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

Archibald R. Lewis


The Continued Failure of Principalties

[337] During the period from 975 to 1050 when militarism was growing in the Midi and the Spanish March and a powerful Church was waging a valiant and not unsuccessful attempt to curb its excesses, little progress was made in government. In many ways, as a matter of fact, the failure of the family system which controlled Southern France and the Spanish March was even more noticeable after 975 than it had been before. For to the failure of families like the dukes of Gascony and the kings of Provence-Burgundy to evolve an effective governing system, we must add similar failures of those noble houses who controlled regions like Carcassonne-Razès, Cerdanya-Besalu, and Provence. By the year 1050 actual control of most of the Midi and Catalonia was in the hands of a large number of less important families, whose authority was local in character and who seem to have borne the general title of principes. They tended to control the castles and churches in their own local areas without much interference from outside authority. Vestiges of greater influence over a wider region still remained for many families like that of Toulouse-Rouergue, but the time had not yet come when such influence could be transformed into effective power. By the end of this period, of all the noble houses of these regions, only the counts of Barcelona and the duke of Aquitaine had begun the process of building toward a more centralized and effective political system, and they were only able to do this in the last decades before 1050. And even for them the real period of political centralization was to come later in the century.

In this Midi and Catalonia of independent nobles and principes it was Cluny and the Papacy, as we have emphasized, who wielded whatever effective outside political influence existed -- not the French monarchs from the North. For instance such rulers gave only a handful of charters to the abbeys of Catalonia in contrast to their earlier activities along this [338] line. All of them seem to have been issued by Lothaire, the last of the Carolingians - - two dating from 981 for Roussillon,(1) one for the abbey of Ripoll in 982,(2) and the last two granted to the abbeys of San-Cugat and Saint-Peter of Rhodes in 986.(3) When the French crown passed to the Capetian house in 987, however, and when it became apparent that no effective assistance against the Moslems could be expected from the rulers of this house, the nobles of the Spanish March dropped any pretense of formal ties to the French crown. Instead they looked for leadership to Rome and the Papacy, or made arrangements and marriage alliances with other local noble houses of the Midi to the north of them. Only a purely theoretical connection remained with the French monarchy. It was one provided by the use of Capetian reigns to date their charters.

If this seems true in Catalonia, it was even truer for the rest of the Midi. When Hugh Capet became king, few lords of Southern France, except the dukes of Aquitaine, supported him, and indeed we have a record of only one abbey, that of Paunat, which solicited his protection.(4) When he and his son Robert, in alliance with the Poitevin dukes, attempted to intervene in the Limousin, the result was a failure. They could not capture castles which belonged to the count of La Marche.(5) Worse still, according to a contemporary chronicle, they had to endure an open insult from this upstart marcher lord, who, when he was taunted as to how he became a count, a title to which he certainly had little right, contemptuously answered, "Who made you kings?"(6)

A few years later Robert, Hugh Capet's successor, was able to do a little better in Auvergne. There he was at least recognized as king. But even here it is difficult to be sure that he was so recognized in his own right.(7) It seems more probable that his influence over this region was more the result of the fact that the countess of Clermont was a sister of his wife Queen Constance. Robert and his successor Henry were able to achieve considerable control over Burgundy,(8) but to the south their role remained [339] a negligible one. Only as their powerful vassals, the dukes of Aquitaine, expanded their authority south were the Capetians able to exert any influence over the Midi, and then only indirectly. By 1050 the Capetian monarchy had become a Northern French institution, for all practical purposes without much meaning for those who lived south of Poitou and Burgundy. And so it was to remain for another century.

Remembering this lack of influence of the Capetian house, let us examine the political system of the Midi and Catalonia. Perhaps the most striking thing of all is the lack of political cohesion which an examination of these regions discloses. The Auvergne region, for instance, during this period seems to have been divided among the counts of Clermont,(9) the counts of Gevaudun,(10) the dukes of Aquitaine,(11) and a series of independent viscontal and seigneurial families like the viscounts of Carlat, the viscounts of LePuy, the lords of Bourbon, the comtors of Nonette, and a number of others.(12) Under the uninspired rule of King Conrad the Peaceful, King Rudolf, and their German imperial successors, the Middle Rhone Valley region, politically speaking, presents an equally chaotic picture. Here a series of new families like the counts of Savoy,(13) the counts of Gravaisdun,(14) and the counts of Lyonnais-Forez(15) were competing for [340] authority with older houses like that of Valence(16) or the viscounts of Lyon and Vienne.(17) In the Limousin viscounts continued to multiply and to compete with each other also without any effective interference from outside overlords.(18) In many ways a similar situation existed in Rouergue, Quercy, and the Albigeois. To the southwest in Gascony the ruling families who controlled Commignes, Bigorre, Astarac, Fézensac, Béarn, and Tartas, to name only a few of the more important ones, quarrelled with one another and made any attempt on the part of its dukes to create a strong duchy an impossibility.(19) The same is true of Pallars and Ribagorça which not only appear to have lost all political cohesion during this period due to internal subdivision of authority, but to have become subject to outside influences that were exerted by the rulers of Aragón and the counts of Urgell and Barcelona.(20)

Amidst this prevailing pattern of political chaos, however, in several parts of the Midi and Catalonia we do find, during this period, attempts to create something approaching territorial states. None of them ultimately succeeded, but by examining them in some detail we can, perhaps, [341] learn something of the nature of these regions' governmental problems and the reasons why they were able to make so little political progress. We will concentrate on seven of these attempts to create principalities. They are Carcassonne-Razès, Anduze, Provence, La Marche, Angoulême, Cerdanya-Besalu, and last of all Toulouse-Rouergue. Since almost every section of the Midi and Spanish March is represented by these examples they should give us an insight into Southern France and Catalonia, as a whole, from the standpoint of government in this period.

The first such potential principality, that of Carcassonne-Razès, was made possible by the earlier failures and weaknesses of the House of Toulouse-Rouergue, which we have examined at some length in a previous chapter. When Count Raymond Pons died about 950 and a little later his wife Countess Garsinde dissipated his holdings in gifts to her friends, her relatives, and a series of abbeys, neither her cousins the counts of Rouergue, nor the Toulousain heir, Count William Taillefer, were able to build up a powerful state in Languedoc or even to maintain the position which Raymond Pons had held.(21) This gave to the House of Carcassonne-Razès, in the person of Count Roger the Old and his wife Countess Adelaise, its opportunity. Controlling important abbeys like Montolieu, Saint-Hilaire, and Caunes,(22) they were able to expand their authority to the west until they not only controlled Razes, but also that area which was to become the county of Foix.(23) They even had an interest in the newly created county of Commignes.(24) Through the Narbonne relatives of Adelaise, countess of Carcassonne, they gained considerable influence over Narbonne and its archbishopric.(25) By marrying their daughter Ermissende to the heir to the county of Barcelona they [342] were able to exert some influence over Catalonia, which explains why one of their sons, Peter, became the bishop of Gerona.(26) No wonder in a charter of 984, Roger could refer to himself as count and marquis.(27)

It seems probable that this new eminence of the House of Carcassonne in the Midi and Catalonia aroused feelings of jealousy and alarm among neighboring ruling families: particularly the House of Cerdanya-Besalu, which coveted territory in Capçir that Count Roger regarded as his own, and which was disturbed by the close ties he had developed with the House of Barcelona. As a result we learn of a conflict between Count Roger and Count Oliba Cabreta of Besalu. Our information concerning these hostilities is scant indeed, being largely confined to information contained in a charter in which Roger, in a gift to the abbey of Saint-Hilaire, mentions his victory.(28) This would imply that he held his own. On the other hand it is equally clear that Count Oliba Cabreta also made gains at Carcassonne's expense in the Pyrenean area to the south.(29) So it seems probable that when Oliba died in Rome about 990, Roger could breathe a sigh of relief. His will, dating from 1002, shows that he still possessed authority over a considerable portion of the Midi.(30)

Yet this potentially powerful principality soon disintegrated, despite Roger's attempts to hold it together as shown in his will. When he left his authority and domains to his heirs he seems to have stipulated that these heirs should only dispose of their holdings to each other and thus keep them in the family.(31) Like most wills this soon became a dead letter. Each of his heirs -- his brother, who was in control of Foix; his son Peter, bishop of Gerona; and his other two sons, Raymond and Bernard -- soon went their separate ways. By 1016 the rival House of Cerdanya-Besalu was able to purchase the archbishopric of Narbonne for Guifred, a member of their family.(32) Carcassonne's domains were divided among a series of heirs until they lost whatever unity they had had earlier.(33) Soon after [343] 1050 even the nuclear county of Carcassonne was to be the subject of contention between the counts of Barcelona, the counts of Toulouse, and the Trencavel house of Béziers. The work of Count Roger the Old had ended in failure.

When we turn to eastern Languedoc we find another interesting case, that which concerns the House of Anduze. The beginnings of this family seem more modest than those of Carcassonne, since this house was descended from a castellan who held Anduze in the last years of the ninth century.(34) After 975 its power and prestige began to increase rapidly. Perhaps the most important cause of this rise was the marriage of Bernard of Anduze to Garsinde, heiress to a major portion of the viscounties of Béziers and Agde.(35) As a result, early in the eleventh century, a potentially powerful principality had been created which stretched from Nîmes to Béziers and north toward Velay. Two sons of Bernard and Garsinde became bishops of Nîmes and LePuy,(36) respectively, and the family appears to have maintained control over the abbeys of Aniane, Gellone, and other religious establishments near Béziers and Agde.(37) In this period Bernard, like Roger of Carcassonne, refers to himself as princeps and marquis and his wife Garsinde bears the title of countess.(38)

Yet it is clear that no permanent principality in eastern Languedoc emerged from all this. Like Carcassonne, the House of Anduze fell upon evil days when its founders, Marquis Bernard and Countess Garsinde, died. The bishoprics of LePuy and Nîmes came under the control of rival houses.(39) The abbeys of Aniane and Gellone regained a large measure of [344] independence.(40) Family lands and rights were divided between Count Perre, who held Béziers, and Bermund, who succeeded to Anduze.(41) By 1050 only traces of Anduze's authority still remained, and much of this soon passed into the hands of the House of Trencavel which succeeded to their authority in Nîmes and Béziers.

Lest this be considered an unusual case, the story of the county of Provence during this period reveals much the same situation. About the year 950 King Conrad of Burgundy-Provence gave this part of his domains to a certain Count Boson of Arles.(42) Boson seems to have had two sons, an older one, Roubaud, and a younger one, William, both of whom bore the title of count.(43) William proved to be a man possessing unusual gifts as a warrior and administrator. With the assistance of the counts of the High Alps and the viscounts of Marseille and Fos, he drove the Moslems from Fraxinetum and reorganized disorderly regions east of the Rhone to create, almost single-handedly, a large, well-organized duchy, marquisate, or county of Provence.(44) He well deserved the title of duke which he bore from this time on, as well as that of Pater Patriae given him in a contemporary charter.(45) He controlled the royal fisc within his domains,(46) and held courts whose competence seems to have been unquestioned, and was in fact in every way independent of his nominal suzerain at Vienne.(47) During his lifetime only he bore the title of marquis while his brother had to content himself with that of count.(48)

When William died, about 994, his brother Count Roubaud seems to have succeeded him as head of the family, and to have assumed William's [345] title of marquis,(49) which be in turn passed on to his own son William II, who bore it as late as 1032.(50) Roubaud's daughter Emma married Count William of Toulouse.(51) Yet the heirs of William I also kept some authority. It seems probable that it was Williams' daughter who was the Constance who married King Robert of France, while her sister became countess of Clermont.(52) Soon after 1018 two other heirs of Marquis William, Bertrand and Josfred, ruled in Provence as co-counts also.(53)

Perhaps it was this diffusion of authority which explains the failure of the family of Provence to maintain the highly organized principality of its founder, Marquis William. However that may be, it seems apparent that by 1030, if not earlier, the power of the counts of Provence had begun to disappear. General courts ceased to be held and the counts of remote regions like Sisteron, Vintmille, Antibes, Nice, and Fréjus began to act in an independent manner.(54) So did the viscounts of Marseille, to whom William's heirs were forced to grant land belonging to the comital fisc as outright allods.(55) By 1050 what had, for the period, been an unusually well-organized principality had become one in which the major noble families were all but independent. It was also one which soon was to see even the marquisate disputed between the counts of Toulouse, the counts of Forcaiquier, and eventually the counts of Barcelona. After an excellent beginning Provence followed the same path leading to dissolution and disintegration which, we have already noted, was followed by Carcassonne and Anduze.

Leaving this southern area of the Midi let us next turn to two other examples of the same sort in a region quite remote from the Mediterranean. I refer to the House of La Marche and that of Angoulême which developed in a part of Western Aquitaine that lay between Poitou and the Garonne and from Limousin to the Atlantic. In choosing these two examples we find a particular interest in the fact that they represent two ruling families which were following diametrically opposed political policies and yet ended up with the same general results. The first example, [346] La Marche, represents a family which led the resistance of the nobles of this region to the advance of the Poitevin dukes of Aquitaine. The second, that of Angoulême, presents a picture of a ruling house which was, perhaps, the leading ally of these dukes during their period of expansion.

Let us first consider the House of La Marche. The origins of the first counts of La Marche are obscure. All we really know is that sometime about 950 a certain Marquis Boson, seems to have been in control of some territory in the northern part of the Limousin.(56) The location of his small principality -- which barred the route that the counts of Poitou had to follow south into the Limousin -- and indeed its very name of La Marche make one suspect that he was given this land by Count Raymond I of Rouergue or other Limousin nobles opposed to Poitevin ambitions. This, however, is only conjecture. It is clear, however, that once established in this region Count or Marquis Boson proved a doughty adversary to the dukes of Aquitaine. So did his successors. As a result of their prestige as warriors and their victories over their Poitevin adversaries they were able to add the county of Périgord to their holdings, by marrying its heiress,(57) and about the year 1000 also made an ally of the viscount of Limoges, whose sister wed one of their house.(58) At the turn of the millenium, under the leadership of Count Boson II, they appear to have headed a powerful coalition of anti-Poitevin nobles in the Limousin.(59) At the same time, their control of Périgord brought them into close alliance with the House of Toulouse who had long maintained friendly relations with the counts of this region. As late as 1040 this alliance was still in existence which helps to explain the marriage between Almodis of La Marche and Count Pons of Toulouse.(60)

Nevertheless the family of La Marche were not destined to create a strong principality in this part of Western Aquitaine. When Count Boson II died his heirs followed the prevailing practice and divided up his domain. Bernard got La Marche and Adelbert, Périgord.(61) So divided Boson's successors could not resist the dukes of Aquitaine, nor [347] were they able to rally the jealous bickering lords of the Limousin to any unity under their leadership. Soon La Marche had to surrender to Duke William the Great of Aquitaine(62) and by 1027 Limoges followed suit.(63) A principality which might have become strong disintegrated. By the end of the century a new Poitevin family held La Marche, the original basis of the power and authority of Count Boson the Old.

Equally instructive is an examination of how the counts of Angoulême failed. These counts were descended from Count Turpio, who established himself in this part of Aquitaine in the late ninth century. In 975 they and their cousins controlled not only Angoulême and Périgord, but also the county of Agen which they received from the counts of Toulouse.(64) They probably also had some authority over Saintonge.(65) When the county of Périgord came into the possession of the counts of La Marche, William Taillefer II, the count of Angoulême, appears to have been unhappy about it. At any rate he countered by allying himself to Count Fulk III of Anjou, whose sister he married.(66) Since Fulk was a close ally of the counts of Poitou this had the effect of making Angoulême the leading ally of the dukes of Aquitaine in this region. This alliance seems to have resulted in some solid advantages for Count William II. Though he did not gain Périgord, which remained in the hands of the family of La Marche, he was able, with Duke William's assistance, to expand his power into the Bordelais,(67) and about 1020 to become a power in Saintonge as well.(68) At the time of his death, about 1028, he was probably the leading magnate in this part of Aquitaine, allied to Poitou and Anjou and having a son who was married to the daughter of the reigning duke of Gascony.

This eminence, however, proved temporary and illusory. Upon the death of Count William in 1028 his sons, Aldouin and Geoffrey, began [348] quarrelling over their inheritance in the Bordelais.(69) Revolts broke out in Saintonge.(70) The duke of Aquitaine and the count of Anjou took advantage of the situation to intervene. As a result Saintonge was lost to the House of Angoulême.(71) Even more important a few years later another heiress to Gascony, the daughter of Duke William Sánchez, married into the House of Poitou.(72) From this union came claims that the dukes of Aquitaine had to Gascony, which in 1053 they were able to vindicate by force of arms. As a result, in this period what had promised to be a strong principality in Angoulême and neighboring regions had disintegrated into weakness. The counts of Angoulême became minor nobles dependent upon the dukes of Aquitaine, who had gained not only all of Aquitaine but Gascony as well.

Our sixth example of a failure to create a strong principality involves Catalonia and the House of Cerdanya-Besalu. By 975 this family, which was descended from Count Miró, the brother of Count Guifred, and who were cousins of the House of Barcelona, had begun to consolidate their authority over their part of Catalonia. In the course of this consolidation they began to adopt policies generally hostile to Count Borell II of Barcelona, who represented the older traditions of Guifred's marquisite over all the counties of Catalonia. By the time Count Oliba Cabreta had personally united his family's possessions in Besalu, Cerdanya, and Berga, a strong principality seemed a reality, not only one which controlled these central Pyrenean counties, but one which exercised an influence far beyond them, through family control of the great abbeys of Ripoll, Cuxa, Saint-Joan de les Abadesses, and even Lagrasse.(73) Yet such authority did not content Count Oliba. He therefore extended his domains to the north at the expense of the House of Carcassonne, and to the west at the expense of the House of Ampurias-Roussillon.(74) He took the title of marquis and [349] began to assume a leadership over Catalonia which the weakness and general incompetence of Count Borell II of Barcelona made it easy for him to do.(75)

When Count Oliba died in 990, by his will he divided his domains among his sons and his widow. One of the sons, Bernard, received Besalu with the promise of Vallespir too when his mother died. Guifred, the other, got Cerdanya, Confluent, and Berga. Perhaps this family's chief weakness had been its failure to control any bishoprics. This was soon remedied. After an effort to get Pope Benedict VIII to set up a new bishopric in Besalu failed,(76) Berengar, one of the family, was made bishop of Elne in 993,(77) A little later in 1008 Oliba, a brother of Counts Bernard and Guifred was made bishop of Ausona,(78) and in 1016 the archbishopric of Narbonne was purchased for a son of Count Guifred, who bore the same name.(79) Still a little later the bishopric of Urgell was purchased for still another scion of this house.(80) By the end of the second decade of the eleventh century the House of Besalu-Cerdanya had come to control bishoprics to the north, the south, the east, and the west of their domains.

Yet such power was not to last. When Count Guifred of Cerdanya-Confluent-Berga died, his will, as was the custom, divided his domains among his seven children.(81) Nor was his brother Count Bernard Taillefer more successful in keeping his lands intact after his death.(82) As this disintegration took place due to inheritance, leadership over Catalonia gradually slipped from the grasp of this family into that of the House of Barcelona, which was now taking charge of the advance south into Moslem territory which we know as the Reconquista. By the time Raymond, count of Cerdanya, was willing to allow the powerful Count Raymond Berengar I of Barcelona to purchase his allegiance in 1058, the future belonged to the House of Barcelona and a powerful principality of Besalu-Cerdanya was no longer a possibility.

There still remains one final family which we need to consider to round out our picture, that of Toulouse-Rouergue. In an earlier chapter we noted [350] how by 975 this family had failed to create an effective principality in the Midi, despite the considerable authority which it possessed.(83) This remained the situation after this period too. Yet throughout these years we can see that they still possessed sufficient authority so that the possibility that they might do so still existed. For example Raymond II of Rouergue was able to pass down to his son, Count Hugh, a certain suzerainty over Rouergue and neighboring regions,(84) and enough power over Narbonne to allow him to receive one-half of the 100,000 solidi purchase price of its archbishopric in 1016.(85)

Even more impressive is evidence of the influence wielded by Count William Taillefer of Toulouse, the cousin of Count Hugh, particularly in Western Languedoc. A charter dating from 1006 refers to William as count of Toulouse, Albi, and Quercy,(86) and another from Lézat, dating from 1015, calls him marchio.(87)Another bearing a date of 1025 from this same abbey refers to him as "marchio prefatus in pago Tholosano."(88) His position was of enough importance so that it was he to whom Pope John XIX appealed to restrain one of his milites who was usurping churches belonging to the abbey of Moissac;(89) and a little later his authority over Toulouse proper is shown in a charter in which he gave up dues which he was levying on the market of this city.(90)

Nor does his influence and authority seem to have been confined to this region alone. The will of the archbishop of Narbonne, Ermengaud, of 1005, mentions him in leaving him some falcons as a legacy,(91) and in the year 1020 he was among the notables present at a ceremony in which Countess Garsinde of Anduze and her sons gave some property to the abbey of Gellone.(92) We also find him present frequently in Provence where his wife Countess Emma had inherited considerable property from her father, Marquis Roubaud. Charters which mention both him and his wife in this region date from 992, 1005, 1006, 1015, and 1024.(93)

After Count William's death his son, Count Pons, seems to have inherited [351] from him and his mother Countess Emma considerable power and a good deal of property. Thus we find him, according to a charter of 1030, possessing considerable rights over certain castles in Albigeois(94) and in 1037 giving to his new wife Majore a number of castles, churches, and other rights in this region and in the Nimois and Provence.(95) In 1040, according to a charter, he gave to Cluny considerable property in Diens,(96) and in 1038 he divided with the Trencavels the purchase price of the bishopric of Albi.(97) In 1047 in still another charter in which he is called count palatine of Toulouse we find him giving to Cluny the abbey of Moissac.(98)

None of this means that either the family of Rouergue or their relatives of the family of Toulouse were able by 1050 to create any principality in this part of the Midi. But it does make clear that they were able to maintain considerable influence and considerable property in the form of rights over castles and Church establishments. Soon after 1050, Raymond of Saint-Gilles, the younger son of Count Pons and Almodis, was to build upon this prestige and such possessions to create a true principality in Languedoc, one that could match in power those which had already been formed by the counts of Barcelona and the dukes of Aquitaine.(99)

This leads us to an important question. What does a careful analysis of the failure of these seven families to create principalities during this period reveal to us concerning the political life of the Midi and Catalonia? Why did the efforts of the Houses of Carcassonne, Anduze, Provence, La Marche, Angouléme, Cerdanya-Besalu, and Toulouse-Rouergue all end in failure, just as had those in the preceding period in the Midi and in Catalonia? What went wrong?

We might begin by emphasizing that it is difficult to find an answer to this question in external enemies, with the possible exception of the House of La Marche. In every other example we have examined it is clear that while all these houses had outside enemies, it was not such enemies which kept principalities of a permanent nature from arising. We must look elsewhere. When we do we at once become aware that each of the [352] families who attempted to form principalities had one thing in common: the way in which they inherited their land and their authority. And in each of the cases which we have examined, disintegration followed hard upon the division of the principality or domains in question among a number of heirs, a process which was sometimes repeated a number of times. In Carcassonne, for instance, the lands and the authority of Count Roger the Old in a very short time found their way into the hands of a number of members of his family:(100) his brother, his nephews, and his sons. The same kind of a divisio followed the death of Marquis Bernard of Anduze and his wife, Countess Garsinde.(101) In Provence in the course of two generations the patrimony of Marquis William I in like manner was divided among his heirs and his relatives, who were descended from Marquis Roubaud, his brother, and the counts of Toulouse.(102) Nor were the ruling families which we have been examining in Aquitaine immune from similar developments. The division of La Marche and Périgord among the heirs of Count Boson II,(103) like the quarrels between Aldouin II and Geoffrey II of Angoulême(104) over their inheritance in the Bordelais, show the same process at work.

Even more striking is the way in which Cerdanya-Besalu disintegrated after first Count Oliba Cabreta and then his heirs divided up their property and authority among their descendants.(105) The division of authority between the House of Rouergue and that of Toulouse seems equally revealing as a cause of failure to develop a principality here.(106) Therefore it seems fair to say that ruling families were unable to create principalities in the Midi and in Catalonia chiefly because of the way they insisted on dividing their domains and their rights among various heirs: widows, brothers, sons, and daughters. Under such a system no principality could last for more than two generations at the most; and all political power and cohesion were lost in the resulting fragmentation which took place.

There seems to be some indication, though, that some of the rulers of this period were aware of this fact and strove to deal with it within the [353] limits which the family system imposed upon them. When Count Roger of Carcassonne provided in his will that none of his property or his rights which he left to his heirs could be disposed of except to another member of the family,(107) he was trying to keep disintegration at a minimum. It also seems possible that the system whereby the oldest member of the family bore the title of marquis after William I's death was used by the House of Provence to keep a certain family unity alive.(108) Perhaps we can see the same kind of family headship in the case of the House of Toulouse-Rouergue with such a position passing from Raymond Pons to Raymond I and II of Rouergue and then back to William Taillefer and Pons of the Toulousain branch of the family.(109) Even in Cerdanya-Besalu we seem to see a pattern whereby brothers rather than sons succeed to certain portions of the family inheritance.(110) In Aquitaine Aldouin II, the older brother, clearly had a position superior to the younger, Geoffrey II.(111)

Examination of certain other ruling families of the Midi seems to show similar efforts to maintain a family solidarity, efforts which were often relatively successful. Such seems to have been the case with the viscounts of Marseille whose family always seems to sign charters en masse,(112) or the counts of Gevaudun where Count Pons and his brother Bertrand seem to act in concert.(113) Such attempts, however, to make a family an effective political instrument seem to have been doomed to failure. No family system [354] could provide the necessary clear political authority which even the government of Southern France in the late tenth century demanded. Family control, however limited and circumscribed it might be, could only end in divided and discordant policies which made government impossible. The prevailing system of family inheritance and control, then, was the principal reason why principalities failed to develop in the period after 975, as they had before that date.

In addition to family control and the system of inheritance it prescribed, however, we find still another tendency which made territorial states of any size all but impossible in the Midi and Catalonia. I refer to the tendencies, even the pressures, which we find to transfer or change to allodial ownership all lands or rights given out by ruling families to those who were their subordinates. This was no new tendency in these regions. It had existed since late Carolingian times, as we have noted. What is important to note is that it continued as a very powerful and pervasive one.

Provence, which had been organized with great care by Marquis William I in the last quarter of the tenth century as an effective principality, gives us a good example of this tendency at work. As far as we can tell Marquis William originally granted to his lords domains over which in most cases he kept some authority, especially over the castles which they built upon them. He also kept in his own hands, judging from charters, a large amount of property which belonged to the fisc of the counts. Soon after his death, however, this situation began to change. Provence's nobles began to transform their rights over such lands and castles into full allodial ownership, and in addition to take over land belonging to the comital fisc. In some cases this appears to have been done with the consent of William's successors. In other cases it was the result of usurpation. When they sought to justify such a change in the status of land we find the usurpers quoting Roman law in the charters of the period, saying that this law gives a man the right to do as he wishes with his property. This seems to have been the case in two charters dating from 1018 and 1028(114) from the region of Nice. Even more specific as an example of this point is a charter of Lérins which dates from 1035 in which an assembly of nobles, including a bishop from this region, agree that any man, miles or rusticus, can freely give land "ex suo honore" to the abbey of Lérins.(115) No clearer evidence of the transformation of feudal to allodial land could be found.

Such statements in charters of the period, however, were merely a justification [355] of what was already a common practice in this part of Provence. In 1002, 1003, and 1032(116) our documents tell us of land given as allods to Church establishments which the owners clearly state were originally grants given them by the counts of Provence. Nor was this practice confined to the area of the Riviera, that is to say Nice, Antibes, Fréjus. In 1010 we find a certain priest called Walbert giving to a church without comital permission land which he states he received from the count "ad medium vestem."(117) And a charter of 1032 tells us of another landowner who is giving property to the abbey of Saint-Victor of Marseille which he states his lord, Count Geoffrey of Provence gave to him "ad proprium allodem."(118)

It is when we examine the relationship which existed between the counts of Provence and their powerful subordinates, the viscounts of Marseille, however, that we can see this change from comital fief and fisc to allodial ownership most clearly illustrated. Some time between 1018 and 1032, for instance, we learn from a charter that King Rudolf of Burgundy, for a payment of four ounces of gold, renounced any rights he might have over a villa which was part of the fisc which belonged to the counts of Provence. What happened to this villa? It was given to Viscountess Odila of Marseille as an allod.(119) A little later, about 1030, we find references in another charter to Count Josfred and Count Bertrand giving up their rights to lands, again belonging to their fisc, which the viscounts of Marseille gave to the abbey of Saint-Victor.(120) By 1044 this practice seems to have become so general that a document dating from this year shows us Count Josfred agreeing in advance to give up his rights over any land which Viscount Fulco of Marseille wishes to give this same monastery.(121) When we add to this evidence from a charter of 1038 in which Count Josfred relinquished his rights over a number of castles and property which were in the possession of the same Viscount Fulco,(122) we can see how slowly but surely the allodial principle triumphed in Provence over the rights which the counts possessed there as overlords.

There is a good deal of evidence that this same tendency to transform [356] fiefs into allods was found in other parts of the Midi as well as in Provence. In Dauphiny a later charter tells how the fortresses of Dauphiny, which belonged allodially to the bishops of Grenoble, were usurped by those who possessed them and transformed into allods during this period.(123) In Aquitaine the specific example of the castle of Blavia is another case in point. Our sources tell us that initially, following its conquest, this fortress was given to the count of Angoulême as a fief by the duke of Aquitaine.(124) Yet a little later it had become the allodial possession of this family, and in 1028 was given out by Count Aldouin as a fief.(125) Just so a new castle at Limoges, belonging to the duke of Aquitaine, soon after 1032 was transformed into an allod belonging to the bishop of Limoges. The wording of the charter which reveals to us this change seems to be a masterpiece of calculated ambiguity. It says "ipse Willelmus comes totum illum fevum dedit mihi in allodem extra episcopum."(126)It should not surprise us to find a charter of this same region dating from 1040 which contains the old phrase that a man has the right to do as he wishes with his property according to Roman law, that usual defense of allodial rights in regions where a resistance to fiefs is to be found.(127)

In Languedoc we notice the same tendency at work. That well-known gift of property made to Guy or Guillem, the founder of the House of Montpellier, by Count Bernard of Melgueil in 985 states that this land is being given to him "pro suo servicio vel benevolentia," but it is clear that the property in question was an allod and not a fief.(128) In 989 a charter, which tells us of the exchange of property between Viscount William of Béziers and Bishop Matfred cites that familiar phrase that a man has a right to dispose of his property as he wishes.(129) In 1006 another somewhat later document shows us a certain Austinde citing Roman law in the same way in exchanging with Abbot Deusdet land which he says he got as an "allodem vel beneficium" from the viscount of Narbonne.(130) Again in 1030 Bishop Peter of Gerona, in violating the will of his father which forbade giving family property outside of the family, cites the authority of Roman law and gives a church and some other property to the abbey [357] of Montolieu.(131) Even such an important lord as Count Pons of Toulouse does the same in citing the authority of Roman, Salic, and Gothic law in a charter giving to his wife Majore a bridal gift in 1037 of a number of castles and churches in Albi, Nîmes, and Provence.(132) Perhaps he did so because his title to such property was doubtful. The land in Albi was probably some over which his Rouergue cousins had certain rights. That in Provence property which he possessed jointly with other members of the family of the counts of Provence.

It does not seem necessary to cite similar examples of property given as allods by the counts of Barcelona to important nobles of Catalonia during the period of the late tenth or early eleventh century. Such grants, suffice it to say, were as numerous as those found elsewhere in the Midi.(133) All of which seems to indicate that the failure of principalities in the Midi and Catalonia during this period was as much the result of the loss of control over land which was given out to supporters as it was to the distribution of land among numerous heirs. By 1050 principalities had failed to develop, in no small measure, because the tendency of property to become allodial still was so strong that it kept rulers from using conditional grants of land as a method of building up their power over a wider region. What helped destroy the fisc of the Carolingians in the Midi and Catalonia continued to deplete those of their successors who were attempting to build up authority in the same regions.

But we need also to recognize two other things which worked against principalities and their formation during this period. The first was the growth of castles, which we mentioned in an earlier chapter. In a society in which the allodial principle was strong, and the government weak and limited in its effectiveness, castles built upon domains belonging to them could often be a positive danger to rulers. Such castles were either from the start allodially owned by those who built them or tended to become [358] so. In either case they gave to their possessors a practical independence which made it difficult for a ruler or potential ruler to deal with them. This, of course, is the reason for those agreements which we have noted between castellans and overlords of such fortresses, agreements which are to be found in our region at such an early date.(134) But if the possessor of such a castle disregarded the agreement which he had made, what then? How could he be brought back to his sworn allegiance or punished by the overlord with whom he had made such an agreement, a man who had limited military power at his disposal and no real means at hand for reducing such a castle by siege. The pages of Ademar of Chabannes, with his accounts of unsuccessful sieges of the fortresses of Aquitaine during this period, show how difficult action in such matters could be for a ruler.(135) Thus wherever they were numerous, castles, in practice, generally tended to reduce the authority of a ruler of a potential principality and enhance the power of those who actually held such fortresses. Castles, then, helped further weaken principalities south of Poitou and Burgundy.

Finally, the attempt on the part of the Church to curb the abuses of the new militarism, as they affected Church property and authority, also helped to make it difficult for ruling families to maintain or even build up principalities. As long as the Church directed its attention to militarism per se this was not so. But when the Church went beyond the Peace of God, with its emphasis upon ending exactions upon Church land, and began to attack secular control of the Church per se this was another matter.(136) Now the Church was striking at the very basis of family rule in the Midi and Catalonia, its control over the land. In so doing, therefore, it helped make principalities more difficult to maintain, for to question family control in the election of Church officials was not a matter which the rulers of the Midi could view lightly. Whether it was an Aniane or a Gellone escaping from the control of the House of Anduze,(137) or a bishopric of Limoges lost to the viscounts of Limoges,(138) or an archbishopric of Bordeaux ceasing to be a local Gascon affair and becoming more Aquitanian,(139) [359] the effect was the same: to diminish the ability of a local family to control its local area and thus also to reduce this family's chances of building and maintaining a principality of some strength and permanence.

By 1050, then, with the exception of the counts of Barcelona and the dukes of Aquitaine, whom we will examine in detail later on, none of the ruling families of the Midi and Catalonia had been able to create strong, permanent principalities. All efforts to do so foundered as a result of the system of inheritance which was used, and the tendency toward allodial possession of land. To these two older tendencies one should add the difficulty of controlling the castles which had appeared and the milites who garrisoned them. Even the Church, the real bulwark of the authority of the ruling families of the Midi and Catalonia also proved to be an uncertain basis of authority as its reaction against the new militarism threatened to take the form of an opposition to all secular control of its establishments.

As a result all government that existed during this period more and more came to rest in the hands of a large class or aristocracy of nobles, churchmen, castellans, and important landowners whom contemporary documents call principes. Whatever order was kept, they kept. Whatever government existed, existed only by their consent. The princeps having failed, the principes took over, and to them we must look to see how the Midi and Catalonia were really governed.

Notes for Chapter 17

1. Catalunya Carolingia, II, 211-213.

2. Ibid., pp. 170-174.

3. Ibid., pp. 197-200, 241-244.

4. Cart. de Paunat, no. 12.

5. Ademar de Chabannes, III, 34, 45, pp. 157, 167. Note that King Robert also failed in attempting to chastize Eudes, princeps of Deols, in 1014 (ibid., III, 51, p. 174).

6. "Aquitaniae Historiae Fragmentum," in ibid., p. 205.

7. On the activities of these early Capetians in this part of France see P. Lauer, Robert I et Raoul de Bourgogne, rois de France.

8. See J. Richard, Les ducs de Bourgogne et la formation du duché du XIe au XIVe siècle, pp. 30-137. See also a charter of 1022 which attests to his authority in Velay (Cart. de Saint-Egidius, no. 98).

9. For information concerning the power of the counts of Clermont in Auvergne see M. Boudet, Introduction to the Cart. de Saint-Flour, p. cxxii. See references to their authority over the abbey of Sauxillanges in Cart. de Sauxillanges, nos. 279, 401, 402, 476, 572, 635. See also Cart. de Lérins, no. 214.

10. Charters of 986, 999, and 1011 attest to the influence of the counts of Gevaudun in Auvergne and especially at Brioude Cart. de Brioude, nos. 91, 321, and Grand Cart. de Brioude, no. LIV.

11. See references to Duke William of Aquitaine in Chartes de Cluny, III, nos. 2277, 2682; Cart. de Brioude, nos. 92, 323; and Chartes de Charroux, no. 4.

12. See Introduction to Cart. de Saint-Flour, pp. cvii-clxvi.

13. See G. de Manteyer, "Les origines de la maison de Savoie en Bourgogne (910-1060)," in Mélanges d'archéologie et d'histoire de l'Ecole française a Rome, XIX (1899), and "Les origines de la maison de Savoie en Bourgogne, Notes additionnelles," in Le Moyen Age, V (1901). See a charter of 1015 which mentions their authority in this region (Cart. de Grenoble, no. 18). For a somewhat later period see Cart. de Sai nt.Sulpice, no. 2; Cart. de Saint-Chaffre, nos. 434, 435; Cart. de Grenoble, nos. 19, 20; and Cart. de Saint-André-le-bas, nos. 211-213.

14. On the counts of Gravaisdun and their authority in Dauphiny which included control of the bishopric of Grenoble see Chartes de Cluny, IV, no. 2307; Cart. de Saint-Chaffre, nos. 355, 356; and Cart. de Grenoble, nos. 15, 16, 34. About 1050 this family seems to have lost control over parts of this Alpine region to the more aggressive counts of Savoy.

15. On Count Artald, who founded a new line of counts in the Lyonnais at the end of the tenth century see E. Fournial, "Recherches sur les comtes de Lyon du IXe et Xe siècles" in Le Moyen Age, LVIII (1952), 231-249. See a charter of 992 which probably refers to Count Artald in Cart. de Savigny, no. 53. For a reference to Count Gerald, who probably succeeded him, note a charter of 1017 in ibid., no. 602.

16. We find references to this noble house which controlled Valence in Chartes de Cluny, III, nos. 1715, 1716; IV, no. 2832; Cart. de Saint-Petrus Aniacensis [LePuy], no. 425; and Cart. de Saint-Chaffre, nos. 314, 315.

17. These viscounts are mentioned in Cart. de Vienne, no. 33, and Cart. de Saint-André-le-bas, no. 140.

18. The pages of Ademar de Chabannes reveal the quarrels of the lords of the Limousin, such as the struggle between Viscount Guido of Limoges and Boson II, count of La Marche in 999 over a castle (Ademar de Chabannes, III, 35, p. 159); or the quarrel between Viscount Guido and the bishop of Angoulême about 1000 (ibid., III, 36, pp. 159-160). Note also the mention of how this same Viscount Guido's castle was besieged by five counts in 990 ("Aquitaniae Flistoriae Fragmenta," in ibid., p. 205).

19. On the situation which existed in Gascony during this period see C. Samarin, "Les institutions en Gascogne" in F. Lot and R. Fawtier, Histoire des Institutions de la France, I, 186-192.

20. On the relations maintained between the counts of Urgell and those of Pallars see Liber Feudorum, nos. 120, 121. On those between the latter and Aragón see ibid., nos. 58, 59. See also ibid., nos. 36, 57, 107, 109, and Hist. Gén. de Lang., V. no. 212. On Ribagorça see Hist. Gén. de Lang., nos. 115 and 117.

21. See Chapter II, Section IV, on the earlier failure of the House of Toulouse-Rouergue to build a principality in this part of the Midi.

22. The will of Count Roger the Old of 1002 mentions certain abadias which he has already given to his son Bishop Peter of Gerona (Hist. Gén. de Lang., V. no. 162). For Roger's earlier control of these abbeys see Cros-Meyrévielle, Documents, nos. 30, 31, 34, and Cart. de Carcassonne, IV, 3. Charters which refer to this family's later control of these church establishments are found in Cros-Meyrévielle, Documents, nos. 38, 39; Cart. de Carcassonne, II, 75; and Hist. Gén. de Lang., V, no. 189.

23. Count Roger's will of 1002 mentions the county of Razes(Hist. Gén. de Lang., no. 162).

24. Roger, son of Count Roger the Old was count of Foix in 1034 (Cros-Meyrévielle, Documents, no. 37).

25. Hist. Gén. de Lang., V, no. 130.

26. Ibid., no. 162, and Cros-Meyrévielle, Documents, no. 37.

27. Hist. Gén. de Lang., V. no. 137. In 1002 he is also called princeps Carcassonne (ibid., no. 161).

28. Cros-Meyrévielle, Documents, no. 31.

29. See P. Ponsch, "Le Confluent et ses comtes du IXe au XIIe siècle" in Etudes Roussillonnaises, 11(1952), 285-295.

30. Hist. Gén. de Lang., V, no. 162.

31. Ibid.

32. Ibid., no. 251.

33. See charter of 1025 (ibid., no. 185); or charters of 1034 (Cros-Meyrévielle, Documents, nos. 37, 38, 39); or charter of 1036 (Hist. Gén. de Lang., V, nos. 208, 209, 210); or that of 1050 (ibid., no. 229). All reveal the progressive fragmentation of the domains of Count Roger the Old.

34. Frédélon, of the castle of Anduze, the progenitor of this clan, is first mentioned in an earlier ninth century charter (Cart. de Nîmes, no. 32).

35. See the will of Viscount William of Béziers dating from 990 (Cart. de Béziers, no. 49). Garsinde is mentioned again in charters of 1006, 1013, 1024, and 1029 (ibid., no. 54; Cart. de Conques, no. 18; Hist. Gén. de Lang., V, no. 182; and Cart. de Gellone, no. 6). The charters of 1013 and 1024 also mention her husband Bernard.

36. Frédélon (or Frédol) of the castle of Anduze is mentioned as being made bishop of LePuy in Chronique de la Monastère de Saint-Pierre de LePuy, in Hist. Gén. de Lang., V, cols. 18-20. He also is referred to in Cart. d'Aniane, nos. 146-148. Both he, his brother, the bishop of Nîmes, and other members of the family are mentioned in a charter of 1020 (Cart. de Nîmes, no. 120).

37. A charter of 1037 mentions that the abbatias of Aniane and Gellone are under the control of Bermund and his brother, Count Peter (Hist. Gén. de Lang., V. no. 206).

38. Bernard is called marquis in 1013 (ibid., no. 171), and princeps in 1024 (ibid., no. 182). Garsinde is called countess in both charters.

39. See Cronique de Saint-Pierre de LePuy, cols. 21-24, on how a new family got control of the bishopric of LePuy. On the change in the bishopric of Nîmes which went to the Trencavel family see Hist. Gén. de Lang., V, no. 217.

40. Hist. Gén. de Lang., V. no. 206.

41. Ibid. On Count Perre's holdings later on see ibid., no. 209.

42. Chartes de Cluny, II, no. 1052.

43. Cart. de Saint-Victor, nos. 29, 598.

44. Raoul Glaber, Historiarum libri V, I, 4, 9, p. 12. See also a later charter of 993 which tells how Count William organized Provence after the expulsion of the Moslems (Cart. de Saint-Victor, no. 77).

45. Charters from 979 and 992 call him marquis (Cart. de Saint-Victor, no. 1042; and Hist. Gén. de Lang., V, no. 153). This later charter calls him princeps also.

46. The charter of 992 says the land William took from the pagans was his as a gift of the king [of Burgundy] (Cart. de Saint-Victor, no. 1042). Other later charters dating from 1018-1032, 1030, and 1040, refer to regalian and comital rights over villas belonging to the fisc (ibid., nos. 1061, 155, 172).

47. Ibid., nos. 29, 654.

48. See note 45.

49. Chartes de Cluny, III, no. 1987.

50. Cart. de Saint-Victor, nos. 652, 656; and Hist. Gén. de Lang., V, no. 172.

51. Cart. de Lérins, no. 149.

52. Raoul Glaber, Historiarum libri V, III, 9, 40, p. 89.

53. A charter of 1030 calls Josfred and Bertrand counts of Provence (Cart. de Saint-Victor, no. 155). In one of 1031 Bertrand calls himself "comes vel gubernator Provintiae regionis" (ibid., no. 455). In 1034 another calls them principes of all Provence (ibid., no. 333).

54. See Cart. de Lérins, nos. 40, 113, 144, 167 for examples of this independence.

55. For an example of this comital bounty see Cart. de Saint-Victor, no. 34.

56. On the possible origins of this house see R. Mertier, La Senechausée de la Basse-Marche, pp. 21-26. It is in 958 that Boson I is first called marchio. (L. Albanes-Chevalier, Gallia Christiana Novissima, II, Inst., col. 169).

57. Ademar de Chabannes, III, 34, p. 156.

58. Ibid., III, 25, p. 148.

59. Ibid., III, 34, pp. 156-157.

60. See reference made to Almodis and her husband Count Pons of Toulouse in a charter of 1047 which gave Moissac to Cluny (Cart. de Tulle, XXII, 154-155).

61. Ademar de Chabannes, III, 20, p. 138.

62. "Aquitaniae Historiae Fragmenta," in ibid., p. 208.

63. See reference to the presence of Duke William of Aquitaine in 1027 (Cart. de Saint-Etienne de Limoges, no. 69).

64. Ademar de Chabannes, III, 20, p. 138.

65. See references to their land and their authority in Saintonge in 1021 (Cart. de Saint-Jean d'Angély, no. 251).

66. Ademar de Chabannes, III, 41, p. 163. Ademar says that Fulk held castles in Poitou and Saintonge "pro bene fico," that is to say as fiefs from them.

67. Ademar tells of the help which Count William Taillefer of Angoulême gave to Duke William of Aquitaine in his struggle with Boson II of La Marche (ibid., III, 42, p. 165). He also mentions his expansion into the Bordelais and his authority there (ibid., III, 68, p. 194).

68. On control over Saintonge exercised by Count William in 1024 see ibid., III, 66, p. 186.

69. Ibid., III, 67, p. 193.

70. Ibid., III, 66, p. 186.

71. Ibid. See also proof of ducal power in this region in Cart, de Saint-Jean d'Angély, nos. 148, 197, 244, 252. See Count Aldouin of Angoulême mentioned as present at a conclave in 1030. He may have still kept some authority there at this period (ibid., no. 158). By 1037, apparently this is no longer true (ibid., no. 42). See also Cart. de Notre Dame de Saintes, no. 1 (which dates from 1947).

72. On this ducal connection with Gascony which preceded its conquest soon after 1050 see Ademar de Chabannes, III, 39, pp. 161-162.

73. See R. de Abadal i de Vinyals, L'abat Oliva, bisbe de Vic, i la seva época, pp. 15-49, and Ponsch, "Le Confluent et ses comtes," pp. 270-290 on the power of the House of Cerdanya-Besalu in this period.

74. Ponsch, "Le Confluent et ses comtes," pp. 290-295.

75. Ibid., pp. 289-290.

76. C. Brousse, "La vicomté de Castellnou," in Etudes Roussillonnaises, III (1953), 115-120.

77. Marca hispanica, I, appendix 176-178.

78. Hist. Gén. de Lang., V, no. 158.

79. Ibid., no. 251.

80. Ponsch, "Le Confluent et ses comtes," p. 295.

81. Dom Luc D'Achery, Spicelegium, III, 392.

82. Ponsch, "Le Confluent et ses comtes," pp. 302-305.

83. See Chapter II, Section IV.

84. Cart. de Conques, no. 8.

85. Hist. Gén. de Lang., V, no. 251. Count Hugh is also mentioned in the Narbonnais in 1032 and 1035 (ibid., nos. 198, 207).

86. Ibid., no. 165.

87. Ibid., no. 173.

88. Ibid., no. 184.

89. Ibid., no. 195.

90. Cart. de Saint-Sernin, nos. 134-137.

91. Hist. Gén. de Lang., V, no. 164.

92. Cart. de Gellone, no. 6.

93. Hist. Gén. de Lang., V. no. 153 (992), 172 (1015), 182 (1024); Cart. de Saint-Victor, nos. 15 (1005), 653 (1006).

94. Cart. de Conques, no. 34. See also the reference to him and to his control over a castle in the Albigeois in 1040 in Hist. Gén. de Lang., V. no. 217.

95. Ibid., no. 206.

96. Chartes de Cluny, IV, nos. 2948, 2949.

97. Hist. Gén. de Lang., V. no. 214.

98. Cart. de Tulle, XXII, 154-155.

99. See an excellent account of the early career of Raymond of Saint-Gilles, the younger son of Count Pons, in J. Hill and L. Hill, Raymond IV de St. Gilles.

100. See Hist. Gén. de Lang., V, nos. 185, 208, 209, 229, and Cros-Meyrévielle, Documents, nos. 37-39.

101. Hist. Gén. de Lang., V, nos. 206, 208.

102. See notes 50 and 51 on how members of this family divided their land and authority.

103. Ademar de Chabannes, III, 45, pp. 167-168.

104. Ibid., III, 66, p. 193.

105. D'Achery, Spicilegium, III, 392.

106. See notes 84-92 on the way this house found its possessions and its authority divided.

107. Hist. Gén. de Lang., V, no. 162.

108. See notes 48-51 on attempts to maintain a head of the House of Provence.

109. Chapter II, Section IV, describes how the headship of the family passed from the Toulousain house to the Rouergue branch before 975. Then, following the death of Count Raymond II of Rouergue the title of marchio passed to Count William of Toulouse (see charter of 1015 in Hist. Gén. de Lang., V, no. 173). William kept this title and passed it on to his son Count Pons after his death.

110. See Abadal, L'abat Oliva and Els Primers Comtes Catalans on this Catalan pattern of inheritance which was repeated a number of times.

111. Ademar de Chabannes, III, 67, p. 193.

112. For examples see a series of charters dating between the years 1000 and 1030 which reveal action by the entire viscontal family of Marseille (Cart. de Saint-Victor, nos. 18, 69, 110, 111, 155, 169, 185, 249, 585). See also a charter of 1029 in which seventeen descendants and heirs of a certain Martin, called Blanchi, join in giving some property to Saint-Victor de Marseille (Ibid., no. 656).

113. Pons and Bertrand of the House of Gevaudun are called "clarissimi consules Aquitaniae" in Cronique de Saint-Pierre de LePuy, col. 14-16. They are also shown acting jointly in two charters which date from 995 and 999 (Chartes de Cluny, III, no. 2305; and Grand Cart. de Brioude, no. LIV). For the distant Riviera region of Provence see a charter of 1041 which shows that Otto and Conrad, as co-counts of Vintmille were acting similarly in concert (Cart. de Lérins, no. 167).

114. Cart. de Nice, no. 11 (1018); Cart. de Saint-Pons de Nice, no. 4 (1028).

115. Cart. de Lérins, no. 74.

116. Cart. de Nice, no. 18 (1002), 19 (1003); Cart. de Saint-Victor, no. 504 (1032).

117. Cart. de Saint-Victor, no. 133. For two other charters mentioning allodial land which was once given out by lords see ibid., nos. 227, 548.

118. Ibid., no. 504.

119. Ibid., no. 1061.

120. Ibid., no. 155.

121. Ibid., no. 34.

122. Ibid., nos. 447, 448.

123. Cart. de Grenoble, no. 16.

124. Ademar de Chabannes, III, 42, pp. 165-166.

125. Ibid., III, 67, p. 193.

126. Cart. de Saint-Etienne de Limoges, no. 28.

127. Cart. de Beaulieu, no. 104.

128. Liber Instrumentorum Memorialium [Montpellier], no. 70.

129. Hist. Gén. de Lang., V, no. 155.

130. Ibid., no. 166.

131. Ibid., no. 189.

132. Ibid., no. 206.

133. See for example a charter of 977 in which Count Borell II sold land and castles to Lord Unifredo (Cart. de San Cugat, no. 126); or a sale of land in Urgell by the same count in 979 to a certain Rescindo (Hist. Gén. de Lang., V, no. 132); or a gift of land and castles as allods in 980 by Count Borell to a son of Viscount Guitard (El Archivo Condal de Barcelona, no. 186); or the sale of land by this same count to Vivas in 989 (ibid., nos. 211, 214, 215); and to Lord Hennego in the same year (Cart. de San Cugat, no. 239). See other examples in charters dating from 989, 990, 992, and 995 in Cart. de San Cugat, nos. 240, 295; and El Archivo Condal de Barcelona, nos. 225, 232. Obviously in this period the count of Barcelona was disposing of his comital fisc as allods to his nobles and fideles.

134. See Chapter I, Section V, on the conditions by which castles were held during this period in the Midi and Catalonia.

135. On these struggles over control of castles in the Limousin and the rest of Aquitaine see Ademar de Chabannes, III, 34, pp. 156-157; 35, p. 159; 42, p. 166; 45, p. 167; 51, p. 174; 66, p. 186; 67, p. 193.

136. See for example the charter of 1039 from the Limousin which forbade lay control of the election of abbots or lay intervention in the affairs of the abbeys of Uzerche and Saint-Martial (Cart. d'Uzerche, no. 1039).

137. Hist. Gén. de Lang., V, no. 206.

138. Cart. de Saint-Etienne de Limoges, nos. 175-177.

139. Ademar de Chabannes, III, 69, p. 194.