Árabes, VIII. See the more recent account in R. de Abadal i de Vinyals, "El paso de Septimania del dominio godo al franco a través de la invasión sarracena," in Cuadernos de Historia de España, XIII.

3. All three Southern French sources -- the Chronicles of Uzès and Moissac and the Annals of Aniane -- mention this victory which Eudes won at Toulouse (Annals of Aniane, col. 2-3; Chron. of Uzès, col. 25; and Chron. of Moissac, p. 290).

4. Only the Chronicles of Moissac, which may reflect special local knowledge of nearby areas, mentions the Moslem conquest of Carcassonne in 725 or 726. Both Zotenburg and Molinier accept the fact that this conquest did take place (H. Zotenburg, "Sur les invasions Arabes dans le Languedoc," in Hist. Gén. de Lang., II, 555-556; and A. Molinier, "Sur les invasions Arabes dans le Languedoc" in ibid., pp. 549-550).

5. Later sources mention Arab raids into Rouergue near Conques and into Velay, but the contemporary Annals of Aniane tells us of this raid as far north as Autun (Annals of Aniane, col. 3-4, and Molinier, "Sur les invasions Arabes," p. 550).

6. This alliance of Eudes with the Moslems of Cerdanya is mentioned in Fredegarii cont., ch. 13, p. 175. See also Zotenburg, "Sur les invasions Arabes," pp. 555-556. Perhaps this alliance with a ruler of a region inhabited by Basques was simply a continuation of Eudes' old pro-Basque policy (see Chapter I, Section I, note 37).

7. Annals of Aniane, col. 4-5; Chron. of Moissac, p. 291; and Fredegarii cont., ch. 13, p. 175.

8. The advance north to Poitiers was a raid or razzia, not a Moslem attempt to conquer this part of France. Thus, some historians have tended to overestimate the importance of Charles Martel's victory. On this see Zotenburg, "Sur les invasions Arabes," p. 557.

9. Fredegarii cont., ch. 14, p. 175.

10. Ibid., ch. 15, pp. 175-176.

11. Chron. of Moissac, p. 291, and Annals of Aniane, col. 4-5. Duprat believes that the Moslems were not invited in by Maurontius until after Charles Martel had intervened in Provence about 736 (E. Duprat in Bouches du Rhône: Encyclopédie départementale, II, 131-132). This agrees with Fredegarii cont., ch. 20, p. 177.

12. Fredegarii cont., ch. 18, pp. 176-177, says that in 737 Charles Martel took Lyon with a Burgundian army and subjugated Marseille and Arles and there "constituit suis judicibus." No local chronicle, however, mentions the conquest of these last two cities during this campaign. It therefore seems more reasonable to believe that they were conquered later in 737 during Childebrand's expedition against Avignon and the rest of Provence proper.

13. Chron. of Moissac, p. 291; Annals of Aniane, col. 5-6; and Fredegarii cont., ch. 20, pp. 177-178. See also Molinier, "Sur les invasions Arabes," pp. 551-552.

14. The Annals of Aniane, in mentioning Waiffre's later attacks on Frankish troops blockading Narbonne during the period 752-759, says that Waiffre attacked Pepin "as his father had done Charles Martel." This seems to imply at the very least a threat to Charles' forces in the Narbonnaise from the duke of Aquitaine in the period 736-739. (Annals of Aniane, col. 6-7).

15. Note 13 contains the references to Charles' withdrawal during which he destroyed the walls of Mauguio and the Arena of Nîmes. The Annals of Aniane also says he took hostages with him. The Fredegarii cont., ch. 20, contains the fullest account and mentions the destruction of Nimes, Agde, and Béziers, and a number of other castra. Then (with Lombard help) he expelled Maurontius from Provence in 738 (Fredegarii cont., ch. 21, p. 178). On this see Duprat in Bouches du Rhône, II, 132. Charles' alliance with the Lombards against the Moslems and the patrician of Provence explains his reluctance during this period to assist the Papacy in Italy.

16. Fredegarii cont., ch. 25, p. 180.

17. Chron. of Moissac, p. 293; Chron. of Uzès, col. 26; Annals of Aniane, col. 6; and Vita Benedicti Anianensis, ed. E. Waitz, ch. I, in Mon. Ger. Hist. Scriptores, XV, 201.

18. Only the Chron. of Uzès, col. 26, mentions the death of Ansemundus in 753.

19. Ibid. Again it is only this local chronicle which mentions the revolt of Nîmes and Uzès against the Franks and its suppression which resulted in these cities' being placed "sub Franchorum dominio" in 754.

20. While the Fredegarii cont., ch. 35, mentions the hostility of Duke Waiffre to Pepin and the fact that he gave asylum to Pepin's enemies in 753 after Pepin had become king of the Franks, it is only local chronicles that mention his attacks on the Frankish forces besieging Narbonne (Chron. of Moissac, p. 294; and Annals of Aniane, col. 6-7).

21. The surrender of Narbonne by its Goths in 759, which is mentioned by both the Annals of Aniane and the Chronicle of Moissac, was an important event. By allowing these Goths their own laws and probably also their own government, Pepin made them his most dependable allies in the Midi. They helped him conquer a hostile Aquitaine. Indeed from this time on an alliance with the Goths of this region remained the cornerstone of Carolingian policy here. It insured Gothic loyalty to the Carolingians, but was probably also the reason for the continued hostility of Gascons and Aquitanians who were old enemies of these Goths.

22. The advance into Roussillon, which was certainly Carolingian by Charlemagne's time, probably took place immediately after the fall of Narbonne. We have no proof of this fact, however. See Molinier, "Sur les invasions Arabes," p. 554.

23. The Chronicle of Uzès says that Rouergue was conquered and a Frankish count installed there in 754 (Chron. of Uzès, col. 27). The Annals of Aniane and the Chronicle of Moissac mention the conquest of Toulouse, Albi, and Rouergue as taking place after the fall of Narbonne in 759. The latter two accounts seem more plausible, since Gothic assistance was probably necessary to secure these regions for Pepin (Annals of Aniane, col. 7; and Chron. of Moissac, p. 294).

24. On the Gascons who fought with Duke Waiffre see Fredegaril cont., ch. 125-126, 130-131, pp. 186-190.

25. Ibid., ch. 124, p. 185.

26. It is interesting to note that Pepin was the aggressor and that Duke Waiffre asked for peace in 760 (ibid.). Only when it became apparent that only complete conquest would satisfy Pepin did the war continue to the death.

27. Ibid., ch. 125, p. 187.

28. Ibid. See also the more general account in Annales Laurissenses, anno 761, in Mon. Ger. Hist. Scriptores, I, 142.

29. Fredegarii cont., ch. 126, pp. 187-188; Annales Laurissenses, anno 762, p. 142.

30. Our sources mention that Franks were placed in fortresses in Berry and that a number of castles were occupied. (Fredegarii cont., ch. 129-130, pp. 189-190; and Annales Laurissenses, anno 763, 765, 766, pp. 144-145). Again in 765 Duke Waiffre attempted in vain to make peace.

31. Fredegarii cont., ch. 131-135, pp. 190-192; Annales Laurissenses, anno 767, 768, p. 146.

32. Anuales Lasirissenses, anno 769, p. 148; Einhard, Vie de Charlemagne, ed. L.Haiphen, ch. 5, p. 445; and Astronomus, Vita Hludovici ímperatoris, ed. G. Pertz, ch. 2, in ibid., II, 608.

33. On this expedition see R. de Abadal i de Vinyals, "La expedición de Carlomagno a Zaragoza, el hecho histórico," in Coloquios de Roncevalles. Augusto 1955.

34. See F. Ganshof, "Une Crise dans le règne de Charlemagne, les années 778 et 779," in Mélanges Charles Gaillard.

35. Fredegarii cont., ch. 13, p. 175; Chron. of Moissac, p.291.

36. Fredegarii cont., ch. 15, 25, pp. 176, 180.

37. Ibid., ch. 35, p. 181; Annals of Aniane, col. 6; and Chron. of Moissac, p. 293.

38. Fredegarii cont., ch. 20, p. 178.

39. Chron. of Uzès, col. 27.

40. Chron. of Moissac, p. 294; Chron. of Uzès, col. 27; and Annals of Aniane, col. 7.

41. See Ganshof, "Une Crise dans le règne de Charlemagne."

42. Fredegarii cont., ch. 14, p. 175.

43. Ibid., ch. 18, pp. 176-177.

44. Chron. of Uzès, col. 26-27.

45. On the occupation of Bourges by Pepin's counts see Fredegarii cont., ch. 126, p. 187. On the occupation of civitates and castella in the rest of Aquitaine between 762 and 765 see ibid., ch. 129, p. 189. On the occupation of Berry in 766 see Annales Laurissenses, anno 766, p. 146. On the building of the castle of Fronsac on the Dordogne see ibid., anno 769, p. 148. See also references to castles fortified by Charlemagne in this region in Annales Regni Francorum, ed. G. Kurze, and in M. Garaud, "La construction des châteaux et les déstinées de la vicaria et du vicarius Carolingiens en Poitou," in Revue historique de droit francais et étranger, XXXI (1953), 54-60.

46. Astronomus, Vita Hiadovici, ch. 3, p. 608. On benefices given such vassals see Capitulare Aquitanicum, anno 768, ch. 5, 9, in Mon. Ger. Hist. Capitularia, I, no. 18, p. 43.

47. A particularly ruthless ecclesiastical policy seems to have been followed in the area of the Upper and Middle Rhone Valley. Yet in Provence in 780 we find a certain Maurontius still bishop of Marseille, obviously of the old family of the patricians of Provence (Cart. de Saint-Victor; no. 31, pp. 43-46).

48. "Castra illis regionis vastabat" is the term used in Fredegarii cont., ch. 21, p. 178.

49. In Aquitaine, where resistance was prolonged, we find in our sources phrases like "maximam partem Aquitaniae ignibus concremant," or "monasteria malta depopulata," or "nullus colonus terram ad laborandam." See ibid., ch. 124, 130, 133, pp. 186, 189, 191.

50. M. Prou, Catalogue des Monnaies Carolingiennes, p. cxi-cxx; and A. Lewis, "Le commerce maritime et la navigation sur les côtes de la Gaule Atlantique du au VIIIe siècle," in Le Moyen Age, LIX (1953), 297.