PART IV
16
In one of the few non-diplomatic letters of Charlemagne (768-814) that has survived, the king wrote to his wife Queen Fastrada in 791. Charles relates that his son, Fastrada’s stepson Pippin king of Italy (781-810), has told him of a victory against the Avars of what is now Hungary, and lists the bishops, dukes, counts and vassals who performed particularly well in the war. (The letter omits their names, unfortunately; it only survives as a model for future writers.) The text then lists the religious litanies that Charlemagne and his court performed for three days, probably immediately after the news of the victory, including a prohibition on eating meat or drinking wine, which however people could buy out of with a graduated payment according to wealth. Charles asks Fastrada to take advice about performing similar litanies, and ends with an injunction to send him more regular communications.
The tone of most of this text is hardly intimate; it reads like a ruler communicating with a high-ranking subordinate, which a queen indeed was. There is no reason to think that it tells us much about the personal relationship between the couple. But in its mixture of military action and religious ritual it reflects what else we know of the tone of early Carolingian politics. It also shows that Charlemagne, even when not actively campaigning (he was probably forty-three in 791, fairly old for campaigns, though he did lead armies for another decade and more), received and expected up-to-date and detailed information from his generals: this information-exchange was a regular part of the political structures of the Carolingian century. The Merovingians had such information, but, as far as we can see, less systematically; it is also significant that this letter has survived when equivalent Merovingian letters have not. It has survived by chance, but in the context of a vast increase in surviving information about the political process in Francia, which reaches its height in the 830s-840s. It is also unlikely that the Merovingians articulated politics through as much penitential ritual as this. Charlemagne was not unusually pious (he was rather earthy, and loved jokes, songs, sex, hunting and swimming, and roast meat - less so drinking, it is claimed), but he introduced an ecclesiastical and moralizing edge to political practice which lasted throughout the Carolingian century and beyond, and which had many ramifications, as we shall see in this chapter and the next.
When Charles Martel (717-41) took over the office of maior of the Frankish kingdoms by force in the civil war of 715-19 (see above, Chapter 5), he re-established the practice of annual summer campaigning that had been intermittent at best for over seventy years. Between 720 and 804 there were only, probably, eight years without a campaign, and in some years there were two or three. Charles fought on all his borders, reabsorbing Provence and blocking Arab advances from Spain as he did so, taking over Frisia, and re-establishing Frankish hegemony in Alsace and Aquitaine. Most important, however, was the total authority he established in the Frankish heartland, thanks to this military aggregation, and to its success - Charles never lost a war. The Merovingian kings were only puppets by now, and the lay aristocracy and the episcopate both followed Charles; he overthrew any potential rivals without qualms or (apparently) difficulties. This continued under his sons Pippin III (741-68) and Carloman I (741-7) - they divided the mayor-ship just as the Merovingians had divided the kingship, until Carloman resigned his office, apparently willingly, and went to Rome, becoming a monk at the monastery of Monte Cassino. So did the annual campaigns, which included the subjection of Alemannia in the bloody battle of Canstatt in 746, extended to Italy in 754-6, and continued with the full reconquest of Aquitaine in a sequence of invasions in 759-69.
In his last years, after 737, Charles Martel ruled without a king. Facing revolts, Pippin and Carloman re-established one, Childeric III, in 743. Nonetheless, after Carloman retired, in the context of disturbances caused by family rivals, Pippin wrote to Pope Zacharias (741-52), to ask (in the words of the official Royal Frankish Annals, written some forty years later) ‘whether it was good or not that the kings in Francia at that time had no royal power’. Zacharias correctly replied ‘that it was better to call him king who had the royal power than the one who did not’, and Pippin took the throne in 751, the first Carolingian king. Childeric was tonsured - that is, had his Merovingian royal hair removed - and imprisoned in a monastery. (The Carolingians henceforth wore short hair and moustaches.) Later Carolingian sources of course depict this as a straightforward succession, buttressed by concord and ceremonial, including the agreement of the Frankish magnates and a formal anointing by Boniface archbishop of Mainz. Pippin was indeed the first Frankish king to be anointed; although this followed Visigothic practice in the late seventh century (and also the traditions of the Old Testament), the innovation clearly shows the need to make the Carolingians special, through a new set of ecclesiastical rituals. But in reality this was a coup, and it presented immediate problems of royal legitimacy. Pippin was able to reinforce the rituals of 751 when the new pope Stephen II (752-7) came north to the Seine valley in 753-4, the first time a pope had ever travelled north of the Alps, to ask for help against the Lombards; Stephen re-anointed him king, and Pippin duly invaded Italy, twice. The fact is that king and pope needed each other, the pope to gain protection against attack, the king to gain legitimate authority; for the Carolingians, although the strongest aristocratic family in Francia by far since the 680s, were not royal until two successive popes - importantly, an external, non-Frankish, moral power - said they were. The two processes went together. Pippin and Carloman were already more concerned than Charles Martel had been with church reform, and called at least four church councils in 742-7, the first since the 670s; this intensified after 751, under the aegis of Chrodegang bishop of Metz (d. 766), a leading adviser of Pippin. In 765 Pippin also introduced compulsory tithes to the church, which dramatically increased the wealth of the episcopal hierarchy everywhere in Francia. The help the church gave Pippin in 751 was already paying off, on a substantial scale.
This was the pattern Charlemagne inherited in 768, together with his brother Carloman II (768-71): the two got on badly, and Carloman’s early death was perhaps not unplanned. Charles Magnus, ‘the Great’, was initially called this to distinguish him from his own son Charles, but already in the ninth century the adjective began to be used to mark his especial charisma, and he is one of the few people in history to find their epithet absorbed into their own name, ‘Charlemagne’ in modern French and English. One of the early signs of this charisma was the fact that two exceptionally forceful rulers, Charles Martel and Pippin III, became reduced to predecessors, and are hard to see clearly in our later eighth-century sources. Charlemagne followed Pippin’s political path, but across his long reign transformed it, transforming the parameters of European politics as he did so, for a longer period - three centuries at least, arguably - than any other single early medieval ruler.
The first element in this was simply war, which certainly continued the practice of the previous two generations, but greatly extended it. Four areas stand out in Charlemagne’s wars. The first is Saxony, Francia’s northern neighbour, and location of border wars for over two centuries. Saxony was pagan; it was also not a single polity, but rather a collection of small tribal territories which met in a single annual assembly and fought in larger or smaller groupings according to choice and need. Charlemagne from 772 onwards set out to conquer it. He started, programmatically, by sacking the major Saxon cult-site, the Irminsul, and taking home a rich booty, but it took him over thirty years to complete his task (in 804; there was also a period of peace, when Charlemagne thought he had won, in 785-93). Saxony was hard to conquer precisely because it was disunited, and it was the theatre of considerable violence, not least for the 4,500 Saxon prisoners massacred in 782 after a Frankish defeat. The conquest was by 780 associated with a conscious process of Christianization; this was one of the few conversion processes openly brought about by force in our period. More important perhaps, Frankish conquest resulted in a social revolution, in which members of the Saxon aristocracy were given for the first time landowning rights over their free neighbours, alongside Frankish incomers and a newly endowed Saxon church system. Saxony remained marginal to Carolingian politics, but the wealth of that aristocracy developed further, and it would be the basis for tenth-century kingship itself in East Francia, as we shall see in Chapter 18.
The second area was Lombard Italy, and it was an easier task. In 773 Charlemagne was asked for his help by Pope Hadrian I (772-95), just as Pippin had been; this time he went the whole way, and annexed the Lombard kingdom in 773-4 in an unusual summer and winter war. Conquering Italy was a controversial decision (several of Charlemagne’s advisers, including his mother Bertrada and his cousin Adalard, were against it), but it turned out to be straightforward once the Lombard capital, Pavia, fell, for the kingdom was sufficiently centralized for resistance to cease almost completely. Again, wealth flowed north to Charlemagne’s treasury. Italy was, however, not absorbed into the Frankish lands in the way Saxony would be (and Alemannia and Aquitaine had already been). Charlemagne took the title of ‘king of the Franks and Lombards’, reflecting the fact that Italy remained conceptually separate, and Pavia remained a separate political centre, the only one in the Carolingian kingdom; after 781 a subordinate king returned to Italy, Charlemagne’s son Pippin. Lombard Italy would nevertheless be a source, not only of wealth, but also of governmental expertise, for Francia. As noted in Chapter 6, only the duchy of Benevento remained independent; in the face of Frankish power its duke, Arichis II, took the title of prince in 774.
Of the old areas of Merovingian rule, the last one still to remain autonomous was Bavaria. Duke Tassilo III (748-88) had begun as a protégé of Pippin III, his mother’s brother, to whom he swore an oath of fidelity in 757 at adulthood; but he stopped participating in Pippin’s wars in 763, and ran an independent politics for two decades; he was particularly close to the last Lombard king, Desiderius. After 781 Charlemagne sought to rein him in, and he threatened invasion in 787. Tassilo’s aristocracy persuaded him to capitulate, and he became Charlemagne’s vassal, or sworn follower. This was not enough, however, and in 788 he was victim of a show trial for disloyalty. A tribunal of Franks, Bavarians, Lombards and Saxons, a rarely invoked image of multi-ethnic cooperation, condemned him to death. Charlemagne then commuted this sentence to forced penance and he was, like Childeric in 751, tonsured and confined in a monastery. The trial of Tassilo in itself marks the Carolingians as different from their predecessors. It has been noticed by historians that, whereas the Merovingians killed those who lost royal favour, the Carolingians often simply imprisoned them, and confiscated their land. This is an exaggeration; the Carolingians often did kill opponents, or else blinded them (following both Visigothic and Byzantine practice: cf. above, Chapter 11). But the ritual of a legal condemnation to death, followed by the ‘milder’ sentence of blinding or imprisonment, did become rather more common, and the deaths by slow torture of the sixth and seventh centuries virtually disappeared. Imprisonment did not always work (people escaped), and death might well then follow, but these changes do show a growing belief that a show of legal process and an elaborate ritual of political exclusion were good ways to marginalize opponents, and that killing was not always necessary. They fit in with other Carolingian changes, as we shall see. In the meantime, Bavaria and the Bavarian aristocracy (who survived almost without exception, apart from the ruling Agilolfings) were absorbed directly into the Frankish political system.
The absorption of Bavaria brought Carolingian borders eastward to the lands of the Avars, and Avar wars began in 791. Avar power was by now far less great than it had been in the early seventh century, but the wealth of the Avar khagan remained enormous. In 795-6 three armies were sent eastwards to the Avar royal residence, the Ring, located somewhere on the Hungarian plain. The sack of the Ring produced booty on such an immense scale that it enriched the Carolingians and their magnates (including the pope) for a generation - Einhard said in his Life of Charlemagne that ‘no one can recall any war . . . that left them richer or better stocked with resources’. The Avars were not conquered, but they soon disappeared, their place taken by newer Sclavenian polities, who remained on the Frankish/Bavarian borderlands (see Chapter 20 for the term Sclavenian).
By 804 the lands ruled by Charlemagne were half again as large as in 768, and over twice the size of those ruled by Charles Martel at his death. Nearly all borders were further away than in 768, even that of Spain, where northern Catalonia had been taken from the Arabs in 785 and 801. This was a fairly thin strip, however, and Charlemagne’s bolder attack on Zaragoza in 778 led to one of the few military setbacks of the reign, the attack on the retreating Frankish rearguard by the Basques at Roncesvalles in the western Pyrenees. The Carolingians had new neighbours now, the Danes, the Arabs, the Beneventans, and half a continent of Sclavenian tribes from the Baltic to the Adriatic. Few of these gave rich pickings, and they were mostly fairly far away. Expansion stopped as a result. Carolingian military activity largely became one of policing, and extracting tribute from, their still independent neighbours, for a generation. It has been plausibly argued that this had bad consequences for the Franks, for their aristocracies now had to aggrandize inside, not outside, the Frankish kingdoms; kings themselves had greater difficulties as a result. But this too was a generation away in 804, and had other roots as well. Charlemagne’s last decade was one of relative peace, and unheard-of prosperity for the ruling élite of Francia by early medieval standards.
It is worth insisting a little more on the roots of this prosperity. Charlemagne had conquered new territories, and seized, not only extensive booty, but the royal treasure of two peoples, the Lombards and the Avars: essential resources for royal generosity in gift-giving, to aristocrats and to foreign rulers, which the Carolingians needed as much as their predecessors. He also now controlled the royal land of Italy and the ducal land of Aquitaine and Bavaria, and the confiscated land of rebels across the whole of Saxony and (to a lesser extent) elsewhere; and also a network of new offices, counties, abbacies and bishoprics, to add to those in the Frankish heartland. (Over all Charlemagne’s lands, there were some 600 counties and 180 dioceses.) All of these could be given out to his supporters as honores, ‘honours’, as both royal land and offices were called. So could the extensive lands of churches and monasteries, which all the Carolingians disposed of without many qualms when they needed. Royal wealth was thus the wealth of aristocrats as well, as long as such men were in the king’s favour. The lands and offices were revocable; Charlemagne gave few permanent landed gifts, preferring to distribute royal and church land as temporary cessions, beneficia or ‘benefices’. Aristocrats hoped to keep these and pass them to their sons, but had to remain committed to the king, faithfully attending court, in order to do so. And there was so much wealth around in these decades that Charlemagne could attract whom he liked to his court, including poets and intellectuals from outside Francia, and endow them as he chose. The self-confidence of the Frankish élite became sufficiently great that it was by the 790s possible for writers to describe them as in effect the new chosen people in succession to the Jews; Old Testament imagery was standard in Carolingian political programmes, and Charlemagne was commonly called David by court intellectuals. Hence or otherwise, it may be added that the Carolingians were notably tolerant of Jews, and Charlemagne’s son Louis the Pious (814-40) in particular protected them, to the great distress of writers like Agobard archbishop of Lyon (d. 840), who came from ex-Visigothic Spain, and had inherited the anti-Semitism of late Visigothic political culture. In less religious imagery, Einhard preserves for us with some smugness a Byzantine proverb, ‘[if] you have a Frank as your friend, [then] he is not your neighbour’, which he actually cites in Greek; the Franks were proud of their greed and aggrandisement, and regarded it as a proof of their virtue.
The court crystallized in two further ways in the 790s. The first is that in the years 794-6 Charlemagne founded his own capital, at Aachen in the heart of Pippinid northern Austrasia, and across the next decades he and his son Louis endowed it with ambitious buildings, one of which, the cathedral-scale palace chapel, still survives. As Charlemagne grew older, he spent more and more time here (it was close to the Ardennes forest, one of the best royal hunting reserves), and it became a stable political and administrative focus for the first time in Frankish history. Kings still moved around, taking their court with them, but two generations of courtiers came to see Aachen as a natural backdrop for politics. The second is that in 800 Charlemagne obtained a new title, emperor, in a ceremony in Rome, in which he was anointed (again) by the pope. The importance of this title should not be exaggerated; it was only honorific. But Charlemagne was proud of it, and was keen to get recognition of it by the Byzantine (as one might say, the ‘real’) emperors, which he achieved in 812 after menacing the still-Byzantine enclave of Venice. Imperial imagery began to infuse Carolingian legislation after 800 as well. The truth is, though, that already by the late 780s, thanks to his military successes, Charlemagne had achieved a western European-wide dominance, and a near unanimity of support from his subjects, a political centrality, that is to say, that no one had matched in those lands since the Roman emperor Valentinian I. Even the strongest Merovingians, Clovis or Dagobert, did not rule as widely or enjoy such long-lasting success. Charles Martel’s military machine, and the luck of four almost unbroken generations of single rulers (for Charlemagne’s sons, between whom he fully intended to divide his lands, all predeceased him except Louis), were the basis of this success, but Charlemagne’s charisma capped it. The question would then be what he would do with it.
It cannot be denied that Charlemagne - and his advisers, but animated beyond doubt by the king himself - had a conscious and ambitious political project. In the widest sense, it was one of ‘reform’ (renovatio), or, a much commoner word, ‘correction’ (correctio), of the inner life as well as the external acts of lay and ecclesiastical subjects alike. It is very clear in one of Charles’s relatively early legislative acts, the General Admonition of 789. In this widely circulated text, the king re-enacted canons from church councils to provide a template for the proper activity of clerics, but also instructed the laity in the necessity of concord, justice, the avoidance of perjury, the avoidance of hatred, and, overall, the necessity of the preaching of the Christian faith. These were keynotes of the moral reform programme of the Carolingian period. They were matched by a systematic education programme, which was (as was the General Admonition) largely the work of the most influential intellectual of the first generation of the Carolingian reform project, the Northumbrian Alcuin (d. 804). Alcuin was at Charlemagne’s court for most of the period 786 to 796, and then continued teaching in one of the several monasteries Charlemagne gave him, Saint-Martin in Tours. As the king said in an open letter of the 780s or 790s, also written by Alcuin, good behaviour and spiritual understanding were impossible without a literary education, for ‘knowing comes before doing’, and even the Bible was full of figures of speech which had to be decoded. The Carolingians promoted basic literacy, but expected more, especially from leading clerics and aristocrats: a proper understanding of the Bible and of theology, without which a path in the Carolingian political world could not properly be walked.
The successes and failures of this project have been very intensively discussed; but that there were successes is not at issue. The whole of the Carolingian élite cared about theology, or had to pretend they did. Already in 794 an assembly of bishops and magnates at Frankfurt could devote much of its time to discussing heresies, Adoptionism and the Byzantine rejection of Iconoclasm (the Franks had greater sympathy with the Iconoclasts), for the first time in the West in two centuries. By the 830s and 840s, the whole political process, including coups and civil wars, could be seen in theological terms. By then, there were two dozen or more political actors who were also active writers, participating in what were often pamphlet wars about the theology of political practice. Some of them were lay aristocrats, including Dhuoda (d. c. 843), wife of the sometime royal chamberlain Bernard of Septimania (d. 844), who wrote a handbook on correct behaviour for her son, suffused with biblical imagery and citing an array of church fathers, which were evidently available to her in Uzès, far in the south of the Frankish lands. This will all be discussed in the next chapter, but it marks the Carolingian period out.
Exactly why this project developed is rather harder to understand. Many of its roots are obvious. The Carolingians had to identify with the church, for it was the church that gave them legitimacy as a ruling family; the coup of 751 was still in living memory at the time of the General Admonition. The church councils, which had become commoner again after the 740s, and which continued without a break thereafter, were a natural source of moralizing enactments, many of them absorbed into royal legislation already under Pippin III. Frankish self-confidence led to Old Testament parallels, as we have seen, and also to Roman parallels, thus encouraging people to look back to the fourth to sixth centuries, when correct belief was a burning political issue (see above, Chapter 3). Although the Merovingian period was not an age of explicit ideological programmes in Francia, seventh-century Visigothic Spain had been, showing that an overtly moralized politics already had potential roots in early medieval western soil; and Theodulf bishop of Orléans (d. c. 826), the major theologian of Charlemagne’s reign, was of Visigothic origins. (It must be said, however, that the Franks, if they borrowed from the Visigoths, did not borrow the Gothic zeal for religious exclusion, as we have already seen.) Once Alcuin, Theodulf, Paul the Deacon from Lombard Italy, the Franks Angilbert of Saint-Riquier and Einhard, and others, combined in Charlemagne’s court in the 780s and 790s, a critical mass of intellectual debate and competitive writing ensued, enough to expand and continue for another three generations. But it is hard not to see a plan at the back of this. It was Charlemagne who invited these intellectuals, and gave them such big gifts that they stayed in or near the court for decades. Programmatic legislation, too, although not, of course, composed by him, went out in his name, and was new. The successes of the 770s (particularly in Italy) seem already to have persuaded the king that he was special, and that he had a mission, not just to rule the Franks and their neighbours, but to save their souls. He may have been educated to this in the already more ecclesiastical political environment of Pippin’s reign - however incompletely; Charlemagne could appreciate poetry and theology, but he never fully learned to write. All the same, it seems to have been his own choice. Charlemagne thus matches Justinian as an innovator in moral-political practice (although he had a better sense of humour than Justinian; his son Louis, famous for not smiling, was a better parallel there). The fascination with him that has resulted in such a dense historiography, unbroken across the centuries but if possible even more elaborate now, is not entirely unjustified.
All kinds of legislation were commoner under Charlemagne. Royal assemblies produced capitularia, ‘chapter-collections’ or ‘capitularies’. These varied in their formality (some were official written texts; some seem to have survived only because participants took private notes of their content); they also varied in their aim, for some were guidelines for local representatives, some were one-off enactments, but others were systematic additions to existing law, Frankish or Lombard. But there were many of them; the standard capitulary edition has eighty-five from the reign of Charlemagne alone, plus some enactments that survive in more fragmentary form. Some of the impetus for this must have come from Italy, for they start in the late 770s, and are matched in frequency earlier than that only by the Lombard laws of Liutprand; church council legislation, which partly overlaps with capitulary legislation (as with the General Admonition, and the 794 synod of Frankfurt), was another model. Charlemagne also reissued the Lex Salicain a new edition, which was widely copied in the ninth century, and made laws for newly conquered peoples such as the Saxons. Not all capitularies were widely copied, it is worth stressing; many survive in only a single manuscript. When Ansegis, abbot of Saint-Wandrille on the Seine, went looking for capitularies to turn into a rearranged collection to present to Louis the Pious in 827, he only found (or used) twenty-nine of them, and only one (the General Admonition) from before 803. As in the Roman empire before theTheodosian Code, it was hard to be sure what laws had been passed, even though the Carolingians, Roman-style, regarded ignorance of the law as no excuse. But some were very carefully circulated, such as the capitulary adding to Salic law of 803, which survives in fifty-three manuscripts (Ansegis used it, too), one of which states that Stephen count of Paris had his copy of it read in a public assembly there, and local political leaders signed their names on it. Such a mixture of oral publication and formal subscription was probably common for the major enactments. The capitulary ‘habit’ continued under Louis the Pious, at least up to 830, and then in West Francia and Italy until the late ninth century; in East Francia, too, the acts of church councils continued to be recorded. In the ninth century, informal capitulary collections begin to be commoner as well, particularly but not only in Italy; they seem to have been intended for use in court. None of them were ‘complete’ sets (capitularies tended, after all, to be repetitive), but they do attest to a recognition that a wide range of new law now existed, and that it was useful to be informed about it.
These laws, and the other sources for Charlemagne’s reign such as annals and letter collections, show that the government of the Carolingian lands was essentially based on old foundations, but that these were fairly carefully reshaped as required. The network of public assemblies that were crucial for the Merovingians and the Lombards remained crucial in the Carolingian period. Royal assemblies were held just before the campaigning season every year and were the points of reference for army muster as well; kings could call smaller or larger assemblies later in the year, too, to prepare policy for the next year or if there was urgent business. Major political figures, lay and ecclesiastical, attended regularly. These were venues for genuine discussion, not just royal instructions; Hincmar archbishop of Reims (d. 882) in his 882 treatise On the Organization of the Palace (which itself drew on Adalard of Corbie’s lost text of c. 812 with the same title) indeed tells us that kings did not attend all assembly discussions, but instead stood outside glad-handing - and Hincmar was one of the major advisers of King Charles the Bald (840-77), as Adalard had been for his cousin Charlemagne, so whichever wrote this would have known. Early in Charles the Bald’s reign, during the preparation for the civil war of 841-2, Charles’s follower and cousin Nithard (d. 845) records in his contemporary history how Charles’s May 841 assembly argued about which way the king and his army should march; Charles went with the minority, not the majority, view - wrongly, in fact, Nithard said - but, either way, he had the benefit of hearing real argument. Even without that argument, participation in assemblies, and in the rituals normal in all of them, powerfully reinforced a sense of collective participation in public affairs.
These national assemblies were matched in every county by local assemblies, placita, meeting two or three times a year under the count’s presidency, in which local élites were brought into the same public network; these heard reports of national deliberation (Count Stephen’s Paris gathering of 803 was one such), and decided on court cases. The Carolingians regularized these assemblies, too, for example determining that local judicial experts should be called scabini everywhere, which by the early ninth century they were indeed coming to be, from the English Channel to Italy. It was also county assemblies that administered the taking of oaths to the king, another older tradition systematized in this period. Charlemagne instituted these in 789 after regional revolts in Hesse and Thuringia in 785-6; in 793 he had them repeated after a second revolt, by his disinherited eldest son Pippin in 792, since some of the rebels said they had not sworn in 789, perhaps because they were too young (not that this did them much good; Charlemagne had them killed). These were the only revolts in Francia in his reign, and they seem to have been fairly small-scale, but the king’s response was to make formal oaths more systematic. Every free man over the age of twelve had to swear, and their names had to be recorded by counts and missi; in 802 these obligations were further extended, as oath-swearers had to swear a much more detailed oath to the emperor. Oaths mattered in this world; oath-breakers were perjurors, and risked damnation, not just secular penalties - dispossession, mutilation and sometimes death. They could be dangerous: Charlemagne banned oaths of association made to anyone except the king and one lord, and in 806 enacted that men who did so should beat each other and cut off each other’s hair (or, in extreme cases, slit their noses). Oaths to the king further added to the intensity of ritual at even the most remote assembly, and to the local presence of royal authority.
The Carolingian empire was huge, larger than any subsequent state in Europe has ever been except for brief years at the height of the power of Napoleon and Hitler, and also extremely diverse, stretching as it did from the half-converted and roadless lands of Saxony to the old urban societies of Provence and Italy. How it could all be controlled, without the elaborate fiscal and administrative system of the Roman empire or the caliphate, was an almost impossible challenge. Assembly politics was one part of it; so was army muster; and the palace, the court of the king or emperor, whether at Aachen or elsewhere, was furthermore a magnet for the ambitious in every period, as they came to seek justice, gifts or preferment. Kings did not just give gifts; they received them too, the ‘annual gifts’ of horses and the like presented at each general assembly. These gifts seem to have had a military edge to them, and were probably associated with the fact that soldiers on campaign had to bring their equipment and three months’ provisions with them, not a small investment. Rather than a proto-tax system, which cannot be identified in the Carolingian period (kings were not short of resources even without taxation, until late in the ninth century), this was another element in the gift-exchange of political participation. Palaces were also the focus of a particularly large amount of collective and increasingly moralized ritual, as we shall see further in the next chapter; the other elements of Carolingian political aggregation had clear roots in the Merovingian period, but this was largely new. But kings did not move around the whole of the empire, except when on campaign; Charlemagne, Louis and Louis’s sons seldom strayed out of the three great ‘royal landscapes’, of the Seine valley, the Middle Rhine valley, and between them the core block of royal and ex-Pippinid estates around Aachen. Not every local leader ever went there; the kings had to reach them too.
One way they did so was by strategically placing their most trusted aristocrats. Counts tended to be from long-standing local élites, except after conquests, as in Alemannia after Canstatt, or in Italy in the early ninth century; so did bishops. But beside these local élites, and interlocking - and intermarrying - with them, there were also greater families, those of the Reichsaristokratie, the ‘imperial aristocracy’, as Gerd Tellenbach called them in 1939. He and his successors identified between forty and fifty such families, who could be found in any part of the empire, and whose members could move around (or be moved around) with some facility. Most of them were from the old Pippinid heartlands of Austrasia, extending southwards into the Middle Rhine and northern Burgundy, though they could come from anywhere except Italy. Very few if any of these families were newly created; but the Carolingians could make favoured members of them rich and powerful beyond any previous imagining, even though Merovingian aristocrats could already be pretty rich, as we saw in Chapter 5. A well-known example of these is the ‘Widonid’ family (as we call them - surnames did not yet exist), originating in the Middle Rhine and Moselle valleys; they seem in the eighth century to be linked to Milo of Trier (see Chapter 8) and to an important church in Mainz. Under Louis the Pious and his sons, they are found simultaneously in the far west of modern France and in the duchy of Spoleto in the central Appennines of Italy, running the frontier marches facing Brittany and Benevento respectively, while keeping their Rhineland links, where they controlled the major monastery of Hornbach. They did not follow a simple family political line (in the crisis of 833-4, which set Louis the Pious against his sons, Guy count of Vannes fought a battle for Louis against his brother Lambert marquis of the Breton march, fighting for Louis’s son Lothar, and was killed), and they could be unscrupulous about establishing themselves locally, as in distant Spoleto, where they ran a largely autonomous politics. All the same, they were loyal to Carolingian ideals, including Carolingian unity - Guy III of Spoleto (d. 895), after Carolingian power ended in 887, tried to make himself king in both West Francia and Italy, and was actually crowned emperor in 891. Without that unity, the geographical range of their power would have ceased to exist, and, indeed, did cease, for the family is not attested after the 890s outside the Rhineland (though there it remained important: the Salian dynasty of German kings was probably descended from it). Kings relied on families such as this a great deal, but the reverse is true too; in many respects the Carolingian empire was an immense oligarchy, and, given the rooted local power of aristocracies both large and small, it had to be. The point will be explored further later.
Not all royal dependants in the provinces were from great families like this. The Carolingians made considerable use of royal vassals, not all of whom were rich, but all of whom had particularly close ceremonial ties to the kings, in rituals of personalized oath-swearing and homage. These could be local men, called to the palace and the army, or else aristocrats, both rich and middling, brought in from outside; either way, they are invoked in legislation as the sort of men kings could particularly rely on. (Aristocrats had, and relied on, their own vassals as well.) Vassalage was the lineal successor of the personal fidelity of the Merovingian world and of Lombard Italy; what was new about it was once again that vassals might be moved around. It is this movement of men, of families, which marks the early Carolingians out from their predecessors.
The kings also, systematically, sent representatives to the provinces. These representatives, missi, were the king’s eyes and ears. They had Merovingian and especially Lombard antecedents too, but Charlemagne regularized them, and the Frankish heartland was in 802 divided into missatica, territories in which pairs of missi, a count and a bishop, regularly toured, to hear appeals against local counts and others. Italy and most of the other conquered lands had missi of their own. Missi were not often outsiders to their territory - local archbishops were popular missi, for example - but they again owed loyalty and responsibility directly to the king, to whom they were expected regularly to report, in writing if necessary. We have some of the court cases in which they held local officials to account, such as the 804 case at Rizana in Istria in which three missi heard the complaints of 172 local leaders against Duke John of Istria’s trampling of local customs; John apologized, and the customs seem to have been restored. It would be wrong to see missi and their territories as fully institutionalized, but kings certainly regarded them as normal until late in the ninth century, except, it seems, in East Francia. And we certainly have chance-surviving evidence of regular written communication, to the provinces and back again, whether through people called missi or other officials, such as the instruction from Hetti archbishop of Trier (as missus) to the bishop of Toul in 817 telling him to mobilize against the revolt of King Bernard of Italy, that very day; or the letters Louis the Pious sent in 832 to tell two vassals to stand by as messengers in case his missus or his count needed to send a message to the emperor; or the demand made by Charles the Bald to his churchmen in 845 for systematic information about his monasteries, which Abbot Lupus of Ferrières sought actively to fulfil; or the lists of men who swore fidelity to Charles the Bald at Reims in 854, attached to a copy of a capitulary by Archbishop Hincmar, who was probably himself the local missus. Men must have been moving around the entire time, looking for the king/emperor, or sometimes, the queen (this was not straightforward, for they moved about too), and informing them; Hincmar indeed supposes in On the Organization of the Palace that receiving them was a major royal task. (Aristocrats and bishops had their own communications networks, to keep abreast of politics, which presumably filled the roads still more.) Without this presumption of regular and detailed communication, again not new but greatly extended, running the empire would not have been possible.
Did this complex network of instructions and accountability actually work? There are two views. One is that the complexity and flexibility of the Carolingian administration was self-supporting. The kings and their advisers were constantly innovating and retouching, and could move quickly; Louis the Pious’s muster against Bernard in 817, for example, was so fast that it caught the rebel entirely by surprise. The ‘system’ of the capitulary legislation or of Hincmar’s On the Organization of the Palace was more flexible in reality, and that was a strength, for it could be moulded to fit the diversity of the provinces. And the centrality of the royal court (or, after 840, courts) remained undiminished, as all political leaders or would-be leaders continued to circle around kings into the 880s, imbibing as they did the elaborately moralized programme of Carolingian correctio; there is good evidence for aristocratic literacy and even book-buying, which backs this argument up. This was further extended into the provinces thanks to the network of rich royal monasteries, from Corbie in modern northern France to St. Gallen and Reichenau in modern southern Germany and on into Italy, and the even denser network of cathedral communities, many of which had extensive libraries, and trained intellectuals who could and did debate about theology and politics until the end of the ninth century, with effects on political practice in some cases.
The other view is that this was all a sham. The aristocracy, secular and ecclesiastical alike, were corrupt and out for themselves, from top to bottom. Theodulf of Orléans wrote a poem around 800 against (among other things) judicial corruption, which would have been incomprehensible to the people of his south French missaticum, given the degree to which litigants apparently pressed gifts on him; many of the abuses missi are recorded as correcting were in fact the oppressive acts of other missi; Adalard of Corbie’s younger brother Wala (d. 836), when a missus for Italy in the 820s, uncovered an elaborate cover-up of the expropriation and later murder of a widow which stretched from top to bottom in the Italian kingdom; Matfrid count of Orléans, one of the major court figures of the 820s, was criticized in about 827 by Agobard of Lyon for providing ‘a wall’ between the emperor and criminals, ‘to protect them from correctio’; there are plenty of other examples of aristocratic bad behaviour from the period, which was in fact also marked by a notable oppression of the poor, as capitularies themselves tell us. As for the imperial project, it was already disintegrating in the 830s and was only fully maintained after that by Charles the Bald and his adviser Hincmar; most other Carolingians soon moved towards the rougher realpolitik of the tenth century. In any case, the ambition of Carolingian reform legislation betrayed its hopeless naivety, and its constant repetition betrayed its failure. (Maybe this was a good thing, Michael Wallace-Hadrill thought, writing in an otherwise sympathetic account: ‘had [Hincmar’s programme] worked out, Carolingian society would have been a police-state.’) The Carolingians were unusual only in their rhetoric, and in their military success, which petered out in the ninth century, leaving the empire open to civil war and demoralizing (because unremunerative) defence against external attack.
The interest of the Carolingian period lies in the fact that both of these views are largely accurate. Aristocrats are always violent, corrupt and greedy, but they were at least aware of the ideology of public responsibility in this period, and presumably - sometimes, as with Dhuoda, demonstrably - linked it to their desire for personal salvation after death, which they certainly always also possessed. The state was ramshackle and far too large for the governmental technologies of the period, but it is, all the same, constantly striking how often it makes its presence felt even in resolutely local document collections. Throughout the ninth century, we have examples of peasants appealing to public courts against their lords, in Italy, Francia, Septimania (modern Languedoc), over personal status, rent levels or seized lands; they almost always lost, but the fact that they bothered to do so at all, in a political system so obviously run by the aristocracy, implies that they knew the system could at least sometimes work as it was supposed to, and such cases are much rarer later. There was a constant dialectic between the state, with its immense patronage powers, and local societies, throughout almost the whole empire (royal power fell back only at the edges, like eastern Bavaria, Spoleto or Catalonia). Local powers had to pay attention to kings, and accept their political guidelines, including whatever ideological programmes they had, not least because kings were also dangerous, and by no means did all the things their own programmes enjoined. We shall explore these contradictions, and their ironies, further in this chapter and the next.
Charlemagne died in 814, and Louis the Pious, who had been crowned emperor by his father the year before, immediately marched north to Aachen from his sub-kingdom in Aquitaine to take over. He represented himself as a new broom, and summarily expelled his sisters, led by Bertha, from the palace, where they had been acting as a sort of collective queen for their father since his last wife died in 800. The imagery of Louis’s early years stresses his moralism, as opposed to the sexual licence of his father’s reign; Charlemagne had had a string of mistresses up to his death, and his daughters, whom he would not allow to marry, had lovers too - Bertha’s was the court scholar Angilbert, by whom she was the mother of the historian Nithard. Louis’s own sex life, once he became an adult, was in fact as far as we know restricted to the marriage bed, unlike most male Carolingians, but his criticism of the sexual immorality of the palace (the ideal moral centre of the polity, thus very vulnerable to such criticism, as we shall see in the next chapter) was a standard part of ninth-century political rhetoric, and would be applied back to Louis’s own court in the 830s. Louis was committed to monastic reform, and his first substantial political initiative was two reform councils at Aachen in 816-17, which revised the Rule of Benedict of Nursia and extended it to all the monasteries of the empire. In 817 he also set out how the empire would be divided at his death between his three sons, which excluded from the succession Bernard, son of his brother Pippin, who was already king of Italy (812-17); Bernard unsurprisingly revolted, with the support of not a few Frankish magnates (including Theodulf of Orléans), but, as we have seen, failed. He was tried in 818 and condemned to death, but, following the common Carolingian pattern, this sentence was commuted to blinding, from which however he died anyway.
After 818, Louis understandably had little opposition for some time, and the next decade can be seen as the apogee of Carolingian self-confidence. Wars were small-scale by now, and the emperor’s attention was focused on an elaborate and complex court politics in Aachen, marked by regular embassies from different neighbours, another dense set of capitularies (many of them collected by Ansegis in 827), and an administrative reordering under the arch-chancellor Helisachar (814-30), who had come with Louis from Aquitaine, and the arch-chaplain Hilduin, abbot of Saint-Denis and four other monasteries (819-30). The emperor’s control of court ritual was marked above all by his decision in the 822 general assembly at Attigny to perform a public penance for the death of Bernard, imitating Theodosius I’s penance of 390, according to one of his biographers. At the same time, he called back the (male) relatives he had exiled from court, notably his cousins and possible rivals Adalard and Wala; Carolingian family reconciliation was to be complete.
The calm of the 820s was, however, broken abruptly in 829-30. Court factions were crystallizing around, on the one side, Louis’s oldest son Lothar (817-55), already emperor (since 824) but with a political remit confined to Italy, and, on the other, Louis’s second wife Judith and her family. In 828 Lothar’s father-in-law Hugh count of Tours and his associate Matfrid of Orléans had lost their offices. In 829 Bernard of Septimania, count of Barcelona, was brought in as chamberlain, an office traditionally very close to the queen, and was for a few months regarded as ‘second to the king’; he was (for unclear reasons) a highly controversial figure, however, and by 830 was accused of adultery with Judith. Lothar gained the support of his brothers Pippin king of Aquitaine (817-38) and Louis king of Bavaria (817-76) to set in motion in April 830 a quiet coup, significantly also supported by the old guard of the court, Helisachar, Hilduin and Wala. Bernard fled and Judith was temporarily exiled, until Louis the Pious regained control in October and brought Judith (but not Bernard) back. In 833 tensions rose again, and much the same occurred; this time, the emperor Louis marched with an army to meet Lothar and his brothers, who were joined by Pope Gregory IV, in Alsace. At the meeting-point, later called the ‘Field of Lies’, Louis’s army melted away, joining Lothar, and Louis was deposed in favour of Lothar. This time his public penance was not voluntary; the best he could do was refuse to take monastic vows when he was confined in Saint-Denis. But, as in 830, Lothar and his brothers fell out - Lothar, like his father, was too clearly committed to being the dominant Carolingian - and Louis was restored in 834. He was ceremonially re-crowned at Metz in 835, and re-established himself, confining Lothar to Italy again, though Louis did not take violent revenge on any of Lothar’s supporters (they merely lost their lands and offices north of the Alps, and some of them, such as Hilduin, soon got them back). Louis then remained in control until his death in 840.
The events of 830-34 certainly greatly disrupted the balances of imperial government and the patronage networks of the Carolingian lands. They have also been typically seen until very recently as a sign of imminent Carolingian breakdown, perhaps fuelled by aristocratic hostility, and also as a sign of the weakness of Louis ‘the Pious’ himself. Louis was not, however, either pliable or accommodating, any more than his sons were - hence, indeed, the fact that the uprising occurred twice; and aristocratic reactions to the crisis show alarm rather than any sense of a new opportunity. Einhard (d. 840), by now in retirement in his monastery of Seligenstadt near Frankfurt, although a supporter of Louis (he preserved in his letter collection a very rude letter to Lothar, written in 830), prudently fell ill during both crisis moments, but then was worried that this might be taken the wrong way by the kings, and wrote to friends at court to ask them to ensure that his loyalty was recognized, by Louis the Pious, but also by Louis of Bavaria (whose power-base was close to Seligenstadt), and even by Lothar; one letter to a dependant in 833 asks him to give the ‘customary gifts’ to the temporarily victorious Lothar, and to report back on how Lothar received them. Einhard was, thanks to his long-standing palace connections, a major local patron and political intermediary, and it is clear in his letters of these years how much mediation would need to be done in a period of sharp political swings, for the kings could and did remove the benefices of the less than fully loyal. So Einhard in late 833 wrote to a friend asking him to intercede with Lothar for a certain Frumold, who had been given a benefice near Geneva by Charlemagne but was too ill to travel to court and commend himself to the new ruler (Geneva was a long way from Seligenstadt; Einhard’s patronage stretched widely); or again, around the same time, to another courtier who might, he hoped, persuade Lothar to let an aristocrat and his brother hold benefices jointly in the kingdoms of both Lothar and Louis of Bavaria. That Einhard kept these letters indicates that they were normal, and also, perhaps, successful: his younger contemporary the poet Walahfrid Strabo (d. 849) wrote a prologue to Einhard’s Life of Charlemagne noting rather wryly how well the author had kept ‘a certain remarkable and divinely inspired distance’ from the crises of Louis’s reign. This was unlike Walahfrid himself, in fact, who was exiled from his monastery of Reichenau by Louis of Bavaria in 839-42; Walahfrid is thus doubly a witness to how hard it was to avoid trouble in the 830s. This was not a crisis period which magnates would easily seek to exploit.
It is probably best to see the crises of the 830s as a product of two underlying problems, a struggle between court factions, and the normal tensions any ruling Carolingian had with adult sons itching to succeed. This confluence was only exacerbated by arguments over theology and political ethics, and the more mundane fact that Judith gave Louis a fourth son, Charles, in 823, who would have to be fitted somewhere into the partitioned empire (he was given Alemannia in 829, a politically tangential area, but in a significant year - Nithard later thought that this was the excuse for Lothar’s first rebellion). It has at least to be said that Louis’s father Charlemagne managed his sons better, and so did Louis’s own sons: Lothar, Louis and Charles each weathered the rivalries of their adult sons without ever losing the initiative. Misjudgements in the crucial years around 830 seem to have marred Louis the Pious’s standard toughness. After Louis’s death in 840, however, it is not hard to see how his heirs fell into civil war. Pippin of Aquitaine had died in 838, allowing Louis to substitute Charles as his heir in the western part of the empire (at the expense of Pippin’s son Pippin the Younger), which ought to have made things easier; but Charles ‘the Bald’ and Louis ‘the German’, as historians from now on call them, were not at all inclined to let Lothar have the leading role which he regarded as his right. It was because of this that civil war ensued in 841-2. A bloody but inconclusive battle at Fontenoy in 841 scared the Frankish magnates, however - another sign that they were by no means ready to exploit crisis - and Lothar, driven out of Aachen in 842, agreed peace; the empire was divided again, rather carefully, at the Treaty of Verdun in 843. Charles took West Francia (including Aquitaine), Louis East Francia (including Bavaria, Alemannia and Saxony), Lothar the lands around Aachen, Burgundy, Provence and Italy. The Frankish heartland, where royal estates were thickest, was divided neatly into three; each brother got one of the ‘royal landscapes’, and was in addition assigned the outlying kingdom in which he was strongest. The fact that the division looks idiotic on a map, much as Merovingian divisions often had, underlines the extent to which all three brothers still saw the empire as a common project; it perhaps also shows that none of the parties really thought it would be permanent. It was permanent, however. The only major exception was the lands around Aachen, named Lotharingia after Lothar’s son Lothar II (855-69) who inherited them, which were divided between Charles and Louis at Lothar II’s death. (Aachen became marginalized after that, as a borderland; in the tenth century Lotharingia was absorbed into East Francia.) Verdun should not be overstated as a dividing point all the same. We know that West Francia eventually became ‘France’, East Francia became ‘Germany’, but contemporaries did not, and the imagery of a single Francia under several rulers survived until after 1000, as we shall see in Chapter 18.
The division of the empire was a return to the norms of the sixth and seventh centuries, and was regarded as inevitable and indeed appropriate by nearly everyone; after all, Charles Martel and Pippin III had both divided their lands temporarily, and Charlemagne would have done so. It was also a return to the bickering and occasional warring of the decades around 600. Lothar’s northern heartland around Aachen looks the quietest, though this may be because the two major continuators of the Royal Frankish Annals, theAnnals of Saint-Bertin and the Annals of Fulda, were written in Charles’s and Louis’s kingdoms respectively. Louis the German, too, seems to have been in full control of East Francia, at least after his bloody quelling of a peasants’ revolt, the Stellinga, in Saxony in 842. Louis spent his long reign (he died in 876) fighting on the eastern frontier, particularly against the Bohemians, and the increasingly powerful Moravian rulers Rastislav (846-70), who was captured and blinded by the Franks, and his successor Sviatopluk or Zwentibald (870-94): these princes had expanded their power into the political vacuum that followed the collapse of the Avars. Zwentibald, in particular, fought the Franks as an equal, and had considerable influence over eastern Bavarian aristocrats by the mid-880s. But the importance of the eastern frontier, and the traditional nature of the campaigns there, allowed Louis to sustain a military effectiveness focused on offensive war that had not been known since Charlemagne’s time. Hence, doubtless, the ease with which he faced down successive revolts by his three sons in 857-73. East Francia was harder to rule, on one level, for very little of it had been part of the Roman empire, so it lacked good communications or cities except in the far south and far west; Louis probably had little direct control in still-peripheral Saxony, and rarely went there. All the same, he ran placita there and did justice, like any Carolingian king, when he did go there, most notably in 852; and, although he did not issue capitularies, and seems to have had a simpler administration than his brothers, his bishops, headed by the influential archbishops of Mainz on the Rhine - a Roman city, and in a Carolingian royal heartland - behaved just like other Carolingian ecclesiastical communities, holding councils and making law. (Louis’s first appointment to Mainz was indeed the influential theologian and biblical commentator Hraban Maur, 847-56.) This, plus Louis’s armies, made the East Frankish kingdom a still-functioning heir of that of Charlemagne and Louis the Pious.
In Italy, too, Lothar’s son Louis II (840-75), who was in sole control of the kingdom (with the imperial title) by 850, operated without recorded difficulty, and seems to have been an effective ruler. He was certainly a practitioner of Carolingian reform, and as early as 850 enacted capitularies and conciliar legislation to combat abuses, the first of an Italian sequence that would only end in 898. He and his wife Angilberga (d. 891), an unusually influential queen, had a more hands-on control over government than most Carolingians; Louis was secure enough to promote Lombard aristocrats for the first time in half a century, alongside three or four major families of the Reichsaristokratie (including his wife’s kin, the ‘Supponids’). He was clearly the heir of kings like Liutprand, while also taking seriously his imperial title; in a letter to the Byzantine emperor, he claimed to represent the whole Carolingian dynasty. Louis II, uniquely among Carolingian rulers, could take the risk of a long unbroken period (866-72) campaigning abroad, against the Arabs who had taken Bari in southern Italy; he took Bari back but was then imprisoned by Prince Adelchis of Benevento (853-78) in 871, who had no reason to welcome Carolingian power stretching so far south. This was a humiliation for Louis, and he had to be re-crowned to counteract it - but he was still unopposed in the north of Italy. Here, too, then, the norms of Carolingian power were not yet under threat.
Charles the Bald faced by far the greatest problems out of the Carolingians of this period. This, plus the extensive documentation for his reign, has meant that he is the best-studied later Carolingian, although he was also the least typical. For a start, his was the only kingdom in 843 with another claimant, Pippin the Younger, who contested Aquitaine rather effectively until 848 and then intermittently until his death in about 864. Secondly, he had to face the most systematic external attack, from Viking raiders. The Vikings in Francia and England were mainly from Denmark (Norwegian Vikings went mostly to Scotland and Ireland). They were standard war-bands of an early medieval type, on the scale of early Anglo-Saxon armies, although they were never as large as Frankish ones, even when they got bigger later in the century. They were private enterprises, in that they were not under the control of the kings of Denmark (at least, this is what the latter said when the Franks upbraided them, and it was plausible enough, given the limits on Danish royal strength: see Chapter 20). They were pagan, so were less inhibited than Christians about sacking churches, major wealth depositories, to the particular horror of ecclesiastical writers. And they were based on ships: this was the big difference from local Frankish border raiding, which was otherwise very similar, for it allowed the Vikings to hit and run, far up rivers into Francia, before any defence army arrived.
Major Viking raids began in 834, with an attack on the Rhine port of Dorestad; ship-owners were also merchants, and knew Dorestad well - as well as also knowing that the Frankish political system was busy in 834. They attacked Dorestad and, more widely, Frisia after that as well, and as early as 841 Dorestad was given in benefice by Lothar to Harald, a Danish royal family-member, and then to his younger brother Rorik. Rorik controlled much of Frisia, and defended it for the Franks more faithfully than not, for most of the period 845-75. Almost certainly as a result, Vikings seldom came further up the Rhine to bother Lothar’s and Louis the German’s heartlands, except for big raids in 881-3. Charles the Bald, however, had to face regular attacks on his long coastline, and up the Seine, Loire and Garonne, without a break from 841. Charles could never get rid of them; they were a permanent wound in his side. Vikings soon over-wintered at river-mouths as well. Charles alternately fought them off and bought them off with tribute (the least popular but most effective response); twice at the end of his reign he actually organized a general tax to pay them. Most effectively of all, perhaps, he fortified bridges over the Seine in 862 and the Loire in 873, to block their path. The major Viking push for fifteen years after 865 was into England, which eased the pressure on Francia a little in Charles’s last years. But the Vikings never really went away.
The aura of military failure, or at least crisis, thus hung over Charles the Bald, and this must be one of the main reasons why he had greater difficulties with his aristocracy than did his brothers and nephews. Charles’s anti-Lothar alliance with his brother Louis broke down in the 850s, and in 854 Louis the German’s son Louis the Younger went to Aquitaine to test out the seriousness of invitations to his father by Aquitanian aristocrats. It turned out to be weak then, but by 858 disaffection was much stronger (it was a bad period in terms of Vikings, and Pippin the Younger had reappeared in Aquitaine), and numerous magnates, lay and ecclesiastical, were prepared to invite Louis the German in. Charles still had support, not least from Hincmar of Reims and most of his other bishops, and Louis retreated; but the episode showed the uncertainties Charles had to face. The pro-Louis group, which included the powerful Robert ‘the Strong’, count of Anjou (d. 866), who was from a major Rhineland imperial aristocratic family, the ‘Rupertines’ or ‘Robertines’, gave in, and retained their honores. Charles did not have to face a revolt like this again, but he had to negotiate with critical aristocrats on other occasions too, such as when, at the end of his reign, he occupied Italy (and took the imperial title) after the death of the son-less Louis II in 875, while simultaneously attacking in 876 - and losing - against Louis the Younger (876-82), who had succeeded his father in most of East Francia. Charles was trying to assert himself as the dominant Carolingian, without securing his base. Hincmar was furious, and several of Charles’s magnates thought he was over-stretching himself. But Charles died in 877, and normal politics resumed.
Charles did remain hegemonic over his aristocracy. He built up the power-bases of his most useful fideles, such as Robert of Anjou, at least before 858, or Bernard marquis of Gothia, a new name for Septimania, who was his mainstay of support in the far south after 865. In particular, he patronized Boso (d. 887), brother of Charles’s second wife Richildis, who was made chamberlain of his son Louis ‘the Stammerer’ in his new sub-kingdom of Aquitaine in 872, as well as count of Bourges and Vienne, and in 876 Charles’s viceroy in Italy and husband of Louis II’s only daughter, Ermingard. But he also removed honores from magnates at will, and moved them around; when Robert died in battle against the Vikings, his son Odo did not inherit Anjou, and lost others of Robert’s counties in 868 - he did not return to royal favour until 882, when he became count of Paris. Similarly, Bernard of Gothia, who rebelled in 878 against Louis the Stammerer (king of West Francia 877-9), was summarily stripped of his lands and offices, and never got them back. Charles was generous with land; he gave out far more estates in full property than did other Carolingians, not just benefices; but he took them back as well with some ease.
Charles also threw himself into the complexities of Carolingian correctio and Carolingian ritual. He developed his palace of Compiègne as another Aachen, including its buildings; he created some original ceremonial, as when he hosted a month-long synod at Ponthion in June- July 876, after his imperial coronation, wearing Frankish costume at the start but Byzantine costume plus a crown at the end. Imperial echoes were already visible in the most substantial of his many capitularies, the 864 Edict of Pîtres, which draws substantially on the Theodosian Code (as well as, explicitly, on Ansegis). Charles was as concerned for administrative refinement as was his father; Pîtres, for example, also involved a coinage reform, which coin-hoards show to have been effectively implemented. His missi still ran as in Charlemagne’s day. And Charles had a court almost as full of intellectuals as Charlemagne’s, including Hincmar of Reims, who wrote much of his legislation and was always at hand for advice, wanted or not, as well as writing some of the longest political tracts of his generation, and twenty years of the Annals of Saint-Bertin. The core of Charles’s ruling was not undermined, for all his military difficulties; and his ambition as a reformer was more elaborate than any other Carolingian after 840. Even Charles the Bald, then, despite many problems, remained on top of his kingdom in most respects, in different ways from Louis the German and Louis II of Italy, but with a similar result. The Carolingian project was still in operation into the late 870s.
But it did not last a decade more. In 887-8 the empire broke up into five kingdoms, with six or seven claimants, only one of whom was a male-line Carolingian. This was seen as an end even by contemporaries, as a takeover by reguli, ‘kinglets’, as the Annals of Fulda put it. Historians have understandably sought long-term explanations for it, mostly in the ‘rise’ and growing autonomy of major aristocratic families, for it was these who provided the new kings of 888, the ‘Robertine’ Odo of Paris in West Francia, the ‘Widonid’ Guy of Spoleto in West Francia and then Italy, Boso’s son Louis in Provence, the ‘Unruoching’ Berengar of Friuli in Italy, and the ‘Welf’ Rudolf, from Queen Judith’s family, in Burgundy. All these were however families very close to the Carolingians, linked by marriage in the last three cases (Louis and Berengar had Carolingian mothers). Only one of them, too, had any serious track record of disloyalty: Boso, who broke with the whole Carolingian tradition in 879 and declared himself king in the Rhone valley (he only lasted until 882 as king, for all the Carolingians combined against him). The others show no signs of seeking power on their own account until the 887-8 crisis itself, which forced them onto the centre stage.
What destroyed Carolingian power was simply genealogy. There had always been too many Carolingians, given the presumption of political division the family had inherited from the Merovingian past. Rulers had developed methods of excluding minor branches from succession, either by force (as with Carloman I’s son Drogo, or Pippin of Italy’s son Bernard) or by agreement (as with Adalard and Wala, who were content to be major players in their cousin’s court, or Bernard of Italy’s son Pippin, count of Beauvais, who effectively turned into a regional aristocrat; his heirs were the central medieval counts of Champagne), or through a growing concern to exclude illegitimate children. Even then, there were still a large number of them; as late as 870 there were eight legitimate adult male Carolingians, all kings or ambitious to become kings. In 885, however, there was only one. None of Lothar’s sons had legitimate male heirs; nor did Louis the German’s; Charles the Bald’s son Louis had three, but two were dead by 884 and the third, Charles ‘the Simple’, born posthumously, was only eight in 887. One by one, as the Carolingians died in the 880s, Louis the German’s last surviving son Charles ‘the Fat’, king of Alemannia (876-87, emperor 881) inherited their kingdoms, until he reunited the whole empire in 884 for the first time since 840.
Charles the Fat has had a bad press. This is and was linked to some over-pragmatic showings against the Vikings, as when Odo of Paris fought off a big siege in 885-6, but Charles paid them to go away; and is coloured above all by hindsight, for he was overthrown by his illegitimate nephew Arnulf in 887, a few weeks before his death in 888. Charles was more able than this implies. But everybody must have known that the world was likely to change, for Charles was ill, and himself had only an illegitimate son, Bernard. (Boso indeed must have seen it coming in 879: most of these genealogical problems were by then predictable.) Lothar II had spent most of his reign trying to legitimate his illegitimate son Hugh, and failing, as we shall see in the next chapter; Charles the Fat had no rivals, but even he could not make Bernard his legal heir. Hugh, who had visible royal ambitions, was caught by Charles and blinded in 885; this, and also Arnulf’s succession, means that Bernard could well have tried to succeed anyway (he did rebel against Arnulf, and was killed, in 891), but Charles did not change the rules fast enough to make illegitimate sons normal royal heirs. Instead, he tried in 887 to divorce his wife Richgard, as Lothar II had also tried, so that he could remarry and aim for legitimate sons; it was then that Arnulf, who had previously been kept away from central power on the Carinthian borderlands of eastern Bavaria, staged his coup and took the East Frankish throne. This coup made the decisions of the most powerful aristocrats of the other sections of the empire easier; Arnulf had some standing in West Francia, Burgundy and Italy, but his genealogical claims did not seem so strong to most political actors outside the eastern kingdom, and someone had to rule. When they did, they varied in their effectiveness; but they did not use most of the Carolingian political practices discussed in this chapter.
More important than the ‘rise’ of an aristocracy was its growing regionalization. This, paradoxically, was a reflection of royal power. Kings could confiscate benefices and offices, honores, and aristocrats feared this. We saw this in Einhard’s letters in the 830s; Nithard in the 840s is still clearer, for the whole of 840-41 was a phoney war in which Lothar and Charles prowled around each other trying to tempt followers from each other by promises, threats and an appearance of future success, which would be convincing enough to persuade worried aristocrats to tolerate losing honores temporarily in order to gain more later. Louis the German’s failed move into Charles’s kingdom in 858 was similarly structured. Each king who did this hoped for a catalytic change that would bring all a rival’s followers running in, as at the Field of Lies in 833; this seldom happened (887 is the only parallel), so what happened instead was usually that the followers of one king lost honores in the lands of the other. They were more likely to keep the land they held in full property, as Matfrid of Orléans did in the case of his family land in northern Francia when he followed Lothar to Italy in 834, or as a group of aristocrats in East Francia did in 861 when Louis the German abruptly expelled them from power. This land could remain very widely spread, as in the case of the ‘Unruoching’ Everard marquis of Friuli in Italy (the father of Berengar, future king of Italy, 888-924), whose will of 863-4, made with his wife Gisela, disposes between his sons and daughters of a book collection and rich treasures, but also estates stretching from Italy up through Alemannia to what is now Belgium. Such wide spreads favoured support for a single political system, as has already been noted for the ‘Widonids’. But Everard and Gisela gave at least their younger sons more geographically restricted territories; they also included explicit provisions for what might happen if political tension made it necessary to divide this land up further. The family regionalized itself as a result; Berengar’s brother Rudolf (d. 892) spent his career, not in Italy, but in Artois and on the English Channel. Similarly, the ‘Welfs’, whose lands lay both in Alemannia and in Burgundy, had to choose between Charles and Louis in 858; it may possibly be that those who chose Charles kept some of their properties in East Francia, but henceforth their careers would be entirely restricted to Burgundy, and their history became totally separate from that of their brothers and cousins who stayed with Louis. The tensions between the Carolingians, that is to say, persuaded prudent imperial aristocrats that it was sensible to have both their honores and their properties in one kingdom, not widely scattered as they had been since Charlemagne’s time. As kingdoms became smaller, this would become still truer.
Aristocrats always wanted to leave all their power-bases - fully owned properties, benefices, rights over monasteries, counties - to their sons. This was only assured for their properties, but already in Charlemagne’s time a loyal aristocrat could assume that his son might well inherit his county. The county of Paris, for example, was probably held by a single family between the 750s and the 850s; kings restricted themselves to choosing which heir took it over. The sons of Louis the Pious actually moved counts around more than their father and grandfather had, but all the Carolingians recognized that the sons of counts should normally get a county somewhere, and as the geography of practical politics contracted it might well be that this might be in or near their father’s county or counties. The sons of counts sometimes actually feuded against men who were given their father’s counties, as happened on the Bavarian eastern frontier in 882-4, admittedly a marginal and somewhat wild area. The memory of former power lingered too; Odo of Paris got some of his father’s Loire counties back in 886, a full twenty years after his father’s death - and very usefully timed, given his takeover of the West Frankish throne in 888. This further aided the process of regionalization. Odo’s father Robert had moved without difficulty from the Rhineland to the Loire in the 840s, when long-distance career moves were still normal, but the ‘Welf’ move to Burgundy in 858 was more controversial, and after that such shifts were rare, or else resented as the irruptions of outsiders. (Perhaps only Boso, who moved from Lotharingia to the Rhône valley and Italy, is a counter-example, but he was a queen’s brother, and anyway a mould-breaker in other ways too.) When Charles the Fat inherited seven separate kingdoms, separate political power networks visibly continued to operate in most or all of them; by now, it would have taken a Charles Martel-style war to unify them, and Charles the Fat did not have time for that. They went their separate ways again in 888. These were, genuinely, long-term causes for the break-up of the empire. They did not make that break-up more likely, but they made it possible, once the Carolingians died off. By then, a sense of empire-wide identity was attached only to the Carolingian family (and, not to be underrated, its army-muster). But aristocratic networks were prepared for a new regionalized politics; which was fortunate, for it was this which faced them now.