6
In October 680, Wamba, Visigothic king of Spain (672-80), fell seriously ill, and thought he was going to die. Like some other kings, he undertook penance, and was tonsured in the presence of his magnates; he designated his successor Ervig (680-87) in writing and in another document asked for him to be anointed as soon as possible (anointing to the kingship was in fact a novelty, introduced, as far as we can tell, by Wamba himself in 672). Wamba did not die; but he was tonsured now, and the sixth church council of Toledo (638) had prohibited anyone who had been tonsured from being king. Ervig quickly called the twelfth council of Toledo, which met in January 681, less than four months later, in midwinter, and as their first act the bishops of the kingdom ratified his succession and all the associated documentation (this is our only source for it, in fact), and cancelled the oath of allegiance the Spanish had sworn to Wamba. As their second act, they discussed what would happen if someone was given penitence and the tonsure while unconscious and, recovering, wished to reject it and return to a secular career: they enacted that the penitence and tonsuring must hold. Like most commentators, I see this as a response to a protest by Wamba that he had been deposed without his consent; but the careful legal framing of an effective coup is nonetheless striking.
The seventh-century Spanish political community were not always as respectful of the forms of law as this. The rules on legitimate succession laid down by the fourth council of Toledo in 633 were almost never followed, for example. But legal enactments, both secular and ecclesiastical, were part of the currency of Spanish political practice. People were aware of them, if they were aristocrats and bishops, at least; and even kings, if their support was weak enough, as was presumably the case in 680, could be trapped by them. This is a marker of a different style of politics from that of Francia: in Visigothic Spain, as to a lesser extent in Lombard Italy, legal principles were important points of reference, as they had also been in the later Roman empire, to which the Visigoths and Lombards were in some respects closer than were the Franks. In the case of Visigothic Spain in particular, historians have indeed often paid too much attention to law, for there are few narratives and documents for the period, and immense quantities of secular and ecclesiastical legislation. Spanish history often looks fairly arid as a result. But we cannot and should not argue that law away; its very quantity tells us something about the values of the Spanish establishment. I shall begin with Spain, move on to Italy, and then compare them; we shall then see better what sort of range of development from Roman practices was possible in the post-Roman West.
Spain (that is, the Iberian peninsula, including what is now Portugal) was partly conquered by the Vandals after 409, and then, after 439, mostly conquered by the Suevi. In 456 the Visigoths invaded and swiftly destroyed Suevic power, confining it to the far north-west. The obscure process of Visigothic conquest began here, speeded up in the 470s, and was probably complete by 483, when King Euric had the main bridge at Mérida, the Roman capital of Spain, repaired, as an inscription attests. The Visigoths were still based in Gaul, however; even after their great defeat by Clovis in 507 their capital remained in Narbonne, in the tiny strip of Mediterranean Gaul (modern Languedoc) that they kept hold of. After 511 Theoderic the Ostrogoth established a regency for the Visigothic child king Amalaric (511-31), and Spain was effectively ruled from Italy until Theoderic’s death in 526. There followed another forty years of relatively weak kings, succeeding each other by coup. Athanagild (551-68), based apparently in Seville in the south, rose up against Agila (549-54) and fought a civil war against him; he asked for Justinian’s help to do so, and this gave the east Romans the excuse to establish a bridgehead in Spain, the south-eastern coastal strip, in 552, which they held until around 628. Athanagild died in his bed, unlike any of his sixth-century predecessors; he was succeeded by Liuva I (568-73), who was again based in Narbonne, but who soon divided his kingdom with his brother Leovigild (569-86), giving the latter the whole of Spain and keeping only Visigothic Gaul.
The mark of the whole period 409-569 in Spain is instability. Perhaps in 483-507 there was relative peace, and also probably in 511-26, but in both periods the peninsula was ruled from outside, from Gaul and then Italy. The empire was not so long gone, when the western Mediterranean had been a single unit, but in our rare sources for this period Spain seems an appendage almost in a colonial sense, and largely left to its own devices. As we saw in Chapter 4, the archaeology for the later fifth century, particularly for the inland plateau of Spain, the Meseta, shows a weakening of rural estate centres, villas, and also a sharp contraction of the scale of ceramic production, which became more localized and simpler. The first of these developments, which became accentuated in the sixth century, might simply show cultural changes, as it did in the militarized northern Gaul of the late fourth century, but the second shows a simplification of the economy as a whole, which implies a decrease in aristocratic demand. The insecurity of the fifth and a great part of the sixth centuries, in some parts of the Iberian peninsula, seems to have hit many of the basic economic structures inherited from the Roman world quite hard.
The other effect of this instability was the fragmentation of the society of the peninsula. Spain is mountainous, with poor communications between the great plateaux and the major river valleys, and very great ecological differences between the wet climate of the north-west, which resembles Cornwall, and the desert of parts of the south-east. It would be easy for it to break into pieces with very different experiences, and this is what seems to have happened in this period. In parts of the north, we find references to semi-autonomous communities, either ruled by local strong-men like the senior Aspidius (575) in the Ourense area, or, more often, apparently collectively run, like the Sappi of Sabaria, perhaps near Zamora (573), or the hardly Romanized tribal groups of parts of the north coast who were generally called Vascones and many of whom spoke Basque. Such communities could have more Roman trappings, however, as was apparent in Cantabria (574), the Ebro valley upstream from Zaragoza, which was ruled by ‘senators’ (major local landowners) and a senate. In the south, it was cities that established autonomy, such as Córdoba (550-72). Southern cities could indeed remain very prosperous in an entirely Roman tradition, as is clear in Mérida, not a fully autonomous centre but for a long time hardly looking at all to the kings, whose bishops and aristocrats maintained considerable wealth (attested in the episcopal saints’ lives for the city), and where several Visigothic-period urban and rural churches and even some villas survive. There were thus two processes of fragmentation in this period. One was the loss to central authority of numerous sections of Spain, up to a third of the peninsula. The other was the development or revival of political practices that were different from those of Rome, more collective, even tribal, in some parts of the peninsula, notably the north. It must be stressed all the same that much of Spain remained very Roman, whether it obeyed the Visigothic kings or not, especially along the Mediterranean coast and in the rich Guadalquivir valley in the south, a zone which extended inland to Mérida. One of the Variae of Cassiodorus from around 524 shows the Ostrogoths taking the land tax, and a document surviving for Barcelona and nearby cities from 594 shows that taxation (in that area it was run by counts and bishops) could, at least locally, be quite high.
It was this doubly fragmented situation that Leovigild faced; he reversed it by military action. The dates in parentheses in the previous paragraph are those of Leovigild’s conquests, which were systematic in the 570s, and which culminated in the overthrow of his son Hermenegild’s five-year Seville-based revolt in 584 and the annexation of the Suevic kingdom in 585. By Leovigild’s death in 586, only the Roman-controlled coastal strip in the south and the Basques in the north remained outside royal authority. As with Charles Martel in Francia in the 720s-730s, the Visigothic power-base cannot have been so reduced, or Leovigild could not have managed this at all, however much more determined he was than his predecessors. It is clear from the Mérida saints’ lives that he wanted to make his power felt inside the lands he controlled as well. Leovigild appointed an Arian bishop, Sunna, to oppose the rich and locally influential Catholic bishop of Mérida, Masona (who was himself a Goth), and eventually summoned Masona to his court at Toledo and exiled him for three years. He exiled and expropriated lay aristocrats, too; and, not least important, he issued a major revision of the law code. Leovigild was not simply a soldier; he was a unifier. Toledo had already become the main royal residence under Athanagild, but under Leovigild it became a focus of political and religious activity, a real capital. The choice of Toledo, not previously a major centre, was itself significant, for it was exactly in the middle of the peninsula: it marked royal ambition. Leovigild founded his own new city, too, Recópolis, to the east of Toledo, as a further sign of prestige, although Recópolis was never very large, as excavations show.
Leovigild also faced up to the problem of religious disunity. The Goths in Spain had remained Arian; Leovigild in a church council at Toledo in 580 sought to soften that Arianism doctrinally, to make it more palatable to Catholics, while also persecuting at least some Catholic activists. This has parallels to Vandal procedures in Africa a century earlier, but the attempts to find a doctrinal middle road more resemble the policies towards Monophysitism of eastern emperors such as Justinian, as we have seen. Essentially, however, Arianism was practised by too few people by now; the Goths were only a small proportion of the population of Spain, a few per cent at most, and not all of them were Arian, as Masona shows. Hermenegild, too, adopted Catholicism in the course of his revolt. Once religious unity came to be seen as desirable, it was most likely to be on Catholic terms. Indeed, Leovigild’s second son and successor, Reccared (586-601), switched to Catholicism almost immediately after he became king, in 587, and at the third council of Toledo in 589 Arianism was outlawed, far more uncompromisingly than Leovigild had sought to oppose Catholicism. Reccared faced a series of revolts and conspiracies as a result, up to 590 at least and perhaps longer. But Arianism must have been weak by now, for it did not reappear as a rallying call in the renewed instability that followed Reccared’s death.
Reccared’s son Liuva II (601-3) did not last long, and between 601 and 642 there were nine kings, only one (Suinthila, 621-31) lasting as much as a decade; three were sons of their predecessors, but they were particularly swiftly overthrown. Fredegar in Francia referred to this constant series of coups rather smugly as the ‘disease of the Goths’ - to a Frank, of course, non-dynastic kingship looked like chaos in itself. But what did not happen in this generation was any reversion to the political fragmentation of the pre-Leovigild period. The kings fought frontier wars, against the Basques, the Franks and the east Romans on their coastal strip, and Suinthila finally conquered the latter region around 628. Internally, the sequence of coups at least shows that the dukes and provincial governors of the kingdom were interested in central kingship, rather than autonomy. The kings themselves, even Suinthila, did not leave much mark; Sisebut (612-21) was an author of poetry, letters and a saint’s life, the only western ruler in this book except Alfred of England to gain a reputation as a writer, as well as being the first serious persecutor of the Jewish population of Spain, but he seems otherwise undistinguished. The only major innovation of this period was the inauguration, with the fourth council of Toledo in 633, of a steady series of plenary councils of bishops, called by kings at Toledo - thirteen from 633 to 702 - which became so crucial a part of the political aggregation of the kingdom that periods without regular full councils, notably 656- 81, were sharply criticized by the church, even if provincial councils had been called in between. The collective role of bishops in the political aggregation of the seventh-century Visigothic kingdom was a specific feature of Spain; neither Francia nor Lombard Italy put as much weight on church councils. Their legislation was secular as well as ecclesiastical, and the king presided, often reading out an initial statement of intent. They contributed greatly to the ceremonial importance of the capital.
The cycle of coups was broken by Chindasuinth (642-53), who took over the throne at the age of nearly eighty, and who curbed the aristocracy by executing 700 of them (Fredegar claims), depriving others of their civil rights, and enacting a draconian law on treason. Chindasuinth was hated for this even by some of his protégés, such as Bishop Eugenius II of Toledo (d. 657), who wrote an abusive epitaph for him. Feelings remained sufficiently strong that once a king succeeded who was in a weak position, Ervig in 680, the thirteenth council in 683 restored the noble status and civil rights of all those who had lost them since 639: aristocratic (and episcopal) solidarity had kept the issue alive for forty years. But conversely the coups ended, or, perhaps better, remained provincial and no longer succeeded at national level; so Reccesuinth (649-72) defeated Froia in 653, Wamba defeated Paul in 673, Egica (687-702) defeated Sisbert in 693. Royal succession became peaceful, even when controversial: Reccesuinth was Chindasuinth’s son; Wamba was elected at Reccesuinth’s deathbed; Ervig’s succession was at least uncontested; his successor Egica was his son-in-law, and Wittiza (694-710) was Egica’s son. Only in 710 was there a contested election, perhaps a coup, with Roderic (710-11) imposed by court officials. This general tendency away from political violence was not lessened by the clear evidence we have that most of these kings were opposed to their predecessors. Ervig with respect to Wamba is one example; Egica with respect to Ervig is even clearer, for at his accession he asked the fifteenth council to let him dispossess Ervig’s family (the council refused). Both Ervig and Egica also took some pleasure in reversing their predecessors’ laws. Wittiza apparently cancelled his father’s expropriations too, and Roderic was later thought to have been opposed by Wittiza’s family. Tensions thus evidently remained, and they could be savage (particularly under Egica), but they were patterned by ceremonies of public solidarity and legislation, not by war.
The last half of the seventh century marks the peak of public activity for the Visigothic kingdom. Reccesuinth and Ervig both revised Leovigild’s law code, and legislated substantially themselves; laws survive for all the other kings except Roderic. The church councils were key moments in royal policy-making as well. And the laws that were made were more and more complex, as well as more and more high-flown. They were posed in all the codes as Gothic law, valid for all people in the kingdom, as law had probably in fact been from the fifth century onwards, even when a distinction between the Gothic and Roman population could be drawn, something which had gone by the mid-seventh century. But the antecedents of much of this ‘Gothic’ law lay in the imperial code of Theodosius II, far more than in other post-Roman kingdoms, and the rising rhetoric of the law looked to Roman models too. It is fairly clear that the late seventh-century Visigoths had the contemporary Byzantine empire as a point of reference as well, at least as a model for ceremonial, and for a close identification between the episcopacy and the king. The importance of religious conformity, implicit since the third council in 589, also became increasingly visible. The major law-givers of the period, Reccesuinth, Ervig and Egica, were fiercely hostile to the main non-Catholic group in Spain, the Jews; they picked up Sisebut’s laws and greatly extended them, banning all Jewish religious practices, restricting Jewish civil rights, and in 694 reducing all Jews to slavery. The seventh century in Byzantium, Italy, even the normally tolerant Francia, saw some sporadic Jewish persecution, but these Visigothic laws have no real equivalent in their violence - and violence of expression - against Jews until the late Middle Ages. It is hard to read them today without hostility and alarm. All the same, they are quite parallel with Roman heresy laws, and they are in a line of legislation which in that respect stresses the Romanizing ambition of the kings only too clearly.
It is at this point, however, that questions arise. The complexity of the ceremonial at Toledo is very evident by 650 or so, and the regularity with which bishops and aristocrats went there is equally clear. The elaborate public humiliations which political losers faced in the capital - Argimund in 590 taken through the streets on a donkey with his hand cut off, Paul in 673 brought in barefoot on a camel - look straight back to the victory ceremonies of the Roman empire. Kings were, as in Francia, rich, not least because of Chindasuinth’s confiscations (they maintained elements of the land tax, too, into the late seventh century at least), and therefore such a focus on the capital was presumably considered profitable by political players. The administration, the officium palatinum, was at least as elaborate as in Francia - although far less than in the Roman world - and hedged about with legal privileges; it included central officials and regional representatives such as dukes, and seems to have had some corporate identity, presumably centred on the king, much as the episcopate did. Indeed, it has been plausibly argued by Dietrich Claude that the aristocracy were, as a whole, more and more involved in palace politics; and the kings could certainly ruin individual aristocrats if they chose. But our sources are so overwhelmingly interested in royal and episcopal aspirations, and tell us so little about what really went on outside Toledo, that it is legitimate - and common - to wonder how much of this legislation was shadow-play. The Jews were so often extirpated, then return to be extirpated again. Wamba’s 673 law on army-service was so severe, Ervig claimed when revising it a decade later, that ‘almost half the population’ had lost their civil rights: do we believe this? Egica in 702 in a law against fugitives said that ‘there is barely a city, fortification, village, estate or dwelling-place’ in which they were not hiding. This is a law which has been taken literally distressingly often, but it at least shows both the tendency of the kings to get carried away rhetorically and their awareness that it might be very different on the ground.
When we get a sight of local realities, they often seem very variegated as well, just as they had been before Leovigild. The archaeological trends of the fifth and early sixth centuries were not reversed later; if anything, they were accentuated, with the Meseta showing an increasingly localized set of economies, imports dropping in the Guadalquivir valley, and much of the Mediterranean coast showing a sharp economic simplification in the seventh century; the Roman south-east coast was no longer supplied from North Africa after Suinthila’s conquest, and it seems to have gone into crisis. Urbanism survived best in some of the southern cities, Mérida, Córdoba, doubtless Seville, and also Toledo in the centre (the latter two have not been excavated, however); much less in most of the north, and only occasionally on the Mediterranean coast (Barcelona and Valencia are candidates). In economic terms, the seventh-century kings thus presided over a set of separate economic realities, with divergent histories and decreasing interconnection. The seventh-century slate documents that have been found in the central mountains south of Salamanca (it is an area with slate rocks, easily usable for writing; the texts are often quite ephemeral estate texts, lists of cheese-rents and animals) seem to reflect a very localized economy as well: they cite very few place names, except, once, Toledo.
This growing local divergence may also explain some of the inconsistencies we can see in social trends. The aristocracy was clearly as militarized as in Francia, and a pattern of private relationships was developing; the late seventh-century army laws show that the army was largely made up of the personal dependants of lords, and church council legislation shows that the image of personal dependence was coming to structure ecclesiastical hierarchies too. The king, indeed, was seen as everyone’s lord; every free man swore a personal oath of fidelity to him, a practice borrowed later by the Franks and Anglo-Saxons. Conversely, we also find institutions and cultural attitudes that were hardly changed from the Roman empire. The obsession with law and with legal delimitation (between aristocrats, honestiores, and non-aristocrats, for example) seems likely to be a Roman survival, even if some of the rhetoric of kings like Ervig and Egica could be seen more as revival. And, above all, the dense Roman culture of major political intellectuals like Isidore, bishop of Seville (599-636), author of theology, history, and the Etymologies, an influential if very strange encyclopedia, as well as animator of the fourth council of Toledo, must show that a traditional educational structure had survived intact in some of Spain’s major cities. The letters of Isidore’s disciple Braulio, bishop of Zaragoza (631-51), which are unusually attractive and human, show that this Roman cultural style existed in other parts of Spain as well, and the letters of Count Bulgar, surviving from the 610s, show that it sometimes extended to the secular aristocracy too, as King Sisebut’s writings further demonstrate. Isidore and Braulio were in any case heavily involved in secular politics; they were both from aristocratic episcopal families, and were very close to kings. Their ‘late late Roman’ political practice, which survived in their successors up to 700 at least, must have been recognizable to a substantial part of the political establishment, and was certainly drawn on by legislators; Braulio indeed seems to have personally contributed to Reccesuinth’s revised law code.
The seventh-century Visigothic kings thus presided over places and social groups where not very much had changed since the days of Augustine, places and social groups characterized by the same sort of militarized - and ruralized - society as in contemporary northern Francia, as well as some much simpler, more collective, societies, surviving in particular in parts of the north, and some areas of economic disintegration on the Mediterranean coast. They handled this diversity with the ambition of Roman emperors, but with a rather less elaborate administrative structure, which would have made detailed intervention rather more difficult. Small wonder their laws were sometimes rather shrill. It was impossible to encompass this diversity with early medieval western means; the kings knew it, and, unlike in Francia, resented it. But we would be wrong to follow the view of some modern historians and conclude that the late seventh century was a period of general crisis for the kingdom. Far from it; in that period the Visigothic state was the strongest in the West.
One of the reasons why the imagery of crisis has been used is that in 711 the Visigothic kingdom was overthrown by an Arab and Berber army invading from North Africa, and most of Spain remained part of a Muslim political community looking to Damascus, Baghdad and Cairo as a result, for the next five centuries and more (see Chapter 14). When kingdoms collapse quickly, historians have often sought to blame them for their defeat; but the answers can just as easily lie in the chance of a single battle, as with the Anglo-Saxon kingdom of England that ended at Hastings in 1066. It is certainly true that Spain fell to pieces in 711. The Arabs were for long only powerful in the far south. The north-east kept a Visigothic king for a decade; the south-east saw its Visigothic governor, Theodemir (d. 744), cut a separate deal with the Arabs in return for autonomy; the far north returned to communitarian and sometimes tribal traditions, as well as choosing an independent Christian king in the Asturias, Pelagius, around 720, the first of a long series of independent kings in the north (see Chapter 20). These different choices certainly reflect the socio-economic divergences already cited. But it took violent conquest to turn them into political realities; before 711 there is no sign of the sliding away of outlying regions, as in late seventh-century Francia. Until then, as far as can be seen, the Visigothic kings kept a firm hegemony over all of them.
Italy had even more Roman traditions than Spain to draw on in the sixth century, but handled them differently. The Gothic war of 536-54 did enormous damage to the infrastructure of the peninsula, and Italy had by no means settled down when a federation headed by the Lombards invaded from Pannonia under Alboin (560-72) in 568-9. The Lombard invasion was one of the more disorganized we know of, however. In 574, after the assassination of two kings, the Lombards abandoned kingship altogether for a decade, and operated as a loose federation of dukes. It is likely that they did this as a result of bribes by the east Roman ruler Tiberius II, and Tiberius and his successors were indeed successful in getting many dukes to fight on the Roman side in the wars of the rest of the century. Tiberius also invited the Franks back to Italy to attack the Lombards. The Lombards, facing this, elected kings again, Authari (584-90), and then, in the teeth of the most substantial Frankish invasion, Agilulf (590-616), formerly duke of Turin. Agilulf withstood the major attacks of 590 and counterattacked himself; he established peace with the emperor Phocas in 605, gave tribute to the Franks, and some stability could return. But that peace revealed an alarmingly divided Italy. The Lombards had not managed to conquer more than separate sections of the peninsula: the inland Po valley in northern Italy; Tuscany, connected to the north only by a single mountain pass; and the duchies of Spoleto and Benevento in the central and southern Appennines. The last two were effectively autonomous duchies, with little link to the kingdom of the north until the eighth century. The Romans clung on to the area around Ravenna in the north, extending along the Adriatic coast in both directions, the west coast around Genoa and Pisa, the area round Rome, the area around Naples, and Puglia and the far south, with Sicily and Sardinia. Italy would not be controlled by a single ruler again until 1871. In this patchwork, the old centres of Roman political power, Rome and Ravenna, and all the major ports, stayed out of Lombard hands, and the Lombards were essentially restricted to inland areas, which had already become structurally separate from the Mediterranean world since the Gothic war.
This division could have allowed the Roman parts of Italy - Byzantine Italy as we can now call it - to maintain imperial traditions without further problem, while leaving the Lombard lands in relative isolation. This did not happen, however. The Lombards tended to remain on the offensive, taking Genoa in the 640s and Puglia in the 670s; another peace in 680 stabilized matters a little, but between 726 and 751 the Ravenna area was taken in a series of wars. Lombard isolation from the coast was steadily eroded, and after 751 Byzantine power on the mainland was restricted to Venice, Rome, Naples and the tips of the heel and toe of Italy. This meant that the Byzantines had to remain heavily militarized to defend themselves, and they did so. The ruler of Byzantine Italy, the exarch based in Ravenna, held a military office, and the aristocracy rapidly reshaped itself into a military hierarchy looking to him; even the citizens of the Byzantine cities could be referred to as milites, soldiers, or as a numerus, an army. They looked more and more like a ‘Romano-Germanic’ society, in fact, whether Lombard or Frankish. Unlike in the rest of the Byzantine empire, even the tax system eroded in much of the peninsula, much as it did in Francia or Spain. And links with Lombard areas, wars or no wars, slowly developed. The northern Lombards had to cross Byzantine territory if they wanted to reach Spoleto or Benevento, and are sometimes mentioned in sources as visiting Ravenna; the Beneventans and the Neapoli tans even shared ownership of public lands in the rich Capua plain on their boundary. In social terms, the various sections of the peninsula developed largely in parallel, and we do not see the divergencies in Italy that are visible in Spain. We shall look shortly at some of the differences that did exist.
The Lombard kings, like the Visigoths, never established a dynasty; even the Agilolfings who ruled from 653 to 712 faced internal coups, and two kings from a rival family in 662-72. But throughout the seventh century they recognized some rough genealogical criteria for succession, if not in the Agilolfing line, then through queens. Authari’s widow Theodelinda (d. c. 620) married Agilulf in 590, and later tradition said she chose him; she was certainly influential in Agilulf’s reign, negotiating with Pope Gregory the Great (590-604) in Rome. Her daughter Gundiperga similarly married two successive kings, Arioald (626-36) and Rothari (636-52); Aripert I (653-61) was her cousin; and Grimoald (662-71), when he overthrew Aripert’s son Perctarit (661-2, 672-88), married the latter’s sister. This did not, all the same, lead to much female political protagonism; Theodelinda remains an exception here, perhaps because she was a Frank, daughter of the Agilolfing duke of Bavaria, at a time of considerable Frankish influence.
The seventh century is poorly documented in Lombard Italy, in the absence of both documents and detailed narratives, for Paul the Deacon’s Lombard history, written in the 790s, is both brief and late; but Agilulf and Rothari stand out. Agilulf stabilized the frontiers, and also established an effective hegemony over the dukes of the cities of the north. The political incoherences of the first thirty years of Lombard Italy ended with him. He used Roman ceremonial imagery, as when he presented his son Adaloald (616-26) as king in 604 in the circus at Milan, and he had Roman administrators and advisers. His wife and son were Catholic, but he was not. All the same, it is significant that we cannot tell from our sources whether he was pagan or Arian; the Lombards included followers of all three religions, and there is no sign from Agilulf onwards that personal religious affiliation had a major political content, unlike in Spain. Rothari, duke of Brescia before his accession, was certainly Arian, but more important than that is that he saw himself as a Lombard legitimist, fighting wars against the Byzantines, and he issued the first Lombard law code, the Edict, in 643. This text lists his seventeen predecessors as Lombard kings, well back into myth, and also Rothari’s own eleven male-line ancestors, and manuscripts of the law include a brief Lombard history, which may have been there in some form from the start. The Edict of Rothari is the longest early medieval code after those of the Visigoths, but much less influenced by Roman law, although the picture of royal authority contained in it is Roman enough. It was really Rothari who created a specifically Lombard imagery for kingship and society in Italy, and there is little in later Lombard ‘ethnic’ identity that can be traced back further than 643. Conversely, it is important to recognize that, as in Spain, this identity was erected on the back of a Roman-influenced administration, based from Rothari’s time at the latest on a stable capital at Pavia, in imitation of Roman/Byzantine centres such as Ravenna (and maybe also of Toledo), as well as on a network of dukes and gastalds (the equivalent of Frankish counts) ruling over each of the traditional Roman city territories of the north of Italy.
Rothari’s successors drew on Lombard imagery, and on Lombard law, but also on the Roman infrastructure that it assumed. They also used the church relatively little; bishops were important in city politics, and are sometimes referred to as royal advisers (under Agilulf in particular), but none of them were major political dealers, unlike in Francia, and councils of bishops had no political or ceremonial role, unlike in Spain, or in the Byzantine empire. After 653 no king was certainly Arian, but the abandonment of Arianism is given little stress in our sources. Slightly more important, perhaps, was the formal abandonment at the synod of Pavia in 698 of the schism of the Three Chapters, which had separated the Catholics of the north of Italy from Rome since the 550s, under the patronage of King Cunipert (679-700), but this did not lead to any increase in the imagery of religious unity in the Lombard kingdom, either.
Liutprand (712-44) was the most powerful Lombard king. Son of the tutor to Cunipert’s son, he could claim a link to the family politics of the seventh century, but he was not genealogically associated with his predecessors, and his reign feels like a new beginning. He legislated extensively to fill out and update Rothari’s Edict, in annual sessions (taking place on 1 March, as in sixth-century Francia); it is clear that he was also regularly acting as a judge, for many of his enactments are the generalizations of specific judgements on quite arcane points of law, such as who is liable if a man is killed when the counterweight from a well falls on his head while water is being drawn, or how much penalty should be paid if a man steals a woman’s clothes while she is bathing. One of his first enactments, in 713, made pious gifts to the church legal, and documents for such gifts and for other matters more or less begin then, making the eighth century as a whole much more visible than the seventh in Lombard Italy. And he made war, almost as regularly as his contemporary and ally Charles Martel, against the Byzantines and also against the southern Lombard dukes. By the 740s Spoleto was permanently brought into the political power-structure of the kingdom. Benevento, further away and richer, had always been the more autonomous of the two (except in the 660s, when its duke, Grimoald, had gained the kingdom), and remained so, but at least Liutprand and his successors chose its dukes several times. By Liutprand’s death the Lombard king was hegemonic in the entire peninsula, and it became for the first time since 568 conceivable that Italy might become a single political unit again.
Liutprand’s successors were the brothers Ratchis (744-9) and Aistulf (749-56), dukes of Friuli in north-east Italy. Both kings legislated, and Aistulf in particular followed Liutprand’s territorial policies. It was Aistulf who finally occupied Ravenna in 751, and in 752 he sought tribute from Rome. But the geopolitical situation had changed by now. It was in 751 that the Carolingian Pippin III claimed the kingship in Francia, and sought ratification by two popes (below, Chapter 16); the debt to the papacy that this represented was quickly called in, as Pope Stephen II appealed for help against Aistulf. Pippin invaded Italy twice in 754-6; he forced Aistulf to leave Rome alone and to hand Ravenna to the pope as well. The next king, Desiderius (757-74), inherited both Aistulf’s aspirations and his constraints; he interfered in Roman politics, and also in Benevento, whose duke, Arichis II (758-87), he chose, but the Frankish threat remained. In the end, Desiderius attacked Rome again in 772, and Pippin’s son Charlemagne invaded Italy in 773-4; this time he overthrew the Lombard king and took all of Lombard Italy for himself, except Benevento, where Arichis in 774 named himself an independent prince.
The mid-eighth-century kings were trapped between their felt need to absorb Rome, as the key to the south, and the certainty of Frankish retribution, even if it has to be added that Italy’s accessibility across the Alpine passes probably means that Charlemagne would have eventually attacked anyway. The Franks were never safe neighbours, and had a history of Italian involvement going back to Theudebert; from the time of Ratchis onward, in particular, the Franks were also keen to welcome Italian exiles, as were the Bavarians. Lombard military activity was probably always on a smaller scale than in Francia, and we have several wills from landowners about to go to war. These hint that actually taking part in fighting was by now not routine, even for large owners, despite all the military imagery that the Lombard aristocracy, just like all their neighbours, now regarded as de rigueur - let alone for the lesser free, who were nonetheless referred to as ‘army men’ in legislation. But there is no sign of political or structural weakness in the Lombard political system in any other respect. Like the Visigoths in 711, they just lost to superior arms, in this case to the strongest army in western Europe. Lombard political practice, indeed, influenced that of the Franks in the next generations, as we shall see later.
The documentation we have for the eighth century shows a Lombard state that intervened in local society in capillary ways. The kings and the dukes or gastalds of the cities remained regular judges for primary court cases and for appeals, and kings made sure their judgements were followed by sending written instructions; we have a case from Lucca in 771 in which the local bishop re-heard a church dispute because the king had instructed him that his first judgement was improperly made. In difficult cases the king sentmissi, emissaries of the royal court, to make enquiry on the spot, as in the disputes between Parma and Piacenza over the boundary of their city territories, resolved after an inquest by Perctarit in 674, and the parallel dispute over diocesan boundaries between Siena and Arezzo in 714-15, resolved after two inquests by Liutprand. It was normal for quite ordinary people to go to Pavia to seek justice, or to Spoleto or Benevento, for which we have similar inquests and judgements. The inhabitants of the Lombard lands were also well informed about royal legislation, which gets cited in documents, even in the duchy of Benevento, unlike in Francia. Writing was an important basis for government. There is relatively little evidence in Lombard Italy for the large-scale ceremonial in the capital that is so visible for Spain, however. It seems that the centrality of Pavia was made easier because of two main features of Lombard society. First, the élites of the kingdom were very largely city-dwelling. They lived in one place, they competed over who was to be duke/gastald or bishop, they regularly attended the courts of both; they were loyal to their cities, indeed, as the boundary disputes mentioned above demonstrate. Even monastic foundations, which begin in the eighth century, were with some prominent exceptions urban. Whereas Frankish historians followed the factional politics of major dealers like Leudegar of Autun, Paul the Deacon, when he described the civil war following the coup of Alahis duke of Trento against Cunipert (c. 688-90), saw it in terms of the political choices of the citizens of Brescia, Pavia, Vicenza, Treviso. All of this meant that local élites were easily accessible, for all political practice took place inside cities, or nearly all.
The second major feature is that most Lombard aristocrats were fairly restricted in their wealth. Almost none of our documents show any of them with more than between five and ten estates, which is close to a minimum for aristocrats in Francia. The king and the ruling dukes of the south had immense lands, of course, and a small number of powerful ducal families, particularly in the north-east, were rich, but the bulk of the élite owned only a handful of properties, usually only in the city territory they lived in, plus perhaps its immediate neighbours, with, quite often, a house in Pavia. This meant that they could not afford the private armed entourages that were the support for factional politics in Francia; it is not chance that nearly all the usurpers in Italy, successful or unsuccessful, were dukes, who had a right to control local armies. It also meant that they would be satisfied by relatively modest gifts by kings, and indeed as far as we can tell royal generosity was not huge in the Lombard period, although the royal treasury was imposing, in Italy as in Francia. Aristocratic identity was also bound up with office-holding, which was in the king’s gift; duchies did not become family patrimonies, except for Spoleto (sometimes) and Benevento. The Lombard kings did not tax, after the first couple of generations of their rule at least. They operated entirely in the framework of a political practice based on land. But inside that framework, their hegemony was very great, and unusually detailed: their capillary power arguably extended to much more modest levels of society than the Frankish or Visigothic kings achieved.
The cities of the Lombard kingdom, despite their social and political importance, were in material terms not particularly striking. They were full of churches by 774, most of them recent foundations by urban notables - Lucca, the best-documented city in Italy, had at least twenty-five - but urban housing was materially nondescript, and commercial exchange for anything except luxuries was local at best. We know less about the duchies of Spoleto and Benevento; the high-mountain core of Spoleto meant that its cities were rather weaker, and its aristocracies more often rural, but Benevento had some rich lowland areas, and the capital there seems to have been a focus for an aristocracy that owned more widely than anywhere else in Italy; Benevento may well have been quite rich and politically coherent. But it was Byzantine cities in Italy, at least major ones like Rome, Ravenna or Naples, that were probably the most economically active. The archaeology for cities like Naples is certainly more impressive - or less unimpressive - than that for Lombard cities, Brescia or Verona or Milan. It was only in the last decades of the Lombard kingdom that even churches, usually the only surviving buildings of the early Middle Ages to show a real monumental aspiration, begin to be architecturally ambitious, as with Desiderius’ prestige monastery of S. Salvatore (later S. Giulia) at Brescia. By contrast, Naples and Ravenna, and above all Rome, could sustain that ambition throughout, and in the Byzantine lands it extended even to private housing, as documents show for eighth-century Rimini, and as recent excavations show for Rome.
Byzantine society in Italy had developed parallel to Lombard society, but it did have some particular features. It was broadly richer and more complex, as just implied. In Byzantine Italy the church was also more of a political protagonist: most obviously in the case of the pope in Rome, but also in Ravenna and Naples, where bishops were major figures. Another difference is that the separate Byzantine provinces of Italy moved towards effective independence in the eighth century, just at the moment when the Lombard lands gained some political coherence. The duke of Naples, Stephen II (755-800), became entirely autonomous from Constantinople (interestingly, he ruled Naples first as duke, and then, after 767, as bishop). By the 740s the dukes of the small lagoon islands crystallizing as Venice were effectively autonomous too; and that decade was probably the key moment in the century-long shift towards independence in the Rome of the popes, which was complete by the 770s. Nostalgia for Byzantine rule could remain; it was very much felt in Istria, taken by the Franks from Byzantine/Venetian control in the late eighth century, as a court case from 804 against the Frankish governor shows (see below, Chapter 16). But Italy was spinning away from Byzantine domination. The only major exception to this was Sicily, stably in imperial hands until the 820s.
Rome remained the least typical city in Italy. Although far smaller than it had been under the empire, it remained by a long way the largest city in the West, maybe twice the size of Ravenna or Naples, and five times the size of Brescia or Lucca (these figures are bald guesses, however). Rome’s territory, roughly the modern region of Lazio, was also much bigger than that of other city-states like Naples or Venice. The popes had always been major players in religious matters, and remained so - although their political-religious interests for a long time remained focused on the East, and they had almost no influence in Merovingian Francia and Visigothic Spain. But when the senate of the city faded out in the late sixth century, the popes emerged as the authority best equipped to rule Rome, as is already visible in the extensive letter collection of Pope Gregory the Great in the 590s. The eastern emperor could still remove a religiously rebellious Martin I in 653 (see Chapter 11), but could not remove Sergius I in 687 (the imperial envoy supposedly had to hide under the pope’s bed to escape the Roman crowd), and in the eighth century the entire imperial infrastructure in Rome steadily became papal. But the wealth of Rome, and of the popes themselves, meant that this infrastructure (and associated ceremonial) remained remarkably elaborate, with dozens of officials in separate hierarchies: far more elaborate than the government in any of the Romano-Germanic kingdoms, and indeed imitating that of Constantinople itself. In the eighth century popes like Gregory II (715-31) and Zacharias (741-52) consolidated papal power inside Lazio; Stephen II (752-7) and Hadrian I (771-95) acted as political protagonists, calling in the Franks against the Lombards, and in Hadrian’s case acting as a regional player, whom Charlemagne treated as a (near) equal. The papacy remained fairly marginal to western European politics for some centuries more, but its more strong-minded occupants could achieve quite an effect, as Nicholas I (858-67) would in Frankish and also Bulgarian affairs. In Italian politics, Rome’s size ensured that the popes would continue to punch above their weight, too; and popes acted as a legitimating element for Carolingian and post-Carolingian rulers, as we shall see later.
Visigothic Spain and Lombard Italy show two coherent alternatives to the Frankish path away from the Roman empire and into the early Middle Ages. Around 700, indeed, Spain looked more successful than Francia, though Spain’s conquest by the Arabs and Charles Martel’s reunification of the Frankish lands in the 710s and later have often led modern historians to conclude otherwise. Italy’s government, too, was effective enough to be a model for the Franks after 774. These three states show sharp divergencies in their political style, in the force of royal ceremonial (strongest in Spain), in the importance of dynastic legitimacy and in the wealth of local aristocracies (strongest in Francia), in the complexity of the links between central government and provincial society (arguably strongest in Italy). Royal aspirations were different, too: only the Frankish kings sought political hegemony over other peoples; only the Visigothic kings sought to rule like Roman emperors. But there are other aspects in which their developments were similar. They all moved towards social and political hierarchies dominated by military identity; civilian aristocracies vanished. (This happened in the Byzantine empire as well, first in Italy, but eventually even in the Byzantine heartland.) Steadily, at different speeds, they lost control of tax-raising, and became essentially land-based political systems, although all three managed to keep aristocratic political practice and even identity firmly concentrated on royal courts. Indeed, even though all three experienced periods of royal weakness and political fragmentation, successful rulers could in each case re-focus the aristocracy on them, Leovigild after 569, Agilulf after 590, Charles Martel after 719. All three also saw their political identity in ethnic terms, as Franks, Goths and Lombards, but ethnicity rapidly became unimportant in practice: by 700 most ‘Franks’ had ancestors who had been Roman, and the same is true for Spain and Italy. Indeed, apart from the continuing importance of assemblies (above, Chapter 4), and the assumption that military service was due from all free males, at least in theory (never in practice), there was not so much that was specifically Germanic in the ‘Romano-Germanic’ kingdoms. Politics, society and culture had moved on from the Roman world, but they can most usefully be understood as products of development from Roman antecedents.