Chris Wickham, Rome's western successors 500–750
From tax to land, declining complexity, public power legitimacy, solidly governed world
Chris Wickham wrote:
Why did the Roman empire fall? The short answer is that it didn’t. Half of the empire, the eastern half (what is now the Balkans, Turkey, the Levant, Egypt), which was ruled from Constantinople, carried on without problems during the period of imperial breakdown and conquest by outsiders in the western half (what is now France, Spain, Italy, north Africa, Britain) in the fifth century; the eastern empire indeed survived even the massive assaults on it in the seventh century … East Rome – from then on we call it the Byzantine empire, although its inhabitants called themselves Romans until its end – continued another thousand years, until the conquest of its last remnants by the Ottoman Turks in the fifteenth century; and then the Ottomans used some of the basic fiscal and administrative structures of the Roman/Byzantine past in their own state-building, focused on their new capital in Constantinople, now Istanbul.
I make this point not to conjure up the image of a past which never changes; there are always elements of the past in the present, but that does not mean that huge alterations have not taken place…
… However Romanised the first century of ‘barbarian’ kingdoms were, some crucial things had indeed changed, and would never, as it turned out, change back. The first is that the Germanic peoples did not call themselves Roman. They clearly saw themselves as separate from the Romans they conquered and ruled, and were in this respect quite unlike any warlords and coup leaders of the past …
… The second shift is that the old unity of the west, everywhere from Hadrian’s Wall to the Sahara, had vanished for ever. …
The third major change was arguably the most important. The Roman empire was governed by a complex bureaucratic structure, paid for by a sophisticated fiscal system, which involved many taxes, but above all a complex and heavy land tax. This system worked, even if it was corrupt, unpopular, and prone to abuse; we have much legislation from emperors who were concerned that the traditional tax collectors, city councillors, were not doing their job, which might imply that tax was not effectively collected, but it was certainly carefully controlled and policed – we have, for example, records from Italy and Egypt of the systematic recording of land transfers, so that the state would be able to tax the new owner accurately, and Egyptian documents also show that even rich and powerful landowners did indeed pay their taxes. This fiscal system was largely used to pay the army, easily the Roman state’s greatest expense (the civilian bureaucracy came a distant second), and this meant that money and goods regularly moved north across the Mediterranean from rich southern provinces such as Africa and Egypt, to the northern frontier regions where armies mostly were located, plus to Rome and Constantinople, capital cities which were kept large for symbolic reasons and largely fed by the state. The paid army was partially separate from the other major set of élites, the imperial (senatorial) aristocracy and the provincial and urban leadership of every part of the empire, who were landowners above all else, and also civilian
The fiscal system thus underpinned the whole Roman state, and was not at any risk at the start of the fifth century. When … the empire in the west was divided into kingdoms, tax revenues ceased abruptly to be moved around, with serious effects on Rome as a city, and on many northern armies.
Furthermore, the new Germanic élites had different aims from the rebellious Roman senior officers of the past. The latter had mostly just wanted higher pay, to go with their claims for political power; but their Germanic successors wanted something different: to be landowners, like the provincial élites they were now dominating and living beside. … It became less and less necessary to pay the now landed army. Tax régimes themselves became less necessary as a result, and since they were both disliked and complex to collect, could eventually shrivel away.
‘Barbarian’ kings continued to tax as long as they could, it is true … and kings can be seen granting tax immunities by then as a standard political privilege; by the 640s the land tax hardly existed any more in Gaul except sporadically … Kings began to rely on the revenues from their own lands … rather than fiscal receipts, except for tolls on commerce. The whole economic basis for political action shifted, from taxation to landowning. This marked a break, not just from the past, but also from the contemporary states in the eastern and southern Mediterranean … The break will also underpin much of the rest of the book, for, as we saw [previously] a land-based politics is less stable and normally less remunerative than a tax-based politics. We will also see that even the revived tax régimes of late medieval western Europe did not fully reverse this shift. …
One important result of this was that the western provinces became less economically complex. Even kings were less rich (although they also had lower expenditures), with the partial exception of the Franks. Nor were the aristocracies of the early middle ages anywhere near as rich as the richest senatorial aristocrats of the Roman world, with their estates spread across the whole Mediterranean (this, with political division, was by now impossible anyway); we find, in most places … very few landowners with estates in more than a couple of city-territories. The tax system had partially underwritten transaction costs for commerce in the later Roman empire; this had gone, which meant that interregional exchange decreased steadily, and in most of the western Mediterranean was restricted to luxury goods by 700 or so. At the same time, since aristocracies were poorer, and, since aristocratic demand fuelled much of the exchange inside regions, plus all luxury trade, commerce at all levels lessened nearly everywhere.
In every western province, archaeology shows clearly that fewer goods moved about, and that élite material culture was far less ambitious. … In one province where the military supply network was particularly important, Britain, the economic crisis was precipitous as soon as the army left in the early fifth century, that is to say even before the Anglo-Saxons came: cities were abandoned, rural villas were as well, and artisanal production beyond the village level ceased almost entirely.
Nothing in Gaul or Spain or Italy matches that crisis, but each of these shows a less extreme economic simplification too. The weakening of the wealth of élites was not all negative by any means. If aristocracies had less land and wealth, even though they still had plenty of tenants (many of them unfree), landowning peasants with less, or no, dependence on aristocrats must have become more numerous, and may have been more prosperous. But they bought fewer goods, and did not prevent the economy from simplifying. Anyone who wants to argue for continuity between the Roman and the post-Roman world must come to terms with these sharp economic changes, which are revealed so clearly in the archaeology.
Whatever continuities there were (and there were many) overlay a much less complex productive and exchange system, which had become less complex as a direct result of the political break-up of the west and the move of armies to the land. These were not structural causes of the end of the western empire; but they were certainly structural results.
The end of the western empire showed crisis … and sharp social and economic change.
… We will look at the three main successor-states to the Roman empire … Frankish Gaul – from now on often called Francia, both then and now – Visigothic Spain, and Lombard Italy.
… But let us begin with some general cultural and sociopolitical structures from the Roman past which continued almost without a break, and helped to define how the early medieval political systems of the west operated: the patterns of Roman provincial society and the Christian church, and the culture and values of public power.
The Roman empire had begun by being a network of largely self-governing cities, linked above all by the army. This had changed by the later empire. City councils weakened and failed in the fifth and sixth century everywhere, both east and west; government became more centralised after around 500 not only in the eastern empire, but, counterinituitively, in the weaker western kingdoms too. But loyalty to city-based societies survived everywhere where cities survived, which was all across the west except in Britain, north-west Spain and along the old frontier in Gaul and southern Germany. Coherent collectivities of local élites existed in the cities of southern Gaul, eastern and southern Spain, Italy; they made up the surviving Roman world that the new Germanic peoples came to rule, and … the two sides could accommodate themselves to each other fairly fast. These city societies were by now increasingly often represented, both in internal politics and with relation to royal power, by bishops. Christianisation was in effect complete in the whole of the former western empire by 500; the only major exceptions were the Jewish communities of parts of Gaul, Italy and, especially, Spain. …
… What was generally accepted was that the leaders of the church were the network of bishops which the Romans had established in every one of the cities of the empire, arranged in a hierarchy, province by province, with ‘metropolitan’ bishops (later called archbishops) at their head, and looking to the five patriarchs of the empire, of which one covered the whole of the west, the pope in Rome. This survived the end of the western empire little changed, except that the influence of the pope was never great outside Italy for many centuries.
Bishops were important in the late empire, but it was in the early middle ages that they really became major political players. Cathedral churches became rich in land, donated by the faithful, which made any bishop more powerful as soon as he took office. Bishops gained further spiritual authority from the cult of the relics of saints, which developed in the fifth century and onwards, for they tended to be in charge of the churches which contained them. They not only controlled urban religious ceremonies, but also became increasingly accepted as local political leaders (in most cases they were from leading local families); their appointment was often the focus for rivalry. And they represented their communities, to kings and other royal officials; kings took them seriously as leaders of these communities, as well as being prepared to hear the religious critiques that, as bishops, it was their task to provide. The new political prominence of bishops was partly because secular urban structures had melted away, and partly because, as a well-organised pressure group, they could make their voice heard in the weaker kingdoms of the post-Roman period better than in the imperial political system that had created them. …
… The other inheritance of Rome which needs to be stressed was a whole conception of political legitimacy which could be called the culture of the ‘public’. Under the empire, the publicum was taxation, imperial property, the bureaucracy, the collective good, just as the ‘public sector’ is today. When the publicum was no longer underpinned by the wealth of the tax system, this concept did not go away. Kings across the post-Roman west used the term routinely to mean rights which belonged to them, plus their officials, law courts, the road system, and so on.
The difference between the public and the private (another Roman and post-Roman word), so clearly maintained, justifies us considering the post-Roman kingdoms as states, even if often weak ones. Kings did not often in this period invoke the imagery of the public good when legislating; that would be for the Carolingians in the eighth and ninth centuries … But the idea that royal power constituted the public sphere was strong; and it could be meant spatially too – justice, for example, was done publice, ‘publicly’, in the sight of all.
The sight of all mattered very greatly in the post-Roman world, in fact. Here, the publicum as the ex-Roman state melded with one clearly non-Roman feature of all of the early medieval kingdoms, the public assembly. Assemblies of the entire political community, national or local, were essential to legitimate royal power, royal acts, and court judgements, throughout post-Roman Europe, north of the Roman frontier and south of it alike: they were called by different names, conventus or placitum, or in Anglo-Saxon England gemot, or in Scandinavia thing, and they can be found in Celtic- and Slavic-speaking communities as much as in Germanic- and Latin-speaking ones.
They seem to have derived from an early assumption, north of the border, that kings were responsible to and legitimised by all the free men (but not women) of their community, and that political practice was at its base collective. In a large post-Roman kingdom, this was impractical (doubtless it was always in part pretence), but even then kings legislated, at least nominally, ‘in the presence of the whole people, in common counsel with us’ (as the Lombard king Liutprand put it in 713), and the imagery of a very wide legitimising community, meeting publice, was a common one from 500 onwards. This was not a Roman concept, then, but the attachment to it of the Roman concept of the public was a natural one, and each reinforced the other. Post-Roman kings may have been at times quite restricted in their practical power, but the public sphere was theirs to dominate, and this distinguished rulers fundamentally from alternative powers in any kingdom. We can find this pattern everywhere in the west, up to the end of the Carolingian period and beyond; and when the culture of the public weakened, along with legitimating public assemblies themselves, from the tenth century onwards, political power would sharply change in nature, as we shall see in later chapters.
The culture of the public, assembly politics, Christianity and the network of bishops, a disappearing tax system and the beginning of the politics of land, a less wealthy aristocracy and a more independent peasantry, a simpler economic system: all these features marked out the post-Roman kingdoms.
So did a landed army run by no-longer-civilian aristocrats, which meant that aristocratic values became highly militarised from then on, and remained so for the rest of the middle ages and beyond; conversely, the literary education of Roman civilian élites became less important. Only the assemblies were not Roman in origin, although many were products of the division of the empire and the collapse of the fiscal system: that is to say they were very different from the Roman past, however much they developed out of it.
These were, anyway, the elements that the political leaders of the post-Roman world had to play with, and the parameters of the world in which they operated. Let us see now how this worked out in the different post-Roman kingdoms.
The Franks were among the least Romanised of the Germanic groups that conquered a slice of the Roman world in the fifth century … and they took over a sector of the empire which had suffered particularly from the troubles of the period, northern Gaul. They were not by any means united at first, and up to the late fifth century there were several separate Frankish kingdoms, intermixed with autonomous army leaders in the Roman tradition. …
… [Even] if from the 670s onwards the Merovingian kings were usually no more than legitimising figures for powerful aristocratic supremos called maiores, they were still essential for that political legitimisation. …
… Maiores gained further in power under the redivided kingdom of the sons of Dagobert after 639. By the mid-seventh century they were contesting the authority of queens regent when kings were children, and by now were sometimes even choosing which Merovingian to make king. They were matched in their power only by a small group of really powerful bishops, many themselves aristocrats, such as Audoin of Rouen (d. c. 684), one of Dagobert’s protégés, and Leudegar of Autun (d. 678), who was brought down and killed by the maior Ebroin. The last Merovingian who was a real protagonist, Childeric II, was murdered in 675, the low point in this sequence of events, and after that aristocratic families had no choice but to fight it out. The Pippinids won out at the battle of Tertry in 687, and a Pippinid maior was always the senior figure in Francia after that. That marked the end of the instability of the mid-century, which had, in the end, only lasted a generation. But Pippin II (d. 714), the victor at Tertry, had less power than many of his predecessors. In the period of trouble, the Franks had lost hegemony over the peoples of Germany, the Bavarians, Alemans and Thuringians, and also the dukes of Aquitaine in south-west Gaul. Even some bishops were beginning to carve out semi-autonomous territories for themselves. After Pippin’s death, his family dissolved into civil war as well, in 715–19, when Pippin’s widow Plectrude, regent for her own grandson as maior, confronted Pippin’s illegitimate son Charles Martel; for a while this must have seemed like the 670s all over again. But Charles’s victory showed that this was not the case; as sole maior (717–41), with only one court by now, he reconquered many of the newly autonomous lands, down to Provence; his sons Pippin III and Carloman I, later called Carolingians after their father, did the same with Alemannia and Aquitaine. The Frankish lands and wider hegemony could thus be reunited again, even after all the travail of the previous period, which indicates that the Frankish polity was pretty solid at its base.
This solidity was partly due to the density of Frankish government. We have more evidence for Francia, particularly in the seventh century, than for other post-Roman political systems, and it is clear from that evidence that its kings were active throughout Francia, intervening a long way from their political centres and moving aristocratic officials around … Merovingian government was complex and document-based, in a very Roman way; Audoin had also been Dagobert’s referendarius, responsible for the production of formal documents for the king. This partially dropped back under Pippin II, and even under Charles Martel to an extent, but Pippin III could begin the process of re-establishing it, and by 800, under his son Charlemagne, governmental complexity was greater than it had ever been. This was certainly an important parameter, and, to repeat, one with a solid Roman (public) tradition at its back. But the staying power of the Frankish political system was also the result of the constraints on aristocratic choices. Aristocratic political strategies, however exuberant and self-interested, above all revolved around the kings (later, the maiores), who were even richer than they, and who provided both patronage (land and money) and legitimacy, at least for the successful. To go it alone was for long impossible, and even after the 670s only aristocrats with a formal regional command, usually dukes, could do it. Aristocrats had a local base, certainly, and we can track regional rivalries in many cases. But in most regions they did not focus on local politics except when they were dukes or bishops, again with formal offices. Indeed, they could move their lands around the Frankish kingdoms; the quantity of them was in some cases more important for political success than their location. This would not change under the Carolingians either, although, when it did change, the structures of political power would shift substantially.
The crucial point here seems to me this. The Frankish political system was the strongest in the post-Roman west; although ramshackle and often violent, it had staying power.
Much of its strength came from Roman administrative traditions, as just noted. But, although the kings were unusually rich by the standards of the post-Roman world, Francia was not a tax-based political system; its armies were increasingly based on the military entourages of aristocrats, too. Kings needed to rule with the consent of these aristocrats, and rulers who did not do so … could be killed. It was normally straightforward to obtain this consent, for aristocrats did not have an alternative political context to operate in, and royal courts were anyway rich and attractive throughout.
The dice were weighted in favour of central power … But it was necessary to seek consent; the politics of land were already in operation, and, even if authority was not yet fragile, it could become so. Assemblies came in here, for they were in the Frankish world the locus of an aristocratic as well as a royal legitimacy. Kings and other rulers routinely sought the collective agreement of assemblies, like the 300 aristocrats who were called together by Fredegund in 585 to swear to the legitimacy of her son Chlotar …
… Visigothic Spain faced the same problems, but dealt with them in a different way. The Visigoths had not yet gained full control of Spain when Clovis seized most of their territory in Gaul, and the next half-century was difficult for them, with a very unstable succession and separatist revolts in the great southern cities of Córdoba and, later, Seville, and even in some rural territories, plus the east Roman conquest of the Mediterranean coast. Leovigild (569–86) however united nearly all Spain by force – all except the coastal strip, which was not retaken until the 620s, and the Basque lands of the western Pyrenees. Leovigild saw himself as a unifier in all respects; he issued a law code which contained the most Roman-influenced legislation of any of the ‘barbarian’ kingdoms, and he tried to address the religious division between Catholics and Arians, which was less tense in Spain than in Vandal Africa but tense enough, by alternately persecuting Catholics (particularly Catholic Goths) and trying to soften Arianism to make it more palatable to Catholics. This latter procedure had parallels in the east Roman attempts to bridge the Chalcedonian–Monophysite divide (it probably copied them) and was equally unsuccessful: religious divisions over the nature of God were never resolvable by compromise. Leovigild’s son Reccared (586–601) dealt with the problem by immediately becoming Catholic and, at the Third church Council of Toledo in 589, outlawing Arianism altogether: in future every Goth should be Catholic (Romans hardly appear in the council’s minutes; already, in effect, nearly everyone in Spain was becoming a Goth in a political sense). The impulse to unity in Spain henceforth took on a highly religious element, as it never did in Francia or Italy …
… The sense of doom which one finds in, in particular, late-seventh-century Gothic legislation has been taken too seriously by historians. They know that in 711 most of Spain would be conquered by Arabs and Berbers after the Visigothic king Roderic was killed in battle, with different parts of the peninsula flying off in different directions, and to them Spain was already breaking up well before this point. Spanish archaeology, too, shows that the economy was becoming very localised, variable, and in many areas fairly simple; our few non-royal sources also indicate very substantial social differences between (for example) a highly urbanised Roman-style south and a rural north with some very uncomplex societies indeed. As a result, the kings could not maintain their fictional homogeneity from their capital at Toledo, and their shrillness perhaps shows that they knew it. This last may be true, but it is at least as likely that the kings were simply influenced by ecclesiastical rhetoric in this very moralised world – as well as by the rhetoric of Roman imperial laws, for the Visigoths kept up a Roman governmental style to the end, with a careful attention to legal form even when the actual politics were messy. In reality, late-seventh-century Spain was very stable … kings died natural deaths, and rebellions failed, until just before the kingdom ended. As in Francia, the aristocracy revolved around the royal court, which was complex and more ceremonialised than elsewhere, and … still collected taxes; …The aristocracy was … much less rich than in Francia, and it is likely that the growing simplicity of the material culture found by archaeologists reflects this too. The rich royal court will have been all the more attractive to its members for that reason, not least because, given that succession was rarely hereditary, one might even become king oneself. So, far from showing weakness, our … evidence shows that, as in Francia … one could maintain a Roman-style governmental practice without the secure fiscal basis that the empire had enjoyed. These practices were further updated, too, for the Visigoths borrowed from the contemporary eastern empire as well.
Lombard Italy, finally, was somewhere in the middle. The Lombards invaded an Italy still disrupted by the Roman–Gothic war in 568–69, and one which the east Romans did not defend well thereafter, but they were a very disorganised invading force, and, after two kings in succession were assassinated in 572–74, broke up into several different political units, led by dukes. They reunited under a single ruler in 584, and their first really forceful king, Agilulf (590–616), defeated most of his rivals and established a capital at Pavia; all the same, when peace was made in 605 with the east Romans, who had maintained themselves in the former Italian capital at Ravenna, Italy was split up into several different pieces. The Romans controlled most of the coasts, and the major cities of Ravenna, Rome and Naples, but three large blocks of Lombard territory, the central-northern kingdom of the Po plain and Tuscany, and two independent duchies in the centre-south around Spoleto north of Rome and Benevento north of Naples, divided the Roman lands from each other. This was clearly a sign of failure, both of the Lombards and of the Romans, and it lasted; Italy was never united again until 1870. Although the Lombards slowly extended their lands across the next 150 years, they never managed to take Rome or Naples, or to unify the three separate polities which regarded themselves as Lombard, even under their two most ambitious and effective kings Liutprand (712–44) and Aistulf (749–56), who absorbed Spoleto and, briefly, Ravenna. The Lombards thus never had the military drive … or the urge to unity of [Frankish kings]. Although they could hold their own against the Romans, they were in trouble when they faced the Franks, who were intermittently hegemonic over them in the late sixth century and early seventh, and who defeated the Lombard army three times in the 750s and 770s, culminating in Charlemagne’s conquest of the Lombard kingdom in 773–74.
This may look unimpressive, but actually the Lombard kingdom was also the most tightly governed of the three main successor-states.
It was much smaller than Francia, so links between Pavia and local city-based societies were easier. It was also less regionally diverse than Spain; the economy was certainly more regionalised and simpler than under the empire, but we do not see the sharp involution of economic complexity which we see in some parts of Spain, and urbanism survived, if at a materially unassuming level, in most parts of the peninsula.
What Italy consisted of was a set of small-scale but stable provincial societies, whose élites were city-dwelling almost without exception. As in Spain, there was not any aristocratic stratum rich enough to make the kings fear their opposition in any systematic way (except, as again in Spain, in the case of successful individuals, usually dukes of one of the cities, who took power by coup), and no member of that stratum would have had the ability to establish a strong local power-base, given the number of rivals there would have been in each city.
The Lombard kingdom was very attached to assembly politics, as was Francia, but here the main function of both royal and local assemblies seems to have been law and justice, more than political deliberation – at least as it appears in our eighth-century sources … People appealed to Pavia, and got royal judgements back, as we can see in texts which show the losers obeying them, as well as in a substantial set of very detailed and one-off royal laws … There is an enthusiasm in some of [the] problem-solving which one does not find in any other law-making of the period – as with what should the penalty be if a man stole a woman’s clothes while she was bathing in the river and forced her to walk back to her house naked? (Answer: he should pay her his full wirigild, blood price, as if he had killed someone, for there would certainly be blood revenge taken otherwise.) This was a pragmatic, fairly low-key and cheap way of ruling, but it seems to have worked. The procedures of Lombard government were indeed borrowed from by the Franks after 774.
We have here got a long way from the complexity, the coherence and the wealth of the later Roman empire. None of these states taxed to any serious extent by 700, and the patterns of government were much simpler as a result. The economy was far simpler too (although northern Gaul maintained more of a network of production and exchange than the others, which fitted the greater wealth of its élites); in the Mediterranean kingdoms, it probably hit its low point in the eighth century.
But this was not an enclosed world – there were always interconnections, and movement, between the kingdoms, and the Lombard kings went so far as to develop a system of passports for travellers entering across the Alps at a time of political tension with the Franks.
And, above all, it was a governed world.
All three of the post-Roman kingdoms used writing-based techniques of government, of different kinds, which had been inherited from the Roman world, together with – in Francia and Italy, rather less in Spain – a tradition of assembly politics which had not. They also developed their own particular practices: in Francia, real deliberative assemblies, as well as an effective and usually regular war machine; in Spain, a tradition of strongly moralised and ceremonialised politics; in Italy, capillary government, both proactive and responsive. All these practices, largely developments of the earliest medieval centuries, would be used by the Carolingians after them.
The Source:
Chris Wickham, Medieval Europe, Yale University Press 2016 [pp. 22-42]
Evolutions of social order from the earliest humans to the present day and future machine age.