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

Archibald R. Lewis


The Breakup of Principalities and the Court System

[195] The story of the decline and disappearance of royal authority and influence in Southern France and Catalonia brings up a most important question. What took its place beyond such influence as could be exerted by the Papacy, Cluny, and the caliphs of Cordova? Were the important noble families who emerged as masters of local regions and the local Church able, as a result, to create principalities on the bones of a decayed Carolingian system? Such seems to be the contention of a recent historian who has examined this question and who appears to postulate such a development as all but automatic.(1) Yet there is little evidence that he is correct. The fact seems to be instead that when Carolingian royal control of the Midi and the Spanish March ended in the last years of the ninth century, the leading magnates were unable to create territorial states in its place, despite all efforts to do so.

What then did happen? The story of the first three quarters of the tenth century is one not of the success but of the failure of the territorial state in these regions, and of a continuing decay of such machinery of government as survived -- except perhaps in Catalonia which in many ways presents a unique situation. By 975 what had emerged in Southern France was not principalities, but a system which was so disorganized in a political sense that it approached anarchy -- an anarchy that only slowly began to develop into a more orderly pattern of government.

Perhaps the best way to make clear this continuing decay of governmental organization and the failure of organized principalities to emerge [196] is to examine our regions in 900 and contrast them with the situation to be found in 975. At the earlier date, as we have already emphasized, political authority had moved from the Northern French monarchs into the hands of a group of noble families. One of these had assumed a royal title and controlled the general region of the Rhone Valley. The other eleven dominated the rest of Southern France and the Spanish March as counts -- two in Septimania, three in Catalonia, one in Gascony, and four in Aquitaine. Assisting them and sharing a measure of authority with them were a number of viscounts who had already made their appearance at Vienne, Nîmes, Béziers, Narbonne, Carcassonne, Toulouse, Roussillon, Catalonia, Albi, the Limousin, Angoulême, and Velay -- at least thirteen in all, if not more.(2)

What do we find by 975? We find that the number of independent comital and viscontal families has multiplied, until we can list some hundred and fifty who hold their authority by hereditary right. In some regions like Provence, the Limousin, Auvergne, and Languedoc these families are so numerous that it is all but impossible to disentangle them or to discover with any accuracy their origins. Nor does a catalogue of such families exhaust the list of those in control of these regions, for by 975 in certain areas like Languedoc, Auvergne, and Catalonia, for instance, we find minor officials, vicars and comtors, whose independent authority over local regions seems almost as great as that exercised by counts and viscounts. The number of families in positions of authority had so multiplied in most of the Midi that the age of the princeps had become the period of the principes. Nor does the above suffice in explaining the extent of the diffusion of political power. For we must add to such secular ruling families ecclesiastical magnates as well; bishops like those of Limoges, Angoulême, LePuy, and Grenoble, or archbishops like those of Lyon, Vienne, and Narbonne. Such churchmen during this period were often the almost unchallenged masters of large parts of the Midi.

How did this decay and fragmentation of principalities come about? Why were families who controlled important portions of the Midi and Catalonia in 900 unable to create any form of principality -- feudal or nonfeudal, which was capable of resisting progressive disintegration? And what happened to the governmental system of the Midi and the Spanish March as a result of such failures? These are the questions which demand answers. One might approach the answers in various ways, but [197] perhaps a good method is to concentrate on a certain number of the ruling families who appear to have controlled the Midi and Catalonia in 900, and to trace their history and that of the regions where they were powerful down to 975. We need not consider all twelve of the families who appear to have been dominant in 900, but only the more important ones. What should emerge is a general pattern which may allow us to make important generalizations.

Interestingly enough, it is not difficult for us to ascertain the relative importance of leading families in the Midi and Catalonia during these years. A number of terms or titles used to describe them in our sources make clear to us what individuals were endowed with more than ordinary authority or prestige. These terms are "king," "dux," "marchio," and "princeps," which, while they are not always precise, distinguish those who bore them from the less important counts and viscounts. Using such titles as a guide we discover that we can with profit focus our attention on the House of Poitou, the family of the Bernards of Auvergne and their successors, the kings of Provence and their principal supporters, the family of Hugh of Arles and those who succeeded them in authority, the House of Toulouse, the counts of Ampurias-Roussillon, the heirs of Counts Guifred and Miró in Catalonia, the counts of Pallars and Ribagorça, and the dukes of Aquitaine. We might also pay some attention to the House of Angoulême-Périgord and the family of Carcassonne-Razès, though they are not exactly similar to the examples cited above.

The eight families listed above, and their successors, all had representatives between 900 and 975 who were called "king," "dux," "marchio," or "princeps." For instance William the Pious is called duke of Aquitaine and the successor to his authority in Auvergne, Count Guy of Clermont, bears the title "princeps Arvenorum."(3) Ebles Manzur and his heirs of the [198] House of Poitou are variously referred to as "count" and "marchio" or "dux" until after 973 they bear the title of dukes of Aquitaine by hereditary right.(4) In the Valley of the Rhone not only is Louis the Blind called "king,"(5) but his chief supporter Hugh of Arles, is called "count," "marchio,"and "dux"(6) in contemporary documents, before he too assumed a royal Italian title. Their successors in authority over these regions bear similar titles. Conrad is called "king,"(7) Geilin of Valence -- Velay is referred to as "marchio" and "princeps,"(8) and William of Provence is variously styled "count" and "marchio" and "dux."(9)

Ermengaud and Raymond Pons of the House of Toulouse-Rouergue in 932 are called "principes Gothiae,"(10) and a little later, in documents dating from these years, the latter is referred to as "Gothorum princeps,"(11) "princeps Aquitanorum,"(12) and "dux Aquitanorum."(13) Just after 975 Count Geoffrey of Ampurias-Roussillon is called "dux,"(14)while Count Guifred and his heirs in Barcelona-Gerona-Ausona are variously styled [199] "count" and "marchio," "princeps," and "dux."(15) Those who succeeded Count Miró in Cerdanya-Besalu seem to have frequently been called "marchio,"(16) as were the counts of Pallars and Ribagorça in this period.(17) Late in the ninth century the hereditary rulers of Gascony are called "dux"(18) and probably continued to bear this title into the tenth century.

The majority of these titles which show a special importance or authority seem to be survivals from Carolingian times when the terms "king," "dux," and "marchio" were used to designate certain officials who possessed a special authority -- particularly in regions close to the frontiers.(19) There is one exception to this -- "princeps." This does not seem to have been found in the ninth century,(20) but is a return to pre-Carolingian eighth-century practices -- a significant fact in many ways, since it seems to reflect the different characteristics which this period had from those found at the time of Charlemagne, Louis the Pious, or Charles the Bald.

Leaving aside this matter of titles, however, let us examine in some detail the families whose representatives bore them. We might well begin with the House of Poitou and the lands ruled by this family down to 975. In doing so we will not consider the county of Poitou itself. Poitou proper was always directly involved in Northern French affairs and had a governmental system which resembles that of Maine and Anjou to the north [200] rather than that common to the rest of the Midi.(21) It will therefore be eliminated from consideration here. What we will consider, however, is the role of this family in their advances south and southwest into the Limousin, Saintonge, Quercy, and Auvergne which are part of our Southern French complex.

Now perhaps the most important thing for us to note in this regard is that for the greater part of this period the counts of Poitou, despite their assumption, at times, of rather grandiose titles and their wide claims, had little authority beyond their own county of Poitou. For a brief period, about 930, Count Ebles Manzur may have conquered the Limousin and entered Saintonge and Quercy,(22) but by the time of his death, the control exercised by his family had shrunk to authority over Limoges, where his son was bishop, and over the nearby abbey of Saint-Martial.(23) If we may believe Ademar of Chabannes, even this power at Limoges had to be exercised through an auxiliary bishop, and the nobles of the Limousin, after the death of Bishop Ebles, recovered control of this see for members of their own families.(24) After 950 or 960 castles constructed in the new march or county of La Marche by Count Boson and his successors seem to have effectively barred the count of Poitou's entrance into the Limousin.(25) As for Auvergne, until 955 a combination of local opposition and Toulousain influence kept the counts of Poitou from achieving control over this region. At least the western portion of it became theirs in 955,(26) but there is a serious question whether such control as they could establish here was very effective until the time of Duke William Iron Arm.(27)[201]

The counts of Toulouse-Rouergue continued to exert a powerful influence in Auvergne for some decades more(28) and so did the bishop of Auvergne, Etienne, and his family at Clermont.(29) Down to 975, then, it seems fair to say that the counts of Poitou had made some progress in Saintonge,(30) had controlled for brief periods parts of the Limousin, and had assumed after 955 a shadowy suzerainty over parts of Auvergne. That is all. The rise of this house to real authority as dukes of Aquitaine must be considered a later phenomenon, for beyond Poitou proper they were still unable, in this period, to create any effective principality -- only claims for the future.

Even more interesting is the case of the family of Bernard Plantevelue and its power over Auvergne and adjoining regions. Bernard's son William the Pious, duke of Aquitaine, was able to retain control over much of the land which his father ruled, Auvergne, Velay, Gevaudun, the Lyonnais, and Autun.(31) Upon his death about 914 his nephew William II seems to have succeeded to all of his honores except Autun.(32) William's successor, Count Acfred, however, was less powerful than his predecessors. He is not mentioned in charters from the Lyonnais or Velay.(33) Even more important his will of 927 left his property not to his relatives, but to his fideles and amicos, the viscounts and important landowners of Auvergne, particularly those who controlled Clermont and LePuy. And when we analyze his estate we are amazed to discover what a limited one it is, [202] consisting of a few villas and other property scattered through Auvergne, Velay, and Gevaudun.(34) What must have represented formidable power and huge estates in Duke William the Pious' time was reduced to limited authority and small property holdings at the time of Count Acfred.

Furthermore following the death of Count Acfred, no true successor to even this authority appeared. Certainly Raymond Pons of Toulouse tried to take Auvergne about 940-941,(35) and Count William Caput Stupe of Poitou achieved a partial control of the region about 955,(36) But neither of these nor native houses like the viscounts of Clermont or those of LePuy succeeded in building up a true principality during this period. Auvergne and Velay remained regions controlled by viscounts and comtors and the prey to the ambitions of neighboring lords in Poitou, Rouergue, and Gevaudun.(37) As for the Lyonnais, by 975 it had passed to the control of a completely new dynasty which was descended from important landowners of this region, who seem to have had no connection with the previous or present rulers of Auvergne.(38)

If Auvergne and those surrounding regions, controlled by the heirs and successors of Bernard Plantevelue, disintegrated into political disunity, what of the nearby kingdom of Provence-Burgundy? Here we find a very similar story. In 900 control over the Valley of the Rhone south of Lyon was officially in the hands of its king, Louis of Provence. Actual authority, however, seems to have been exercised by Hugh of Arles and his brother Boson, who, as far as we can tell, distributed honores and acted in the name of this blind monarch, particularly after his return from Italy.(39) [203] With Louis' death, however, they were unable to keep their authority, despite efforts to give their lands to relatives and to carve out careers for themselves in Italy.(40) In 938, Conrad became king of a new kingdom of Burgundy which included in it all Louis' holdings.(41) The Middle Rhone region south of Vienne fell into the hands of Count Geilin, who also dominated the Vivarais and Velay for a time,(42) while Provence went to a certain Boson, the father of Count William of Provence, who was to make his authority effective over this region and drive the Moslems from Fraxinetum in 972.(43)

As these changes took place, however, it seems clear that King Conrad's actual authority did not extend much beyond his capital of Vienne and its immediate vicinity. Nor is there much evidence that any of the kings or magnates of the Rhone Valley, like Count Hugh, Count Boson, or Count Geilin ever exercised much authority over Alpine regions east of the Rhone. These areas remained, as far as we can tell, disorganized and lawless, the prey of brigands and Moslem freebooters down to the end of our period. Only then do we find a group of new families in Savoy, Dauphiny, and Provence, aided by churchmen like Bishop Isarn of Grenoble, begin the task of restoring order on a local basis by the building of castles, the cultivation of abandoned soil, and a restoration of churches and abbeys.(44) But this colonization seems to have proceeded with little assistance given to such nobles by either the kings of Provence or their principal vassals.(45) As principalities, the kingdom of Provence and its successor the kingdom of Burgundy must be rated as failures.

[204] In some ways an examination of that vast extent of territory which the House of Toulouse controlled in 900 is even more interesting. This territory stretched from the Rhone to Périgord and Angoulême, and from the shores of the Mediterranean, the borders of Carcassonne, and the Pyrenees to Auvergne and the northern borders of the Limousin. It had few frontiers in any formal sense of the word, which was one reason for the friction between the Toulousain house and its neighbors in Carcassonne-Razès, Auvergne, Poitou, and to some extent in Gascony. As far as we can tell, early in the tenth century this whole region formed a rather complex family co-dominion which was controlled by two sons of Count Eudes, Raymond and Ermengaud.(46) Upon Raymond's death his son Raymond Pons represented the Toulousain branch of the family and Ermengaud the Rouergue branch.(47) Of the two, Raymond Pons, down to his death about 950, seems to have been the more powerful,(48) though a certain Count Hugh, a nephew, perhaps had some authority also.(49) After his death his widow Countess Garsinde seems to have been more powerful than Raymond Pons' heir, Count William Taillefer, judging from her will which dates from 972.(50) But these years saw an even greater family authority exercised by the Rouergue branch consisting of Count Raymond I of Rouergue,(51) his son Raymond II,(52) and the latter's brother Bishop [205] Hugh of Toulouse.(53) This Rouergue predominance seems to have continued down to the early years of the eleventh century.

What concerns us here, however, is not the complexities of authority and influence exercised within the House of Toulouse-Rouergue, but rather something else -- evidence of their steadily shrinking authority over Languedoc. We first see this clearly when we examine the Narbonne-Carcassonne region. Here in 918, 922, and 924(54) charters seem to show that the counts of Toulouse exercised a large measure of control over these regions. Yet in 933 Count Raymond Pons was forced by a court of leading lay and ecclesiastical magnates of the region to give up his efforts to enforce his rights over lands belonging to the abbey of Montolieu.(55) Thereafter direct control by the Toulousain family seems to have been largely confined to the abbey of Saint-Pons of Thomières(56) and after 949 even this authority seems to have been a limited one. In nearby Béziers we find the same story repeated. In the year 937, when Raymond Pons gave land to Saint-Nazaire of Béziers, he seems to have acted with considerable authority over this region.(57) After this date this ceases to be true, and down to 975 his house is not even mentioned in the local charters of this part of Septimania.(58)

Lest this be thought unusual, an examination of the Nîmes region reveals the same pattern of development. Early in the century both Count Raymond and his son Raymond Pons were the powerful overlords of Nimes and were mentioned as such in certain lawsuits which took place in [206] this area in 902, 915, and 928.(59) Then all is silence down to 961 and 965, when we find a record of certain gifts to the cathedral church of Notre-Dame of Nîmes by Countess Bertha, widow of Count Raymond I of Rouergue. In the charters referring to this gift of villas, Countess Bertha in 961, following what was the common practice of the period, forbade their alienation from church control, saying that if this was done they would revert to her son, Count Raymond II of Rouergue or "ipsam potestatem de Nemauso publice."(60) In 965 this curious phrase is explained by a repetition of the preceding phrase except that "ipsam potestatem publice" is identified. It is the viscount of Nîmes.(61) By 965, then, even the Toulouse-Rouergue family had to publicly recognize those who really controlled Nîmes -- not they but the hereditary viscounts of the city.

When we examine what happened to the power of the Toulousain family in the northern and western portions of their domains we find the same thing happening. We do find a reference to their authority in Rouergue in 910 and 919(62) and in the Albigeois in 926, 934, 935, and 943.(63) Then all is silence, and from local records it seems clear that these regions were controlled by viscontal families, like those of Albi, Rodez, and Carlat, just as similar families controlled Auvergne. Similarly in Quercy and the Limousin we have references to Toulousain overlordship in 930 and 932, as well as one in 972 which refers to an earlier period about 930 or 940.(64) Yet after this period, except for a court held by Raymond I of Rouergue in the Limousin in 960,(65) we find no other proof of effective overlordship. The viscounts of Cahors and the Limousin obviously now control these regions in their own right.

The same story seems to have been repeated toward the south and southwest of these domains. We are told that about 950 the count of Toulouse gave the county of Agen to the House of Angoulême-Périgord.(66) And at about this same time we learn of a certain Arnaud, who was probably related to the family of Carcassonne, who became the ruler of that [207] region which late in the century we know as the county of Commignes.(67) None of the few charters which we possess for this region between 950 and 975 even mention the Toulousain house.

The best proof, however, of the declining fortunes of the family of Toulouse-Rouergue and of their failure to establish a territorial state comes from a series of wills which date from late in the period -- one of Raymond I of Rouergue's dating from 961,(68) one of Countess Garsinde's from 972,(69) and a little later one left by Hugh, bishop of Tou1ouse.(70) Naturally, these wills put forward the best case possible for family and personal possessions, and they do prove to us that down to 975 this family was still a powerful one, controlling a number of castles, churches, and allods through all of Languedoc. But when we examine these wills in detail we find that actual possession of most of this property, as opposed to its ownership, is generally in the hands of magnates and churchmen of these regions, not the House of Toulouse.(71) By 975, then, not only had this family failed to create a principality in the Midi, but they had been forced to watch both their authority and much of their property slip away into the hands of others. After 975 for many decades it was the counts of Commignes, of Carcassonne-Razès, and of Périgord, and the viscounts of the Limousin, of Cahors, Rodez, Albi, Nîmes, Béziers, Lautrec, and Narbonne, who really controlled Languedoc. All that was retained by the House of Toulouse was control over some scattered abbeys, castles, and allods, and a tradition of overlordship, which in the next century could be used by Raymond of Saint-Gilles to create a new and different principality.

When we turn from Languedoc to Gascony again disintegration seems to have been the order of the day, though we lack detailed information concerning how it happened. In the ninth century Gascony seems to have been a vast rather politically amorphous region, consisting of a county of Fézensac, some areas under the control of the counts of Toulouse and a large area ruled by the hereditary dukes of Gascony.(72) By 975 this had [208] ceased to be the pattern. Instead, in eastern Gascony we find a whole series of independent counties and viscounties: Commignes, Bigorre, Astarac, and Fézensac, and in the western part Béarn, Lomargue, Tartas, Tursan, Marsan, and several others.(73) Probably the dukes of Gascony still exercised some control over western Gascony,(74) but eastern counties and vis-counties had become independent of ducal authority. So they were to continue down into the eleventh century.

In some ways, however, the Spanish March, consisting of Pallars, Ribagorça, and Catalonia, provides us with the most interesting case study of all, for here in the late ninth century Count Guifred had established the most workable and successful large principality of all as an heir to Carolingian authority. What happened then to his work during the first three quarters of the tenth century? As far as Pallars and Ribagorça are concerned the answer is a simple one -- a progressive and almost complete disintegration followed the death of its first Count Ramón. First Pallars and Ribagorça were divided into separate counties, and then subdivided as three and four co-counts divided authority still further and left it to their heirs.(75) Ampurias-Roussillon stayed relatively stable and unchanged throughout this period.(76)

When we consider the rest of Catalonia, though -- that part ruled by the heirs of Count Guifred and Count Miró -- we begin to notice some interesting things happening. In a certain sense throughout this period a large amount of Catalan unity was maintained by this family, and Guifred's direct heirs, who controlled the Barcelona -- Gerona-Ausona region, could claim some primacy over the rest by bearing the title of princeps, even after 975.(77) They could intervene in some cases in the affairs of abbeys which were located beyond regions which they ruled directly.(78) About 950, [209] however, despite the tradition of Barcelona's supremacy, we begin to notice a certain separation taking place between the counties of Cerdanya-Besal-Confluent-Berga under Miró's descendants and Barcelona-Gerona-Ausona-Urgell controlled by Count Borell and his brother Count Ermengol.(79) This situation, a little later, seems to have led to an outright hostility between Count Oliba Cabreta who controlled the former and Count Borell who controlled the latter, with Urgell, under Count Ermengol, following still a third path of its own.(80) By 975 the forces of disintegration were beginning to affect a Catalonia which had not known them earlier.

This disintegration which was affecting the ruling family's unity was made worse, it would seem, by a growing independence which more and more began to be assumed by viscounts, vicars, and other important landowners who controlled castles and broad lands along Catalonia's frontiers. In 928, for instance, we find Viscount Leopardo selling to his son for some 1,000 solidi a castle and its surrounding territory without bothering to secure the permission of one of the counts.(81) In 940 another important lord, Sendredo, freely exchanged the castle of Tarrassa for some other property which belonged to the bishop of Barcelona.(82) In 954 the viscounts of Cérdanya and Urgell, neither bothering to secure a count's permission, made an agreement covering their frontier castles of Queralt and Miralès.(83) Two years later another marcher lord, Hennego, freely bought from a certain Aigo a turre and much property nearby for 800 solidi,(84) and in 957 left to the abbey of San-Cugat land and the rights to justice he had over the territory of the castle of Arampuña.(85) Two years later Daniel, another frontier lord, also left a turre and land near the castle of Olerdula to San-Cugat without securing Count Borell's permission.(86) [110] In 963 Hennego bought from Count Miró the castle of Masquefa, near Barcelona's frontier, for some 1,000 solidi.(87) All this makes it clear that along the frontiers of Catalonia there existed a good deal of independence -- an independence that definitely limited the authority of the counts of Catalonia. So Catalonia, too, began to be affected by the same forces of disintegration which we find in the rest of the Midi, though down to 975 such forces were weaker here than elsewhere.

This examination of the leading families of the Midi and the Spanish March makes it dear that no principalities arose to take the place of Carolingian government. But it leaves unanswered the basic question as to why this was so. The first and most important reason we can give has to do with the operation of the family system in Southern France and Catalonia. As we noted earlier in all our regions, by 900, families had become the controlling element in government, as they took over what in Carolingian times had been honores. During the first generation or so family solidarity tended to preserve a certain unity in regions which a particular family controlled, witness Catalonia immediately following Count Guifred's demise;(88) or Languedoc down to the death of Raymond Pons and Ermengaud;(89) the Valley of the Rhone under the control of Count Hugh of Arles and his brother Boson;(90) or Angoulême-Périgord where, according to Ademar of Chabannes, Count Bernard and his brother, Count William Taillefer, "communem habuerunt totum honorem eorum ipse."(91) In some ways this family system continued to work effectively [211] even after 975, as in Cerdanya-Besalu, Carcassonne-Razès,(92) and Provence under the immediate heirs of Count Boson of Arles.(93)

But eventually the principle of divisio and the habit of allowing widows to continue to control their husbands' estates triumphed over family unity. Countess Garsinde, for instance, in 972 appears to have given to her own family in Narbonne, and to special friends and relatives, land that belonged to the House of Toulouse, and even threatened her nephew Hugh with disinheritance if he disputed her distribution. In this will she also seems to have disinherited Count William Taillefer of Toulouse, her son or stepson. The division of land in Gascony among the Aznar heirs and their relatives effectively dismembered this region into a whole series of counties and viscounties.(94) On a minor scale a similar divisio resulted in the county of Béziers being divided between two viscounts, one at Agde and one at Béziers,(95) and began to spawn a series of minuscule viscounties in the Limousin out of what originally was a viscounty of Limoges,(96) a county of La Marche,(97) and a county of Turenne.(98) It was later to break up [212] the county of Provence and that of Carcassonne-Razès. Familial control in which wives, sons, and daughters shared in a principality and in which no primogeniture existed, ended any possibility of organized states emerging in the Midi and the Spanish March during these years.

There was, however, still another force at work in this process of political fragmentation. I refer to the tendency to transform into allods land which was given to fideles by their lords. Some of this land, no doubt, was from the start given in this manner, but the rest was transformed in the course of this period. Such a tendency, of course, was nothing new. We have commented on it earlier in the distribution of land from the royal fisc by Carolingian rulers. It continued under their successors, the independent nobles of the Midi and Catalonia. Only this can explain why Count Acfred possessed so little property in Auvergne and neighboring regions when he made his will in 927, a will incidentally which left even this property to fideles instead of to his own family. By 930 there was no comital fisc left in Auvergne.(99)

Exactly the same thing happened in the Valley of the Rhone. In 902, for instance, we find King Louis of Provence giving land belonging to his royal fisc to his fidelis, Viscount Berilon, "jure pro prietario"(100) and doing the same with other family property in Rodez which he distributed to two other fideles, Teutbert and Bernard.(101) In 903 he went further and gave to all his nobles the right to freely dispose of, sell, or will their own property as they pleased.(102) Just before 910 he gave considerable property to Count Hugh of Arles(103) and followed it up with distribution of land from the fisc to his fidelis Gerard in 915,(104) to Inglebert in 923(105) and to the church of Vienne in 927.(106) Probably it was land which originally was part of the royal fisc which Hugh, now king of Italy, gave to his nephew in 937 -- some 700 mansi according to the charter(107) -- and from this same source came those five villas with which he endowed the abbey of Saint-Barnard de Romans in 928.(108) It seems probable that even before King [213] Conrad of Burgundy succeeded to Louis' position in the valley the greater part of the royal fisc had been dissipated, and with it royal control over this region. Yet King Conrad continued to act in a similar fashion with what little was left. In 943 we find him giving the abbey of Saint-Genesius to a certain Hermarus,(109) a turre to Saint-Victor of Marseille(110) and as late as 975 giving to a nobleman, Artald, two villas as a benefice, which, since it was a grant for two lifetimes, probably soon became the latter's allodial property.(111)

In regions controlled by the House of Toulouse we find the same tendency at work, though there can be little doubt that this family was more reluctant to alienate its property than the kings who controlled the Valley of the Rhone. Nevertheless our sources tell us of a certain Repertus in 923 who controlled an allod at Nîmes which had at one time been fiscal land.(112) A charter of 930 from the Limousin contains a statement by Viscount Ademar of Scalas that the land which it refers to, an allod, originally belonged to Count Raymond who gave it to the viscount's father.(113) Our sources also mention another allod, a villa, owned by Arnold of Commignes which the latter received from Count Hugh of the Toulousain house,(114) and a fief or feus in Quercy which a certain Raimulf in 972 says Count Hugh gave to his grandfather.(115) Still another charter dating from 942 concerns an honorem given to the abbey of Saint-Pons of Thomières by Viscount Aton of Albi which the latter states was a gift to him from Raymond Pons,(116) while the will of Raymond I of Rouergue in 961 mentions a fief given to Sánchez in Gascony by this count which is to be changed into the latter's allod.(117) Such dissipation of property, which ended in becoming allodial land in the hands of important local magnates, then, was a real feature of the failure of the House of Toulouse to create a strong principality in Languedoc during this period.

There is little need to burden these pages with the even more numerous examples of this tendency which could be cited from Catalonia and Pallars and Ribagorça, where a continuing policy of granting aprisiones meant the exhaustion of vacant land belonging to the comital fisc and its transformation [214] into allods which were owned by both large and small landowners. Suffice it to say that dozens of examples could be cited, particularly from regions close to the frontiers of the Spanish March.(118) Enough has already been said, though, to show how the tendency of land granted to fideles to become allodial was a feature of this period, and one which must bear a large share of the responsibility, along with the workings of the family system, for the failure to develop territorial states south of Poitou and Burgundy.

As attempts to create principalities failed and those which existed disintegrated into small, amorphous and often unstable counties and viscounties, what happened to the machinery of government which had survived the end of Carolingian authority? Here an examination of law and the court system is particularly pertinent, since it seems to have survived more intact than any other aspect of Carolingian government down to the first part of the tenth century. What happened to it now during this period of further governmental disintegration?

Our first impression seems to convince us that very little change took place during the tenth century in the surviving legal system of most of Southern France and Catalonia. Personality of the law continued in the Valley of the Rhone, in Aquitaine and in Septimania where we find references in a number of our documents to Roman, Gothic and Salic law and precedents.(119) Even more important, certain courts continued to be held in the Carolingian fashion and are presided over by the official representatives of secular authority, that is to say by kings, counts, viscounts, missi, and vicars. We find two such tribunals which were held in the Valley of the Rhone in 912 and 926(120) over which King Louis of Provence and Hugh of Arles respectively presided, and two others at Arles in 965 and 967-968(121) which met under the aegis of Count Boson of Provence and his son Marquis William. At Nîmes we find a record of similar tribunals which met in 902, 909, 915, and 928(122) under the legally [215] constituted authority of the count of Toulouse and his viscounts and missi Abbo and Frédélon. Church and lay officials jointly presided over similar courts which met in the Narbonne-Carcassonne region in 918, 933, and 955,(123) and there is also a record of a vicar's court which was held-in Razes in 958.(124) To the south in Catalonia we find the same sort of tribunals which were held by the counts of this region in 904 and 912.(125) Though our record in this respect is not as complete for the turbulent lands which made up Aquitaine, we do find in our documents a reference to a vicar's court which met in Rouergue in 934,(126) and one which in 960 was held in the Limousin and which was presided over by Raymond I of Rouergue.(127)

As far as we can tell, too, all these courts seem to follow the same general pattern as those found in the earlier Carolingian period. In each one the official or officials presiding over the court were assisted by minor officers, such as viscounts and vicars, and by judices and that group of leading landowners of the region where the court was held, who still went by the name of boni homines, rachimburgs, scabini, or in one case principes.(128) Despite the presence of some men called fideles(129) at some of these tribunals and even the mention of vassi dominici at a Provençal court,(130) it is clear that these courts were not feudal ones. They still represent the regional organization of earlier Carolingian times based upon a district, and the boni homines who attend do so because of their residence in a region and not because of a personal tie linking them to the count or viscount who was presiding. Formal courts, then, in this period still seem regional in essence rather than personal.(131)

There were other tribunals too which seem to have survived from the Carolingian period, ecclesiastical ones. Some of them seem to follow the usual pattern of such courts in the ninth century, which means that their jurisdiction lay in the privileges granted them in immunities from the [216] crown, one of the more important being a church or an abbey's right to maintain its own tribunal. For instance the important church council that met at Azulan in 902 acted in just such a fashion in forcing a certain cleric called Tetbald to give up land he was holding unjustly which belonged to the church of Sainte-Eulalia.(132) So did a tribunal of the abbey of Saint-Joan de les Abbadesses which in 912 forced a certain Tudesco to acknowledge Abbess Emma's ownership of land that he claimed.(133) Even that curious court presided over by Bishop Hucbert of Nîmes in 921, in which the churchmen present as boni homines are called vassi in our documents, seems essentially like a court of Carolingian times since it dealt with tithes unjustly appropriated by landowners who seem to have been dependent on the church of Nîmes -- and thus within the competence of this court.(134)

In this age of decaying governmental forms, however, certain other church tribunals seem somewhat more unusual, particularly in parts of Central France. An example of such a court is one held by Bishop Etienne II of Auvergne, who certainly had a court of his own as early as 945,(135) and also in 958.(136) This court, according to our sources, was one attended by churchmen and important laymen or principes of Auvergne, and forced a certain Calistus to give back land which he had usurped from Amblard, a canon of Clermont. Three years later this same prelate held still another court, with his fideles in attendance, at the church of Saint-Martin and forced his vassal Gerard to return a mansus which he had unjustly taken from the abbey of Sauxillanges.(137) What makes these tribunals unusual is the fact that Bishop Etienne, who presided, seems to be acting not as a churchman who could hold such courts by virtue of a royal immunity, but as the embodiment of public authority in Auvergne replacing the secular government of the period. Furthermore if the court of 961 seems an unusually feudal one for this period, that of 958 does not fit such a category, since its presiding officer, in this case the bishop, is still assisted by neighboring boni homines, called principes here, who, as in Carolingian times, represent the force of local public opinion in giving their backing to the verdict reached by the court.

As governmental authority and forms decayed still further in most of [217] our regions, except perhaps Catalonia and Provence, the destiny of the Midi was to see more such unusual tribunals like that held by Bishop Etienne in 958. Such courts were not feudal, but rather seem to have been informal gatherings of leading landowners and churchmen of a particular region, who met together to decide matters which earlier in Carolingian times had come before a mallus publicus. Sometimes such assemblies or gatherings are called courts, sometimes they are not. But irrespective of their title, they performed one of the major functions of public tribunals, that of forcing an individual landowner to justify his ownership or occupancy of a certain piece of property, and, if he could not furnish satisfactory proof, to formally return it to its rightful claimant in an action or ceremony known as guirpitio or werpio, which frequently included a written carta evacuonis.(138) One might well think of such tribunals as courts without presiding officers or a public authority in the Carolingian sense of the word.

No study of this informal guirpitio system, which we begin to find in the Midi during this period,(139) has ever been made, so a few observations concerning it seem very much in order. It seems dear that in part it was derived from the Carolingian court system, which also provided a very similar method of giving up land illegally held before a mallus publicus which was attended by boni homines and presided over by an official, and the formal writing of a carta evacuonis at the conclusion of the case.(140) One might then simply consider the guirpitio system a Carolingian court meeting informally without a presiding officer; a system made necessary by the breakdown of government.

Nevertheless it seems wise also to note another possible origin of guirpitios which we find in the Midi in increasing numbers; a private origin. All through the ninth and tenth centuries in the Midi landowners made it a practice to gather together groups of important witnesses to attest the validity of their private charters and wills, particularly those in which important property was at stake. This was particularly true of charters giving land to church establishments. The formula seems clear: an individual and members of his or her immediate family swear to [218] their gift and they ask others present to affirm or to witness their oath. The number and importance of these witnesses seem to be in direct proportion to the importance of the land in question. They are referred to in such charters as friends, neighbors, or noblemen or boni homines -- and their names are almost always given.(141) The validity of a private legal act involving transfer of property rights in this period seems more and more to be dependent in practice upon a public opinion represented by the presence and affirmation of members of one's family, one's friends, and one's important neighbors.

What could be more natural, then, as the family system superseded government in the Midi and public courts broke down, than to gather together the same sort of boni homines, nobiles viri, and neighbors to decide matters concerning disputed land, since such assemblies were already being used in a private capacity to attest the grants of such lands by individuals or their wills in the charters of the period. Certain remnants of the public court system in the Midi and these private procedures then coalesced in the tenth century to create the guirpitio system. Groups of boni homines deriving their authority from their habit of attending the public courts and their use as witnesses to private agreements and family charters began to informally assume the right to judge disputes concerning land, where no other clear jurisdiction existed. Since the authority of these boni homines and others who met informally in such assemblies rested essentially upon public opinion, they strove to arbitrate disputes rather than use force, but this does not mean they were without power. They had a great deal, and continued to use it in this period and later on.

We find in the Midi, as public tribunals disappear, an increasing number of such guirpitios used to settle disputes over land. Perhaps those witnesses who attested the return of land to the abbey of Tulle in 922 by Ademar, count of Scalas, formed such an informal tribunal,(142) as did those who, a charter shows us, were present sometime between 923 and 935 when Gotfred returned the church of Stanquillières to the abbey of [219] Beaulieu in this same part of the Limousin.(143) It seems probable too that such an informal group handled the trial by combat in 946 which decided a dispute over a church between Tulle and Viscount Odalric of Saint-Cyrico.(144) Whatever we may decide concerning these earlier guirpitios, real or potential, judging from our charters there can be little doubt concerning later ones. Thus we can be sure that such an informal court procedure was used in the Limousin in examples which date from 951, 959, and 969-970.(145) We find it in use in the Périgord area in 970,(146) in Rouergue in 958 and 964(147) and in Auvergne about 950 and in 961 and 962.(148) In Septimania we find examples of it in charters dating from 939, 972,(149) and shortly after 975, though only later does it appear to have taken firm root in the Narbonnaise or Catalonia. It is found in the Valley of the Rhone in 926, 932, 968, and 970.(150) After 975 it became the normal system of handling disputes over property in most regions north of Catalonia.

Thus as government disappeared in the Midi and Catalonia the older court system decayed also. As it did more and more disputes which had been carried before public tribunals had to be handled in a new way. This new way was the use of gatherings of an informal nature attended by boni homines, nobiles viri, or principes. Such gatherings were not feudal, though feudal elements sometimes attended them. They were regional or local reflections of public opinion as represented by the leaders or powerful men of a particular district. And their verdicts bad the force of law through their very importance. The fate of the Carolingian legal system in the Midi, then, was not to disappear into feudalized courts, but to disappear and to be replaced by assemblies of important landholders and church magnates who rendered such rough justice as was possible under the circumstances and who kept disorders to a minimum at a time when there was little judicial authority upon which society could depend.

Notes for Chapter 11

1. See J. Dhondt, Étude sur la naissance des principalités territoriales en France, for an expression of this point of view. In this work Dhondt makes the assumption, without proof, that the decline of royal power automatically made for principalities.

2. See Chapter II, Section III, for a description of how viscounts came to develop in the Midi and Catalonia.

3. William is called "count" and "marquis" in a charter which dates from 898 to 911 (Cart. de Brioude, no. 64). Another charter dating from 907 refers to him as "duke" (ibid., no. 228). Again a charter of 917 which established the abbey of Sauxillanges calls him "count" and "marquis" (Cart. de Sauxillanges, no. 146). His heir, Count William II, seems to bear no title other than "count," however (see Cart. de Brioude, nos. 66, 318, 324). The next ruler of this line, Acfred, is only called "duke" once in a charter dating from 928 (Cart. de Sauxillanges, no. 13). In other charters of 922, 923, 926, and 927 he is simply called "count" (Cart. de Brioude, nos. 15, 30, 167, 315, and Grand Cart. de Brioude, no. CCCCXXXIII). Charters from Auvergne from late in this period dating from 954 to 986 mention Guido or Guy (of the Clermont family) as "count" (Cart. de Sauxillanges, no. 363). Note this latter is during that period when some historians claim Duke William Caput Stupe of Aquitaine controlled Auvergne.

4. A charter of 930 from the Limousin calls Ebles Manzur "count" (Cart. de Tulle, no. 14), while in 942 another refers to his son William as "count" and "marchio" (Cart. de Saint-Sernin, no. 289). In still another given to Saint-Jean d'Angély in 945 he is called "duke" (Cart. de Saint-Jean d'Angély, no. 242). A later charter dating from 971 from this same region also refers to his son William Iron Arm as "duke" (ibid., no. 192). On members of this house continuing to bear the title of "duke" after 973 see F. Lot, Fidèles ou vassaux?, p. 67.

5. See the use of this title in charters dating from 902 and 903 in Cart. de Vienne, nos. 11, 12, 13, and in 923 in Chartes de Cluny, I, no. 237.

6. See charters of 910 (Cart. de Vienne, no. 14); of 917 (Cart. de Saint-Barnard de Romans, no. 9); of 920 (Cart. de Saint-André-le-bas, no. 124); and of 928 (Cart. de Saint-Barnard de Romans, no. 11).

7. Cart. de Saint-Barnard de Romans, no. 11.

8. See references to Count Geilin as "nobilissimus vir et potens" and "princeps noster" in Hist. Gén. de Lang., V. no; 110 and as "marquis" in a charter of 937 (ibid., no. 20).

9. He is called "dux" in Raoul Glaber, Historiarum libri V. ed. R. Prou, I, IV, 9, p. 12.

10. Flodoard, Annales, ed. P. Lauer, anno 932.

11. In 933 a charter calls him "count" and "marchio" (Hist. Gén. de Lang., V. no. 57). A charter issued by King Louis IV in 939 again refers to him as "marchio" (ibid., no. 59). Still another local charter dating from 924 calls him "Narbonensium principem" and "marquis" (ibid., no. 50).

12. Cart. de Brioude, no. 338, dated 941. This charter was issued by King Louis IV.

13. See charters dating from 937 and 939 which Raymond himself issued (Hist. Gén. de Lang., V, nos. 67, 69, 74).

14. Catalunya Carolingia, II, 210, 394.

15. A charter dating from 906 calls the count of Barcelona "count and marchio" (J. Villanueva, Viaje literario a las Iglesias de España, XI, appendix 11). Another of 908 calls him "princeps et marchio" (Hist. Gén. de Lang., V. no. 35). Still a third dating from 904 calls him "count" and "marchio" (Cart. de San Cugat, no. 2). See the use of the term "princeps" for Count Borell in charters dating from 947 and 974 in Marca hispanica, appendix 84, 118. See same in a letter of Gerbert dated 985 in J. Havet, Lettres de Gerbert, no. 45, p. 296. He is also called "dux Gothiae" in 972 (Marca hispanica, ap. 112, 113).

16. On use of the title "marchio" by the counts of Cerdanya-Besalu see R. de Abadal i de Vinyals, Els Primers Comtes Catalans, pp. 294-296.

17. See a number of charters in Abadal, Els Comtats de Pallars i Ribagorça, for use of this term in this region by members of the comital house.

18. See C. Samarin, "Les institutions en Gascogne," in R. Fawtier and F. Lot, Histoire des institutions français au Moyen Age, I, 186-190.

19. Judging from these examples, there seems little doubt that the title "marchio" no longer meant a count holding a frontier region, as had been the case earlier. Now it meant a very important nobleman.

20. The first use of "princeps" in late Carolingian times seems to come from a charter dating from 878 which refers to Bernard, "princeps Gothorum" (undoubtedly Bernard of Gothia), and a "princeps Urso" (perhaps the count of Provence) (Hist. Gén. de. Lang., V, no. 1). This charter is of local not royal provenance, however.

21. On the survival of Carolingian elements in Poitou which remained strong in contrast to the weakness of these elements in the rest of the Midi see M. Garaud, "Les circonscriptions administratives du comté de Poitiers et les auxiliaires du comté au Xe siècle," in Le Moyen Age, LIX (1953), 11-61.

22. See Cart. de Tulle, no. 15.

23. See Ademar de Chabannes, III, 24, pp. 146-147, on the building of castles at Limoges by Bishop Ebles.

24. On the control of the bishopric of See Chartes de Cluny, I, no. 825, on the nobility of Auvergne doing homage to William, count of Poitou, in 955. On Ebles' earlier supposed control of Auvergne see Ademar de Chabannes, III, 23, p. 143. Limoges by its viscontal family see G. Tenant de la Tour, L'homme et la terre de Charlemagne a Saint Louis, pp. 260-280.

25. On the appearance of these counts in the tenth century see G. Thomas, Les comtes de la Marche de la maison de Charroux (Xe siècle-1177).

26. See Chartes de Cluny, I, no. 825, on the nobility of Auvergne doing homage to William, count of Poitou, in 955. On Ebles' earlier supposed control of Auvergne see Ademar de Chabannes, III, 23, p. 143.

27. Since a few years later a charter, which does not mention the Poitevin house, calls Guy of Clermont "princeps Arvenorum," it seems that their control over this region was minimal (Cart. de Sauxillanges, no. 363). Since there is also no mention of this house in charters of Brioude or Sauxillanges during these years, this would seem to indicate that they had little effective control over Auvergne.

28. See will of Raymond I of Rouergue which mentions land in Auvergne belonging to the Toulousain House (Hist. Gén. de Lang., V, no. 111).

29. See the charters of the Cart. de Brioude and the Cart. de Sauxillanges which reveal the power of this family in Auvergne throughout this period.

30. See the charter of Saint-Jean d'Angély of 945 in which the count of Poitou is called a duke (Cart. de Saint-Jean d'Angély, no. 242).

31. On the power of Duke William the Pious in the Lyonnais see E. Fournial, "Recherches sur les comtes de Lyon aux IXe et Xe siècles," in Le Moyen Age, LVIII (1952). See references to his power in Auvergne in Cart. de Brioude, nos. 64, 228, and Cart de Sauxillanges, no. 146.

32. On William II's power in Auvergne see Cart. de Brioude, nos. 66, 318, 324. On his control of the Lyonnais see Fournial, "Recherches sur les comtes de Lyon," and reference to him as count of Lyon in 921 in Cart. de Savigny, no. 12. For Velay see Hist. Gén. de Lang., V, no. 49.

33. See reference to Acfred's land which was located in Auvergne, Velay, and Gevaudun according to his will of 927 (Grand Cart. de Brioude, no. CCCCXXXIII). In a charter dating from 928 he is referred to as "duke" (Cart. de Sauxillanges, no. 13).

34. Grand Cart. de Brioude, no. CCCCXXXIII.

35. Note that Raymond was called "princeps Aquitanorum" when he was present with the lay and secular "procures" of the province when Louis IV gave a charter to the abbey of Chanteuges in 941 (Cart. de Brioude, no. 338).

36. Chartes de Cluny, I, no. 825. Yet a few years later Aldegardis of the Clermont family is called "countess" (Cart. de Sauxillanges, no, 188).

37. See the excellent descriptions of Auvergne during this period in "Etude historique sur la vicomté de Carlat" in Documents historiques relatifs a la vicomté de Carlat, ed. G. Saigne, II, lix-lxxvii, and Introduction to the Cartulaire du Peuré de Saint-Flour, ed. M. Boudet, pp. lvii-clxvi.

38. See Fournial, "Recherches sur les comtes de Lyon," on the appearance of this new family of counts in the Lyonnais.

39. See R. Poupardin, Le Royaume de Bourgogne (888-1038), pp. 1-138, on the career of Louis of Provence, and ibid., pp. 138-160 on Count or Duke Hugh. See also charters dating from 908, 912, 917, and 928 in Cart. de Saint-Barnard de Romans, nos. 6, 9, 11, and Cart. de Grenoble, no. 24. There is an additional reference to Hugh in a charter dating from 920 in Cart. de Saint-André-le-bas, no. 124.

40. In 937 Hugh attempted to give land belonging to the royal fisc and consisting of 700 mansi to a nephew (Cart. de Vienne, no. 22, pp. 232-233). This did not prevent Conrad from becoming king of Provence-Burgundy the next year.

41. On the circumstances surrounding Conrad's assumption of authority in this region see Poupardin, Le Royaume de Bourgogne.

42. A charter of 937 from LePuy calls Geilin a "marquis" (Hist. Gén. de Lang., V. no. 70), and another dating from 961 calls him "princeps" and "nobilissimus vir et potens" (Cart. de Saint-Chaffre, no. 345).

43. See reference to Count Boson in 958 in Chartes de Cluny, II, no. 1052, and again in 965 in Cart. de Saint-Victor, no. 29. On Count William's expulsion of the Moslems see Raoul Glaber, Historiarum Libri V. I, IV, 9, p. 12.

44. A posterior charter of 1100 refers to the work of Bishop Isarn (950-974) in colonizing Alpine regions (Cart. de Grenoble, no. 16). In a more contemporary document of 938 there is a reference to land which has been devastated by the pagans (perhaps the Hungarians), which is given to Canon Ubold by Archbishop Sobbo of Vienne "in prestaria" (Cart. de Vienne, no. 23).

45. Neither King Conrad nor the principal lay nobles of the Middle Rhone region seem to have played any role in the colonizing activities mentioned in note 44. In contrast to the churchmen who were active here, in Provence it was Count William and the secular nobility who took the lead.

46. See reference to Count Raymond in 919 in Cart. de Conques, no. 322, and to "Ermengaudo principe" in 934 in Hist. Gén. de Lang., V, no. 54.

47. In 932 Raymond Pons and Ermengaud, called "principes Gothiae," did homage to King Raoul (Flodoard, Annales, anno 932). A charter of 922 refers also to "comites nostri Ermengaudi et Raymundo" Hist, Gén. de Lang., V, no. 47.

48. Hist. Gén. de Lang., V, nos. 50, 55, 59, 69, 74, and Cart. de Brioude, no. 338, all refer to Raymond as "marquis," "duke," or "princeps."

49. See mention of Hugh in 934 in Hist. Gén. de Lang., V, no. 54. In 940 he is referred to as a witness to a charter from the Narbonne region (ibid., no. 74), and again in 949 (ibid., no. 89). It is probably he who is referred to in Countess Garsinde's will of 972 (ibid., no. 126).

50. Ibid., no. 126.

51. Raymond I's will which dates from 961 shows he exercised considerable authority over the whole Midi (ibid., no. 111). See evidence of his influence in the Limousin in 960 as evidenced in his presiding over a court in 960 according to a local charter (Cart. de Beaulieu, no. 47).

52. For references to Raymond II of Rouergue in 961 and 965 see Cart. de Nîmes, nos. 61, 66. For another reference in 972 see Hist. Gén. de Lang., V, no. 125.

53. For an early reference to Bishop Hugh of Toulouse, whose will we possess from later in the century, see a charter of 949 in Hist. Gén. de Lang., V, no. 89.

54. In 918 Viscount Bernard, missus of Count Raymond, settled a dispute between the abbey of Montolieu and the vicar of the castle of Alzonne (Cart. de Carcassonne, I, 188-189). A letter of 922 from the archbishop of Narbonne speaks of "comites nostri" Ermengaud and Raymond (Hist. Gén. de Lang., V, no. 47). In 924 Eudes, viscount of Narbonne gave to the abbey of Montolieu an allod which his father received from Charles the Bald. He does so only with the consent of Raymond, count of Toulouse who is called Pons, "Narbonensium principem" (ibid., no. 50).

55. Ibid., no. 57.

56. See ibid., no. 74, on Pons' founding of Saint-Pons de Thomières. For his widow's continued control of this abbey as late as 969 see ibid., no. 117.

57. Ibid., no. 68.

58. See charters of 967 and 969 which mention only the viscounts of Béziers and not the Toulousain House in Cart. de Béziers, no. 30, and Hist. Gén. de Lang., V, no. 118. The second charter is the will of Viscount Reginald.

59. Cart. de Nîmes, nos. 9 (902), 16 (915), 32 (928).

60. Ibid., no. 61.

61. Ibid., no. 66.

62. Cart. de Conques, no. 322.

63. Hist. Gén. de Lang., II, no. 207 (926); V, nos. 59 (934), 63 (935), 78 (943).

64. Cart. de Tulle, no. 14 (930); Cart. de Beaulieu, no. 48 (932); Hist. Gén. de Lang., V, no. 122 (972).

65. Cart. de Beaulieu, no. 47.

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

67. See C. Higounet, Le Comté de Commignes de ses origines à son annexation à la Couronne, I, 22-27, and P. Ourliac, "L'origine des comtes de Commignes," in Recueil de travaux offerts à M. C. Brunel, II.

68. Hist. Gén. de Lang., V, no. 111.

69. Ibid., no. 126.

70. Ibid., no. 183.

71. See references to wills in notes 69 and 70 above.

72. Amalvinus is called "dux Bordelensis" in 906 (F. Lot, Etudes sur le règne de Hugh Capet, p. 378 n), and in 904 a certain García Sánchez is called "comtes et marchio in limitibus oceani" Gallia Christiana, I, inst., col. 170. These attest to the vagueness of any over-all ducal control of Gascony in the early years of the tenth century. See also Samarin, "Les institutions en Gascogne," I, 186-189.

73. Samarin, "Les institutions en Gascogne," I, 187-190, and Higounet, Le Comte de Commignes, I, 21-36.

74. Samarin, "Les institutions en Gascogne," I, 186-187.

75. See Abadal, Introduction to Els Comtats de Pallars i Ribagorça on this fragmentation of authority.

76. R. de Abadal i de Vinyals, Els Primers Comtes Catalans, pp. 297-298.

77. "Count Borell of Barcelona is called "Hispanica duce, marchisco, comtes" in 988 in Cart. de San Cugat, no. 217, and in 992 in Catalunya Carolingia, II, 72.

78. See account of the election of the bishop of Gerona in 908 in the presence of "princeps marchio" Guifred II (Hist. Gén. de Lang., V, no. 38). On the court in Ausona jointly presided over by Counts Miró of Cerdanya and Sunyer of Barcelona in 913 see El Archivo Condal de Barcelona, no. 38. Or note Count Borell's role in 949 in choosing the abbess of Saint-Joan in ibid., no. 128, or the intervention of Counts Borell and Miró of Barcelona in the affairs of the cathedral of Vich in 957, in ibid., no. 138.

79. On the growth of the authority of the Cerdanya-Besalu branch of the Catalan comital house see P. Ponsch, "Le Confluent et ses comtes du IXe au XIIe siècle," in Etudes Roussillonnaises, I (1951), 190-309.

80. Ibid., pp. 270-282.

81. El Archivo Condal de Barcelona, no. 91.

82. El Archivo Catedral de Barcelona, no. 20.

83. Hist, Gén. de Lang., II, no. 209.

84. Cart. de San Cugat, no. 45.

85. Ibid., no. 49.

86. Ibid., no. 56.

87. Ibid., no. 65.

88. Note the cooperation in 904 between Count Miró of Cerdanya-Besalu and his niece, Abbess Emma, the daughter of Count Guifred, in a matter concerning land which belonged to the nunnery of Saint-Joan (El Archivo Condal de Barcelona, no. 16). Note similar close cooperation between Count Miró and Count Sunyer of Barcelona in a court case in 913, previously mentioned (ibid., no. 38). In the next generation this remains the pattern also as seen in the cooperation between Count Borell of Barcelona and Count Sunifred of Urgell in 949 (ibid., no. 128).

89. A charter of 922, which speaks of "comites nostri Ermengaudius et Raymundus," shows this close relationship (Hist. Gén. de Lang., V, no. 47). So does Flodoard's account of how Raymond Pons and Ermengaud as "principes Gothiae" came together to do homage to King Raoul for their lands in 932 (Flodoard, Annales, anno 932).

90. For an example of such family cooperation see the gift of a villa and many mansi which Hugh of Arles made to his nephew and heirs in 937 (Cart. de Vienne, no. 22).

91. Ademar de Chabannes, III, 23, p. 145.

92. For the Cerdanya-Besalu comital family see the charter of 964 in which Sunifred, count of Cerdanya, Oliba Cabreta, count of Besalu and Miró -- all brothers -- join in giving land to the nunnery of Saint-Joan (El Archivo Condal de Barcelona, no. 162). See also action in concert taken by Countess Arsinde of Carcassonne and her sons in 957 and 959. Hist. Gén. de Lang., V, nos. 106, 109.

93. See how Count Boson and his sons Counts Roubaud and William jointly presided over a court at Arles in 965 (Cart. de Saint-Victor, no. 29), and the close cooperation between Counts William and Roubaud revealed by another charter of 970 (ibid., no. 598).

94. For a typical divisio see a charter which shows how two sons divide family land in Uzès in 946 (Charres de Cluny, I, no. 693). Or the divisio set up by the viscount and viscountess of Narbonne in 966 for their sons (Hist. Gén. de Lang., V, no. 115). Or another shown in a charter from the Limousin dating from 913 (Cart. de Beaulieu, no. 89).

95. See reference to Viscounts Eudes and Teudo of Béziers and Agde, who were brothers and heirs of Bishop Reginald of Béziers in 933 (Hist. Gén. de Lang., V, no. 58).

96. See also a charter of 930 mentioning Viscounts Gausbert and Ademar, who were perhaps brothers (Cart. de Tulle, no. 346). They are mentioned in a charter of the same year as well from Beaulieu (Cart. de Beaulieu, no. 49).

97. See reference to Viscounts Ademar, Boson (of la Marche?), Odolrico, Gausbert, Boson, and John, all of this region, in a charter of 931 (Cart. de Tulle, no. 641).

98. The family of Count Robert of Turenne or Cahors is mentioned in a charter of 932 (Cart. de Beaulieu, no. 108). Viscounts who may be of the same family are referred to in another charter which dates from 946 (Cart. de Tulle, no. 487).

99. See the distribution of the comital fisc of Auvergne and Gevaudun in 927 in Grand Cart. de Brioude, no. CCCCXXXIII.

100. Cart. de Vienne, no. 11.

101. Hist. Gén. de Lang., V, no. 29.

102. Cart. de Vienne, no. 12.

103. Ibid., no. 14.

104. Ibid., no. 16.

105. Chartes de Cluny, I, no. 237.

106. Cart. de Vienne, no. 21.

107. Ibid., no. 22.

108. Cart. de Saint-Barnard de Romans, no. 11.

109. Cart. de Vienne, no. 24.

110. Cart. de Saint-Victor, no. 1041.

111. Cart. de Vienne, no. 239.

112. Hist. Gén. de Lang., V, no. 89.

113. Cart. de Nîmes, no. 22.

114. Cart. de Tulle, no. 14.

115. Hist. Gén. de Lang., V, no. 122.

116. Ibid., no. 77.

117. Ibid., no. 111.

118. See Cart. de San Cugat, nos. 5 and 206 (912), 7 (914), 9 (917), 20 (941), 47 (956), 69 (964), 79 (965), 86 (966), 98 and 102 (972); El Archivo Condal de Barcelona, nos. 58 (918), 60 (919), 81 (975), 84 (927), 110 (937, 942), 124 (948), 151 (961), 157 and 159 (962), 160 (963); Catalunya Carolingia, II, 379-380 (922); F. Miguel Rossell, Liber Feudorum Maior, no. 395 (939) (Hereafter cited as Liber Feodorum).

119. For example see the court held in 918 at Alzonne (Cart. de Carcassonne, I, 188-189), or that held at Arles in 967-968 (Cart. de Saint-Victor, no. 290).

120. Cart. de Grenoble, no. 24 (912); Chartes de Cluny, I, 256 (926).

121. Cart. de Saint-Victor, nos. 29 (965), 290, (967-968).

122. Cart. de Nîmes, no.9 (902), 12 (909), 16 (915), 32 (928).

123. Cart. de Carcassonne, I, 188-189 (918); Hist. Gén. de Lang., V, no. 57 (933), 98(955).

124. Ibid., no. 104.

125. El Archivo Condal de Barcelona, nos. 16 (904), 35 (912).

126. Cart. de Conques, no. 155.

127. Cart. de Beaulieu, no. 47.

128. The term "principes" is used for some of those who attended the court held at Arles in 965 (Cart. de Saint-Victor, no. 29).

129. See record of court held in Dauphiny in 912 in Cart. de Grenoble, no. 24.

130. This is in that court held at Arles in 967-968 (Cart. de Saint-Victor, no. 29).

131. See comments on a similar situation in Burgundy during this same period in F. Estey, "The Fideles in the County of Macon" in Speculum, XXV (1955).

132. Hist. Gén. de Lang., V, no. 28.

133. El Archivo Condal de Barcelona, no. 35.

134. Cart. de Nîmes, no. 21.

135. Grand Cart. de Brioude, no. CCCXXXV.

136. "Archives du Puy du Dome" (unprinted), XVIII, soc. A, cat. 4.

137. Cart. de Sauxillanges, no. 86.

138. We find a series of such cartae evacuonis in the documents referred to in most of the notes found here from no. 145 to no. 150.

139. Ganshof defines a guirpitio as a ceremony whereby a vassal gives land back through a judicial act (F. Ganshof, Qu'est-ce que la féodalité? p. 167). Such a definition seems out of place for the Midi in this period, where anyone, vassal or not, can act in this fashion before a formal or informal tribunal.

140. See record of court cases found in Section III dating from the late ninth century for examples of this procedure.

141. For some examples of this see the charter of 930 from the Limousin which tells how Gausbert, viscount of Tulle, and his fellow viscount, Ademar, gave the abbey of Tulle a villa. The charter says that they do so in the presence of "boni homines" (Cart. de Tulle, no. 396). Or see the charter recording a gift of land to Saint-Augustin of Limoges by Bishop Turpin between 936 and 942 which says he does so "consentientibus nostris consanguineis seu optimatibus Limovicensi pago degredientibus" (Cart. de Saint-Etienne de Limoges, no. 15). Or see the charter which reports a gift to Cluny by Archbishop Gerard of Aix in 948, which says this is an "actum apud Sanctum Saturninum publice" (Chartes de Cluny, I, 724).

142. Cart. de Tulle, no. 12.

143. Cart. de Beaulieu, no. 56.

144. Cart. de Tulle, no. 599.

145. Cart. d'Uzerche, no. 120 (951); Cart. de Saint-Etienne de Limoges, no. 157 (959); Cart. de Tulle, no. 345 (969-970).

146. Cartulaire du prieuré conventual de Saint-Pierre de la Réolle [Gironde], ed. C. Grellet-Balguerie, no. 171, in Archives Historiques de la Gironde, V (1864), 171-172.

147. Cart. de Conques, nos. 293 (958), 193 (964).

148. Cart. de Sauxillanges, nos. 58 (950), 86 (961); Cart de Brioude, no. 336 (962).

149. Cart. de Gellone, no. 420 (939); Hist. Gén. de Lang., V. no. 121 (972).

150. Chartes de Cluny, I, no. 272 (926); Cart. de Saint-Barnard de Romans, no. 13 (932); Cart. de Savigny, nos. 116 (968), 158 (970).