Ancient History & Civilisation





ROME: THE MORNING OF 25 DECEMBER in the year of our Lord 800. The Frankish king Charles the Great – Karolus Magnus: Charlemagne – is making a visit to the old imperial capital and enters St Peter’s to celebrate Christmas Mass. What happens next is described in the broadly contemporary life of Pope Leo III in the Liber Pontificalis (Book of the Pontiffs):

Then with his own hands the venerable bountiful pontiff, crowned [Charles] with a precious crown, and all the faithful Romans, seeing how much he defended and how greatly he loved the holy Roman Church and its vicar, at God’s bidding and that of St Peter, key-bearer of the kingdom of heaven, cried aloud with one accord: ‘To Charles, pious Augustus, crowned by God, great and pacific Emperor, life and victory.’ Three times this was said in front of St Peter’s sacred confessio, with the invocation of many saints; and by them all he was established as Emperor of the Romans.

324 years after the deposition of Romulus Augustulus, the lands of the old Roman West had a new emperor, in both name and fact, a figure of such towering beneficence that one of his court poets could style him Pater Europae, ‘the Father of Europe’. And yet, according to Einhard, Charlemagne’s biographer who knew him well, Charlemagne would later say that:

If he had known in advance of the Pope’s plan, he would not have entered the church that day, even though it was a great feast day.1

What is the explanation of this extraordinary statement, and how exactly had the Roman Empire come to be reborn in the person of a ruler of the Franks?

A restoration of empire based on a Frankish power base had nearly happened over 200 years earlier, when successive generations of a previous Frankish royal dynasty – the Merovingians – came within a cat’s whisker of claiming the imperial title for themselves. Any real understanding of who Charlemagne was, and why the imperial title eventually came to him on Christmas Day 800, requires us to delve (briefly) into this deeper Frankish past. As the alert reader will already be recalling, we have met the Merovingians before, in the person of Clovis who went head to head with the Theoderic the Amal for bragging rights over the former Roman West in the decades either side of the year 500. Theoderic came out on top, but Clovis has long enjoyed the greater historical reputation. This is due not so much to his own achievements, prodigious though these were, as to what happened after his death. Both the rise of his Merovingian dynasty and the reasons why it never quite grasped the imperial torch help bring Charlemagne properly into focus.


There are many parallels between the backstory to Theoderic’s Amal dynasty and the rise of the Merovingians among the Salian Franks. The main difference is that Attila’s Hunnic Empire had little direct part to play in the Frankish story. Attila tried to interfere in the odd Frankish succession dispute, but the Franks were physically beyond his reach, and there is no record that any Frankish groups ever fought for him, or had their leadership structures rearranged at his command.

That aside, the parallels are striking. For one thing, both dynasties emerge into the light of something approximating history at the same moment. The first Merovingian of whom we know anything at all definite is Clovis’ father Childeric, who died in c.480 and belonged to the same political generation as Theoderic’s father and uncle who first united the Pannonian Goths and steered them into a Constantinopolitan orbit. Childeric, however, did not get quite so far along the road to monarchical authority as Valamer, a relative obscurity which perhaps explains some of the mystery which surrounds him. Apart from having – reputedly – a sea monster for a grandfather, the main puzzle he poses us is that the (highly fragmentary) reports of his activities in the 460s and 470s place him more in central Gaul, participating in events at Orleans, Angers and the mouth of the Loire: bound up in the swiftly changing alliances which mark the last generation of the old Roman West (apart, that is, from a period of exile when his followers got fed up with him seducing their women). But the power base he left his son was centred much further northeast in what is now Belgium, on the old Roman legionary base at Tournai. And this, in the end, is the most securely attested fact of all, since his tomb was uncovered there in the seventeenth century, containing – amongst many other items – a seal ring conveniently inscribed with his name.2

Historical records and the contents of the tomb combine to suggest that he was commander of a force powerful enough to be courted, mobilized and rewarded as the last generation of West Roman leadership struggled to hold things together. But then, like most of the other players of this complex game, he eventually realized that it was time to draw a line under things imperial and operate independently, since the Roman centre had ceased to control any assets worth worrying about. We lack a precise chronology for Childeric’s moment of truth, but all the other players we know about gave up shortly after 468 when the defeat of Constantinople’s great Vandal armada removed all hope of shoring up imperial authority in the West.3 Within the constellation of figures who started to go it entirely alone at this point, however, Childeric was clearly still only a relatively minor player. Despite the impressive wealth submerged alongside him in Tournai, his son started his political life as only one of several Frankish leaders of similar stature

The pattern, however, was set for decisive change in the lifetime of his son and successor: Clovis. The new king’s mother was a princess of the Thuringians called Basina, whom Childeric had met – reputedly – during the involuntary exile brought on by his sex addiction. The later seventh-century chronicler Fredegar (also the source of the sea monster story) records what happened on the night of Clovis’ conception. Three times Basina woke up her husband and sent him outside to see what he could see. On the first occasion, he saw lions, unicorns and leopards; then wolves and bears; and the third time lesser beasts like dogs. You can imagine that Childeric may have been befuddled by lack of sleep, as well as slightly irritated at this point, but the queen helpfully made everything clear. The whole parade, like the vision revealed by Hecate to Macbeth, was an account of what was to come, though this time the line of Childeric’s own descendants rather than someone else’s. Clovis himself was the lion, of course, and three elements of his career are central to our unfolding story of Frankish Empire. Most obviously, he completed a series of military victories which greatly expanded the territory under his control. First in the firing line was a certain Syagrius, whose defeat is traditionally seen as extending Clovis’ territories as far south-west as Paris, or thereabouts. This was followed by other victories which brought the Burgundians of the Rhone valley at least temporarily under Frankish hegemony, and then, in the crisis which would generate Theoderic’s finest hour, massive victories over both the Alamanni and the Visigoths. Theoderic eventually imposed some counterbalancing influence over the Burgundians, and forced Clovis back too from Provence and the Mediterranean. Nonetheless, in the course of his maybe thirty-year reign, Clovis expanded his control from an originally pretty restricted corner of Belgium, to take in most of old Roman Gaul and a significant chunk of territory east of the Rhine (Figure 11).


At the same time, paralleling the efforts of Valamer and Theoderic, Clovis eliminated all the rival warband leaders within his sphere. As with much of the chronology of his reign, it is not completely clear when this happened. Our only source, the Histories of Gregory of Tours, presents the process as a series of Mafia-like hits organized in part through insiders, but certainly in a rush, at the end of the reign, in the aftermath of his victory over the Visigoths; that is, between 507 and 511. This may or may not be right, but there is no doubt that he effectively eliminated all his major rivals, many or most of whom – with that uplifting care for family we have encountered before among royal dynasties – were his collateral relatives. His sons alone would now inherit their father’s much expanded regnum.4

The third key element of Clovis’ career is that he eventually converted to Orthodox, Catholic Christianity. In this case, Gregory of Tours is certainly some way from the truth. He presents the king as going straight from convinced paganism to Catholic Christianity in the middle of the first decade of the sixth century, and winning thereby God’s favour for his upcoming campaign against the Visigoths. The latter, like Theoderic, were non-Nicene Christians of the type often mistakenly labelled Arians, so that via this construction Gregory was able to present Clovis in 507 as conducting a victorious Catholic crusade against hated heretics. All very satisfactory to a Catholic churchman, but Gregory was writing three generations after the event, and more contemporary sources suggest a significantly different picture. For one thing, Clovis seems to have declared his Catholic allegiance only after his great victory, and, even more interesting, the context of this declaration was one of intense debate at the Frankish court in which it at one point looked likely that the king might throw his hat into the Arian ring: a dimension of the story which had been entirely suppressed by Gregory’s time. In the end, Clovis opted for Catholicism and this too helped set the Franks on the path towards empire.5 His new, much enlarged Frankish power base combined territories both west and east of the Rhine, a core area in what is now France, Benelux and western Germany that was destined for a long and influential history, even if the exercise of political power within it would be far from smooth.

When Clovis died in 511, the only blot on his almost imperial record was the failure to outface his great Gothic rival. His death occurred precisely in Theoderic’s annus mirabilis, and news of it can only have added a bit of extra lustre to that creation of the united Gothic kingdom of Italy, Spain and southern Gaul (page 78). Nor was Clovis’ Catholicism any trump card. In 511, Theoderic’s relations with his Catholic clergy could hardly have been better, except for the fact that their best moment, when he brokered Constantinople’s return to the religious fold in the early years of Justinian, was still to come. Prodigious as they were, Clovis’ achievements did not rival Theoderic’s, and I don’t find it surprising that none of the Gallo-Roman aristocrats who had by this stage fallen under Frankish rule felt the same compunction as their Italo-Roman peers to hail their new ruler as semper Augustus.

However, as soon as Theoderic’s united kingdom broke apart on his death, and even the Italian portion of it was riven with the political struggles first for control of Athalaric and afterwards for the throne itself, Frankish expansion – on hold since 507 – could begin again in earnest. In the early 530s, the Thuringian and Burgundian kingdoms, deprived of the support of an effective Gothic counterweight to Frankish power, were swallowed up, and more was to follow with first Provence and then extensive territories in the northern Alpine foothills being formally ceded to the Franks by Wittigis as he desperately tried to mobilize more troops for the war (above, page 156).

In addition, the sons and grandsons of Clovis – leopards, unicorns, wolves and bears – were busy on their own account east of the Rhine (Figure 11). When Clovis’ grandson Theudebert wrote to Justinian in the year 540, he could with justification declare himself ‘the ruler of many peoples’, including Visigoths, Thuringians, Saxons and Jutes, as well, of course, as the Franks themselves. By the mid-sixth century, Clovis’ descendants were close – so close they could almost smell it – to imperial power in the West, in both fact and legal claim. In one ideological dimension, they went even further than Theoderic had dared. It had long been a prerogative of Roman emperors – stretching back into the more distant past when many constituent cities of the empire had issued their own coinages – that they alone could issue gold coins. In the generations after 476, rulers of the successor states essentially respected this prerogative, maintaining only base metal coinages. Theoderic himself minted only the occasional gold medallion (Plate 8). When Merovingian kings started to issue gold denominations regularly, this represented a major break with tradition, and an affront that Procopius noted with disgust. Nonetheless, for all that they were mighty rulers, and praised with many of the epithets that used to be trotted out about emperors, still no one hailed even Clovis’ grandsons as truly imperial Augusti, and there is no sign, in the extensive court-based literature of the third quarter of the sixth century, of any conscious project of imperial restoration of the kind undertaken by Theoderic. Although they got close, therefore, no member of the Merovingian dynasty ever quite managed the jump to imperial lightspeed.

The reasons they didn’t, I think, are straightforward. Extensive as the territories were that fell under Frankish sway in the sixth century, they were only rarely united under a single ruler. As it developed over the sixth century, the Merovingian kingdom came to have five core components – Austrasia, Neustria, Burgundy, Aquitaine and Provence – with differing degrees of hegemony being exercised over satellite territories in Thuringia and Alamannia, Bavaria, Frisia and Saxony (Figure 12). It was standard Merovingian practice even for the core areas of the kingdom to be divided between a ruler’s surviving adult sons, whoever the mother, as they were on Clovis’ death in 511. Many pages could be written (and have been) detailing the ensuing political struggles – featuring brief periods of unity, lots of potential heirs and nephews, early deaths and straightforward executions – but we don’t need to worry about them here. Suffice it to say that any Merovingian pretensions to claim a Western imperial title were materially hampered by the fact that only occasionally was the Frankish kingdom’s potentially mighty power base in the hands of just one ruler.6

Equally important, the kings of the Franks had to share the former Western imperial airspace with other, entirely legitimate Christian rulers. Especially after their formal conversion from Arianism to Catholic Christianity was declared at the Third Council of Toledo in 589, the Visigothic rulers of Spain and Septimania (an arc of territory in what is now south-eastern France north of the Pyrenees) drenched themselves in a Roman and Christian sacrality, which determinedly echoed that of the emperors in Constantinople. They looked for all the world like perfectly legitimate rulers of a substantial part of the old Western empire. The exact religious status of the Lombards who intruded themselves into northern Italy as Avar power began to entrench itself in the middle Danube region in the late 560s is harder to judge. The sources preserve confused – indeed confusing – indications about whether Lombard rulers were generally Catholic, Arian, or something else. But, at least on occasion, Lombard kings could do a pretty decent impression of respectable Roman-type rulership, not least in 612 when King Agilulf presented Adaloald, his son and chosen heir, to his subjects in the circus at Milan (an old imperial capital) in direct imitation of Roman ceremonial practice. And all this, of course, is to take no account of the fact that, up to the mid-sixth century, the East Roman state – complete with its own ‘proper’ emperor – remained a highly active force within the Italian peninsula (controlling large swathes of territory around Ravenna, Rome and in the south), not to mention maintaining a thoroughly vital governmental structure over such echt bits of the old Western Empire as North Africa, Sicily, and the Adriatic coast.7

At times, the Franks waged major wars against all of these peers, but, in the later sixth century, Frankish intrusions south of both the Alps and the Pyrenees tended to take the more restricted form of raiding, and there were no further serious attempts to increase the amount of former Roman territory under Merovingian control. As a result, no Merovingian was ever dominant – even by a mixture of direct and hegemonic rule – over so large a portion of the Roman West as Theoderic had been in his pomp, and, in general, Frankish imperial credentials lacked sufficient clout to overpower the claims of the other successor states, and indeed of East Rome itself.

If Clovis’ sons and grandsons were mighty but not quite imperial, traditional accounts of his seventh- and early eighth-century descendants, the lesser beasts of Basina’s dream, have been overwhelmingly shaped by historical works produced in the time of the Carolingian dynasty to which Charlemagne belonged. In particular, Einhard, once again writing early in the reign of Charlemagne’s son, has left us an irresistible pen portrait of the last Merovingians.

Nothing was left for the king to do except sit on the throne with his hair long and his beard uncut, satisfied [to hold] the name of king only and pretending to rule … he listened to representatives who came from various lands and, as they departed, he seemed to give them decisions of his own, which he had been taught or rather ordered [to pronounce]. Except for the empty name of ‘king’ and a meagre living allowance which the [mayor of the palace] extended to him as it suited him, he possessed nothing else of his own but one estate with a very small income. On that estate, he had a house and servants who ministered to his needs and obeyed him, but there were few of them. He travelled about on a cart that was pulled by yoked oxen and led … by a herdsman … In this way he used to go to the palace and so also to the public assembly of his people, which was held annually for the good of the kingdom, and in this manner he also returned home. But it was the [mayor of the palace] who took care of everything, either at home or abroad, that needed to be done and arranged for the administration of the kingdom.

Once the extraordinary history of Gregory of Tours gives out in the mid-590s, no narrative survives with anything like the same scale of circumstantial detail until you reach the Carolingian period, so that there is little with which to challenge Einhard’s vivid characterization. In general terms, therefore, historians have often been happy to see later Merovingian kings as lesser beasts, their lack of authority condemning the Frankish world to disorder, allowing power steadily to coalesce in the hands of second-rank figures: the mayors of the palace. This was the role filled by Charlemagne’s ancestors in the northeastern part of the kingdom, Austrasia (Figure 12), until his father Pippin the Short decided that enough was enough in 751 and made himself king of the Franks instead.8

But Einhard was writing in the third political generation after Pippin had become king, and was straightforwardly a Carolingian loyalist. What could be more convenient for the dynasty he served than the idea that the Merovingians had forfeited their right to rule through incapacity? Looked at more closely, as the last generation or two of modern scholarship has done, the sources have yielded a rather different story: one where both the later Merovingians are less useless than Einhard would have us believe and the rise of the Carolingians has become a much more contingent outcome.

Even taking this into account, though, there is not the slightest doubt that representatives of the dynasty ruling between, say, 675 and 700, were essentially occupying a much less powerful political position than their predecessors of a hundred years before. In particular, the interim had seen the rise of well-entrenched groupings of regional nobilities in all the core regions of the Frankish world, through whom any later king had to work, and the power of these nobles insulated many of the localities within those regions from the direct reach of the king. The most vivid picture we have of any of them comes not from Austrasia, but from adjacent Neustria to the west (Figure 12), where a key source, the Book of Histories of the Franks (the Liber Historiae Francorum), shows us that any Merovingian monarch was expected (not least by the author of the text) to work with and through an interrelated grouping of around half a dozen noble clans. These spent their time marrying and competing with each other for dominance within Neustria, which they exercised by holding the position of Neustrian Mayor of the Palace, and then, increasingly, in trying to fight off Carolingian interference.9 Each of the core regions, in other words, had thrown up their own equivalent of the Carolingian family, and, collectively, these nexuses of nobility greatly reduced the effective power of any king, whatever his personal abilities.

This means that although the ruling dynasty was weakened, the Carolingian replacement of the Merovingians was no simple palace coup d’état, where the king’s chief minister finally put himself in a position to oust the last member of the old line. What we see, taking the sixth, seventh and eighth centuries together, is a major restructuring of power, in which the Carolingians’ seizure of control over the entire Frankish world in the eighth century represented a distinct second stage. In stage one, Charlemagne’s ancestors were no more than the winners of a regional, intra-Austrasian competition, and a set of peer rivals were busy winning similar competitions in the other main regions of the Frankish kingdom.

I strongly suspect that if the Carolingians hadn’t called time on the increasingly delicate balancing act that was the late Merovingian system, then a rival noble lineage from one of the other core Frankish regions would have done so instead. This does not sound like a recipe for long-term stability, since someone, somewhere was likely to get the idea of reorienting the entire political system around their own authority, if only out of fear that a competitor might do so in their place. A similar sense of instability also shows up on the periphery. Here the sixth-century Merovingians also exercised authority – partly directly, partly by occasional assertions of military hegemony – over a series of satellite areas: Alamannia, Thuringia, Bavaria and even, to a lesser extent, Saxony, which was forced to pay tributes.10 As the seventh century wore on, these satellite areas all broke away from central control, and this too, as much as the growth of regional noble power, reflects a breaking down of the old system. A sequence of highly contingent events underlay the ability of the Carolingian line to seize total control over Francia, but, by c.700, the Merovingian line was anyway destined for the dustbin of history.

Even if Einhard’s characterization of the last Merovingians was satirical, he is at least pointing us in the right direction. The period c.550 to 700 saw a massive net transfer of power from king to regional nobilities, and the result was both overall political instability, and kings who, while remaining important within the system, had lost their capacity to dominate it. This was not the result of dynastic decline that had turned lions into pussycats, and some Merovingian rulers, such as Dagobert I (who died in 639), had not forgotten how to roar; but the later Merovingians were fighting a hopeless cause in the face of structural shifts in the location of power in the Frankish world. By 700, the outer reaches of the Frankish quasi-empire of the sixth century had already fallen out of orbit and the core was fragmenting. How the Carolingians emerged from this process of central enfeeblement brings us face to face with the career of the first great Charles of the new dynasty: Martel, ‘the Hammer’.


A centrepiece of any tour of Versailles is the Gallery of Battles. 120 metres long and thirteen wide, it was created out of a series of smaller salons by Louis Philippe in the 1830s. This – almost entirely pacific – king was desperate to show that he belonged to a martial French tradition that had culminated so recently in the career of Napoleon. As I remarked helpfully to a French former penfriend, there’s no picture of Waterloo; that aside, the great battles of l’empereur are all there.

But the thirty-four major paintings cover a much longer time frame than this. First on the left on the standard tour is a modest 5.42m by 4.65m number depicting Charles Martel at the battle of Poitiers (or Tours) traditionally dated to 732, although it might have been 733. This famous battle was fought by Charles as an ally of Eudo, Duke (dux) of Aquitaine, who had called Charles in to help him fight off a Muslim army which had penetrated north over the Pyrenees, the old Visigothic kingdom of Spain having been wrapped up by Muslim invaders in the decade or so after their initial victory in 711. After a lengthy stand-off, battle was eventually joined and Charles was victorious, his Muslim opponent Abd ar-Rahman being left dead on the field. Traditionally, this was viewed as the moment when a potential Islamic conquest of the whole of Western Christendom was frustrated, and inspired from Edward Gibbon one of his most famous passages:

A victorious line of march had been prolonged above a thousand miles from the rock of Gibraltar to the banks of the Loire; the repetition of an equal space would have carried the Saracens to the confines of Poland and the Highlands of Scotland; the Rhine is not more impassable than the Nile or Euphrates, and the Arabian fleet might have sailed without a naval combat into the mouth of the Thames. Perhaps the interpretation of the Koran would now be taught in the schools of Oxford, and her pulpits might demonstrate to a circumcised people the sanctity and truth of the revelation of Mahomet.

If preserving Christendom was not enough for a day’s work, then there was what lay behind Charles’ great victory. In traditional narratives of the birth of feudalism, Charles was seen as a prime architect in the emergence of the heavy armoured cavalry which would be such a fundamental feature of the central European middle ages. The downside here was that mailcoats and chargers are both mighty expensive items, so Charles was also remembered as a great nationalizer of Church property, using the wealth of religious institutions to pay for the military forces which had allowed him to fend off Islam.

As so often proves to be the case with traditional views of the past – most of which became entrenched in their national psyches from the later nineteenth century as heightened nationalism and mass education coincided – neither of these twin pillars of Charles’ reputation has quite stood up to closer, post-nationalist scrutiny. Abd ar-Rahman was not leading the force of certainly many tens of thousands (and sometimes, amazingly, hundreds of thousands) that he was often credited with, did not suffer catastrophic losses of manpower (although he himself certainly died), and was probably engaged in a little profitable raiding, rather than a determined war of conquest. Whatever else it may have been, Poitiers was no titanic struggle to decide the religious fate of Europe. Nor, sadly, did Charles single-handedly invent feudalism and solve the resulting funding issue. He had a reputation in the century after his death of putting considerable amounts of Church land into lay hands, but this was not a new practice. Many pieces of notionally Church land in the early Middle Ages were held by laymen in semipermanent tenancies called precaria, these laymen often being ‘friends’ of the religious institution in question, or even relatives of its founder. This went on long before and after Charles, and the big problem with his reputation may be that his victorious career allowed him to grant these tenancies to supporters who had no track record of association with the particular institutions concerned. And while heavy armoured cavalry was in the process of becoming a defining characteristic of Frankish warfare in the eighth century, this seems to have been a steady evolution, not a one-generation revolution like the emergence of the armoured division in the 1930s.11 But even if we can’t hang on to all the trappings of the traditional picture of Charles Martel, there is still no doubt that his career set the Carolingian dynasty decisively on the road to imperial power.


Not that such an outcome looked remotely likely on the death of his father Pippin of Heristal in December 714. This is the first recorded moment when Carolingian skeletons start tumbling out of cupboards, and a nasty round of familial in-fighting plays a major role in the story. The immediate problem was Pippin himself. He had produced two sons with his current wife Plectrude: Drogo and Grimoald. Both, however, predeceased him. Drogo died in 707, and Grimoald just before his father earlier in 714, Grimoald leaving one male heir of his own: Theudoald. But Pippin also had two other adult sons by his concubine Alpaida, Charles (our Charles) and Childebrand. Desperate to protect her grandson’s interests, Plectrude had her stepsons, Theudoald’s half-uncles, imprisoned, although the redoubtable Charles soon escaped to rally a formidable body of Austrasian supporters.

Although a normal enough kind of intra-familial spat – which would be a regular feature of Carolingian history from now on – the dispute had dire consequences in 714 because Pippin’s dominance over the Frankish world had had distinct limitations. As a result, a long list of the aggrieved parties lined up to exploit the political paralysis engendered by the quarrel between Plectrude and her stepson Charles. First in line, not surprisingly, were the Neustrians in the old Frankish heartlands. At the time of Pippin’s death, Theudoald was notionally mayor of the palace for Neustria, but Pippin’s control there had really only been negative – making sure that no serious rival emerged – and, once death had removed him from the scene, elements of the Neustrian nobility quickly asserted themselves, immediately throwing off any residual allegiance to Theudoald. Beyond the core, the Aquitanians were very ready to use the crisis to assert a greater degree of independence and, as quickly as 715, appear as participants in the action under their own Duke Eudo, as do the Provençales under dux Antenor. Even forces from outside the traditional Frankish ring were drawn into the action. Pippin’s pre-eminence had always been partly based on successful campaigning beyond Frankish borders. One of his victims had been the Frisians under King Radbod, defeated heavily in 690 and 695. But Radbod remained sufficiently powerful afterwards for Pippin to marry his son Grimoald to Radbod’s daughter, so that Theudoald was Radbod’s grandson as well as Pippin’s. Radbod, therefore, was ready to weigh in, quite probably with entirely mixed motives in mind, since a decisive reassertion of independence was an attractive goal for him, whether this was achieved by putting a grateful grandson in power, or by defeating Charles and his Austrasian supporters.

In 715, Charles’ position was extremely perilous. Many of his father’s henchmen had rallied to his standard, but the coalition of his enemies was potentially overwhelming, and the first omens were far from good. Radbod inflicted some kind of defeat on Charles fairly early in 715, but maybe his Austrasian support was not fully mobilized at that point. Certainly, Charles was able to return the favour later in the year and drive Radbod’s forces back into Frisia, and that is the last we hear of Radbod in the war, so it was presumably a major victory. The Neustrians, meanwhile, had been far from idle. One Ragenfrid quickly emerged there as mayor of the palace, operating under a Merovingian flag of convenience in the persons first of Dagobert III, and, then, after his death, of Chilperic II. Two substantial but indecisive battles were fought in April 716 and spring 717, by which time Charles had found his own Merovingian, Chlothar IV.

The real showdown came in spring 718 by which time the Aquitanians under Eudo had joined in the struggle to throw over the Austrasian hegemony that Charles now represented. But this time Charles’ victory proved decisive, and in its aftermath he was able to stomp right across the Neustrian heartland, capturing the city of Orleans.

Immediate military victory was also turned into lasting political control. In particular, he appointed key supporters to leading positions in the Neustrian heartland, which his father had never done, including his half-nephew Hugh, Drogo’s son, who was made abbot of two rich monasteries and five sees including Paris and Rouen. By 720, Charles’ position was further consolidated (Chlothar IV having died) when Eudo handed over to him Chilperic II and a substantial quantity of treasure.12

Although it had been touch and go, Charles was firmly in control by the early 720s. At that point, the internal Carolingian succession dispute was done and dusted. Drogo’s son Hugh received an astonishing pay-off, but he had obviously declared his loyalty to Charles at an early date. Other male descendants of Drogo and Grimoald were less fortunate. Charles didn’t have them killed, not even Theudoald, but by 723 they were all safely in custody and there they remained. Moreover, Charles had not only re-established Austrasian hegemony over Neustria, but taken it to a new level. There was now no Neustrian mayor of the palace at all, and Charles was appointing key supporters to rich and influential positions in the Neustrian heartland, a move which had been well beyond his father’s political reach. As yet, the old core regions of Burgundy, Provence and Aquitaine (not to mention the satellite duchies) were independent of his control, but much of this would follow before Charles’ death in 741.

The details of his subsequent campaigns don’t need to delay us, not least since it is impossible to reconstruct them in great detail anyway. But the most plausible outline runs as follows. By the 730s, Charles’ effective control had reached Burgundy, where he was now appointing more of his supporters to key positions. Whether this political pre-eminence had required much in the way of actual campaigning, is unclear. Some aristocratic elements within Provence, however, certainly were ready to resist. They had their own duke, one Maurontius, and when Charles started to flex his muscles in their direction, some were ready to fight. Maurontius also negotiated some assistance from his Muslim neighbours, who had taken over, along with the rest of the kingdom, the old Visigothic toehold north of the Pyrenees: the province of Septimania centred on Narbonne. But Charles had allies within Provence too. One of the most interesting documents to survive from this era is the will of the Patrician Abbo (from 739), a major Provençale landowner, who campaigned alongside Charles and received plentiful rewards in return. Of the satellite duchies, the Hammer’s domination over Alamannia was pretty much complete by 734, when he expelled the sitting ducal line, and relations with Bavaria – following some initial conflict – were close and amicable, Charles having taken Swanahild, a Bavarian princess, as his second wife in the 720s. Elsewhere, he had asserted his power regularly over the Frisians, who still retained some autonomy, but had made only occasional noises in the direction of the Saxons.13

By the time of his death on 22 October 741, Charles’ overall achievement was enormous. Of the old core regions of Merovingian Francia, only Aquitaine was beyond his direct control, and his influence over the satellite duchies was considerable. A clear sign of the self-confidence he felt in his later years was that, when his final choice of Merovingian frontman, Theuderic IV, died in 737, he felt no need to appoint another. For the last four years of his life, Charles ruled Francia in his own name, not as king, but as Duke and Prince of the Franks (Dux et Princeps Francorum), and it does seem likely enough that, from the 720s onwards, the Merovingians were finally reduced to the kind of job description set out for them by Einhard. But one important question remains: why was he so successful?

What really needs to be explained, in fact, is how he managed his initial victories. Once he had started to win, then a straightforward logic set in, which meant that one success would generate another. This worked both in Francia and elsewhere in early medieval Europe too, and the reason behind it was simple enough. After the disappearance of the military structures of the old Roman Empire, military forces came to consist in large measure not of full-time professionals, but of landowners accompanied by contingents of their more capable dependants, which probably – for richer landowners at least – included a handful of more or less professional heavies. These landowners had a recognized duty to provide military service, but served much more readily under a leader who was likely to bring victory in his train; that is, one who already had a convincing track record. This was not just because you were less likely to be killed, but also because a victorious leader was expected to reward loyal service with great generosity, both in the form of moveable wealth (as exemplified in the astonishing gold wealth of the seventh-century Staffordshire hoard recently unearthed in England – Plate 00 – and Frankish kings were much richer than their English counterparts), and – for the more prominent – landed estates of one kind or another. Both forms of wealth tended anyway to be much more available to a winner than a loser, since winners were entirely within their rights to confiscate treasure and landed assets from defeated opponents. This is why Charles had negotiated hard with Eudo for the return of both Chilperic and his treasure at the end of the war, and it was precisely the process of rewarding loyal supporters (such as Drogo’s son Hugh) with Neustrian assets which generated the Hammer’s reputation for being overly liberal with Church property.

From the moment of his decisive victory in 718, Charles’ record of success became a massive encouragement to potential new supporters to sign up with him, since he already had plenty to give away. And this same willingness to serve him on the part of further bodies of militarized landowners (such as the network of Abbo and his supporters in Provence) in turn made it more likely that there would be more victories and yet more rewards in due course. But for this virtuous circle to kick in, you do have to win that initial victory, and this is precisely what our sources fail to explain. It is possible to offer a couple of observations, however, which must be pretty much along the right lines. First, it was not really that surprising that most of the militarized landowners of Austrasia decided to back Charles Martel as Pippin’s heir, rather than his grandson Theudoald. We don’t know when exactly Theudoald was born, but he was only a child, and certainly not fit to provide the kind of effective military leadership required in the political crisis unleashed by Pippin’s death. Charles was an adult male of twenty-eight, and a much more plausible candidate. He had also been playing some kind of public role in his father’s entourage for many years, probably serving on some of the campaigns. He had every chance to build up working relationships with at least some of his father’s key supporters, and this is presumably why they supported him rather than his step-nephew.

The choices of these militarized Austrasian landowners also had a second dimension of importance. Not only did they pick Charles, but they seem to have served him as a coherent block with a considerable degree of group loyalty. At least, there is no sign in the sources that any Austrasians saw a Neustrian leader as potentially a better source of rewards than their own man. Even in the crisis years of 715 and 716, Charles appears to have been able to count on solid Austrasian support (although, of course, it was important that he managed not to suffer any major defeats at this juncture). This has to reflect the relative political solidity that his father Pippin had built up among these key supporters during his own lifetime of political success. From at least the time of his decisive victory over all-comers at the battle of Tertry in 687 – celebrated in the Carolingian histories of Charlemagne’s time as the start of the dynasty’s effective domination of the Frankish world – Pippin had generated loyalty by handing out rewards aplenty, obtained both within Francia and by regular campaigns outside it (whose importance we will look at in the next chapter). It was probably this broader inheritance of his father’s militarily resilient force which got Charles through the initial testing years.14

The only drawback to Charles’ reign was that, on his death in October 741, he left his domains facing a succession which was much more complicated, in its way, than his own had been some twenty-seven years before. On one level, things were easier. No major Neustrian magnate, or one from anywhere else within the heartlands of Francia for that matter, attempted to use succession as an opportunity to overthrow Carolingian dynastic dominion. And that is a real measure of Charles’ political achievement within Francia, because the intra-dynastic quarrel that soon followed was so messy that, had the Hammer’s rule not been so securely established, you really would have expected some aristocratic networks from somewhere within a nexus of Neustria, Burgundy or Provence to have at least attempted to reassert their independence.

Within the dynasty itself, however, the fallout from Charles’ demise was both spectacular and far-reaching. The Duke and Prince of the Franks left three adult sons by two legitimate marriages: Carloman and Pippin from his marriage to Rotrude (daughter of the Bishop of Trier – best not to ask …) and Grifo from his second wife, the Bavarian princess Swanahild. He also had three other sons with his concubine Ruodhaid, who never figured in anyone’s political calculations as far as we can see. According to the chronicler Fredegar (actually a continuator adding to the original seventh-century text), Charles decided just before his death, in 740, that his domains should be shared only between the two sons of Rotrude. But this continuation was written later, and there are good indications that Charles had a three-way split in mind, including Grifo in the mix.

Carloman and Pippin, however, didn’t hesitate, and the body count begins to rise. They executed poor old Theudoald, the last remaining male descendant of Pippin and Plectrude who had long languished in a kind of protective custody, showing far less compunction than their father. Their half-brother Grifo was imprisoned at more or less the same time, and his mother, Swanahild, packed off to a nunnery. Having cleared out the dynastic undergrowth, they then divided their father’s lands between them in 742, before, interestingly, finding themselves a tame Merovingian frontman – the last ever, Childeric III by name – and appointing him king in 743.

It’s sometimes thought that this was to forestall any possible political unrest from other, non-Austrasian sections of the Frankish nobility, who might have seen succession as offering a last-ditch opportunity to resist the hitherto seemingly unstoppable rise of the Carolingian line. But there’s no other sign of serious internal Frankish unrest, and my own suspicion is that appointing a Merovingian was meant to help keep the peace between two brothers who – to judge by the amount of fraternal amity knocking around within the family – must always have been potential rivals once they had cleared out Grifo and the rest. Whatever the precise driver, the brothers then showed themselves completely ruthless towards outsiders as well. The Aquitanians were determined to hold on to their autonomy, so battle commenced on that front, while attempted resistance on the part of some of the Alamanni led Carloman to call a council of all the nobility of the region to the old fortress of Canstatt in 746. There he assassinated them, liberating their goods and lands for distribution to his own largely Austrasian henchmen.15

At this point, matters seemed settled, but this turned out to be only stage one of a two-stage succession process. For in 747, the year after the Canstatt massacre (so it is known), Carloman took the extraordinary decision to withdraw from Frankish politics and head off to Rome for the good of his soul (not, to start with, actually to become a monk). Whether this was linked to Canstatt in some way is unknowable; it has been suggested, for instance, that he might have been filled with remorse. Possibly, but an alternative scenario would be that the massacre and financial redistributions were designed to generate the warmest possible feelings of loyalty towards himself and his line at a moment when he was planning to disappear. For Carloman’s plan was that his son Drogo should succeed him, and for any such plan to work, as Plectrude had found to her cost in 715, you needed consent from a critical mass of your warrior aristocracy.

To start with, all seemed well. Carloman trotted off to Rome and there are documents from 747–8 suggesting that Pippin initially ruled equably enough alongside his nephew Drogo; both, at least, held reforming Church councils in 748 (a highly significant point to which we will return in Part IV). But then the pattern of Frankish politics changed forever. With very little fanfare or even explanation in our sources, Pippin suddenly emerges – by himself – as king of the Franks, after a consecration of some kind was held probably in 751 (although it could just about have been 752). For such a massive, game-breaking event as the deposition of the last Merovingian and the crowning of the first Carolingian monarch, you would expect a great deal of historical noise. The fact that it’s not there, and that in some places our sources preserve slightly contradictory accounts of what happened, speaks volumes. Pippin’s accession to the ranks of royalty involved considerable self-assertion on his part, assertion which was – and this is the potentially embarrassing part – directly at the expense of other close members of his family.

The process began late in 748, when Drogo disappears from the sources. In April of that year, Pippin’s wife Betrada had given birth to Pippin’s first son, the future Charlemagne, and some have wondered if the birth was a significant moment in the development of Pippin’s calculations, giving him an obvious reason to eliminate his nephew from the picture. My own preference would be more along the lines that Pippin couldn’t make a move against Drogo until he had first prepared the ground by building up the right kind of contacts with enough of his brother’s former key supporters. When push came to shove, would his father’s men stand up and fight for Drogo or would they accept Pippin’s leadership? It looks as though Pippin did his job extremely well, in fact, for there is no record of any major conflict in 749–50 and I suspect that a real civil war – however embarrassing – would have crept into the historical record. Drogo had enough support, it seems, to stay out of Pippin’s clutches, but not enough to resist him on the field of battle, and someone else took the opportunity to muddy the waters still further by helping Grifo to escape from custody as well.

But none of this was enough to prevent Pippin’s political bandwagon rumbling towards its final destination. Later sources make great play of the positive reply received to a letter sent by ‘the Franks’ to the Pope asking whether the man holding real power should hold the royal title, rather than a powerless cipher. But it is unclear that any such letter was ever sent (there are inconsistencies in the accounts), and, as a myth, it was an excellent device for smearing a veneer of divinely ordained legitimacy over Pippin’s aggressive self-assertion. Pippin’s original coronation in 751/2 was a purely Frankish affair, which may or may not have involved an Old Testament-type anointing. Only subsequently was the Pope brought in to add his blessing. Pope Stephen had his own reasons for travelling to Francia in 753/4, as we shall see in a moment, but, for Pippin, it was all extremely convenient. A second ceremony was arranged for 754, this one certainly involving an anointing, which helped tie up the final loose ends. In 753, with the Pope on the road, Drogo and his never-named younger brother were definitively brought into custody, and Grifo was assassinated. Carloman himself also came back to Francia in the Pope’s wake, but was promptly despatched to the monastery where he would die in 755.

Carloman’s final journey has generated endless speculation. Was he trying to save his son? There is also a late report that he came as an emissary on behalf of the Lombards. But after Pippin’s coronation there was no possible way back for Drogo to a share of power, and I strongly suspect that Carloman’s return to Francia was demanded by Pippin as part of the price for the alliance that the Pope was trying to stitch together. For, having got rid of his offspring, Carloman represented the final loose end that Pippin would not want to leave dangling. Perhaps Carloman’s return was the price he had to pay to ensure that his son might live out his days in a monastery (as he seems to have done) and not share the fate of the assassinated Grifo. Either way, the papal visit was no more than a bit of useful icing on a fundamentally Frankish cake, and most – or enough – of Carloman’s old supporters had clearly accepted their invitations to the ceremony.16

With that, the line of Austrasian mayors set the seal on their rise to royal power. Charles Martel blew away the opposition, placing his line on a pedestal far above their former peers in the other core regions of Francia. Pippin’s political manoeuvring then helped everyone reach the natural conclusion to his father’s career: Carolingians and not Merovingians should now supply the kings of the Franks.


Like our two previous Western imperial revival acts, the stench of contingency hangs over the events leading up to Christmas Day 800. There is no sign that Pippin entertained even the vaguest of imperial ambitions, and many accidents had to coincide before the thought took shape for his son. But it is true nonetheless that the reasons which made Pope Stephen willing to travel to Francia, to play his walk-on part in the second ceremonial expression of Pippin’s acquisition of the mantle of royalty, form the crucial backdrop to Charlemagne’s imperial title.

They had their roots in events 2,000 kilometres and more to the east. As we saw in the last chapter, the first round of Muslim conquests in the middle decades of the seventh century robbed the East Roman Empire of its richest Middle and Near Eastern possessions, and demoted it from world to regional power status. But this was not yet the end of its struggles. Conflict with the caliphate intensified again from the 690s and the early 700s, with the imperial capital facing besieging Muslim forces for a full year in 717–18. It survived, but only by the skin of its teeth, and further debilitating losses in Asia Minor followed in the 720s and 730s. The eastern front simply had to be Constantinople’s chief priority, and there was no military or economic capacity to spare for the still substantial parts of Italy which had been left in its control in the immediate aftermath of Lombard invasion. As we shall see in more detail in Chapter 7, this both allowed the region around the city of Rome to emerge as an independent state under papal control, which called itself the Republic of St Peter, and encouraged Lombard kings to develop ever more expansionary ambitions within the peninsula.

Initially, the papacy had found that an alliance with the independent Lombard Duke of Spoleto was sufficient to fend off the Lombard monarchy in the north, but by the 740s Pope Zachariah was forced to deal directly with them, negotiating peace (at the cost of some cessation of territory) with King Liutprand (712–44). But as the Byzantine position continued to weaken, by 751 – when Pippin was having himself crowned king – Liutprand’s successor Aistulf was marching victoriously into Ravenna and the neighbouring duchy of the Pentapolis (based on five Adriatic coastal cities between Rimini and Ancona: Figure 13). Aistulf’s forces were also moving into the Istrian peninsula and had already made a separate peace with the Venetians. His eyes then turned greedily towards Rome.17

It was this new strategic situation which set Pope Stephen on the road to Francia in 753. He was happy enough to see Constantinopolitan influence reduced, but had no desire for Byzantine domination to be replaced by that of Lombard monarchs. In his eyes, the Franks were the obvious counterweight to Lombard ambitions, and, because he wanted the Pope’s formal blessing for his action, Pippin was ready to play the prescribed role. Ready, that is, up to a point. In both 755 and 756, he led armies across the Alps and besieged the Lombard king in his capital at Pavia. In 755, Aistulf had agreed to hand over to the Pope control of the territories of the old exarchate and Pentapolis, but, as soon as Pippin left for home, Aistulf went back on his word and launched his own siege of Rome. Pippin therefore returned for a repeat of the same Pavian routine, and a similar outcome. Aistulf agreed to the same deal once again (as did his successor Desiderius in turn in 757), and Pippin returned to Francia where he faced much more pressing issues.

And there, for the remainder of Pippin’s reign, the situation rested. Nothing was done to enforce the agreed handover of assets, and the Lombard monarchy was content broadly to leave Rome in peace, so that a new equilibrium had been reached in post-Byzantine northern and central Italy. This suited Pippin and his magnates who were much more interested in their own affairs. The odd campaign against the Saxons aside (such as in 753 and 758), Pippin’s focus was for the remainder of his reign firmly set on securing his control of Gaul. From 752 to 759, first of all, a series of ultimately successful campaigns were fought to bring the old Visigothic province of Septimania, now under Muslim management, under his control. Then attention turned to Aquitaine, which had been fully a part of the sixth-century Merovingian kingdom, but where a tenacious ducal house had established an independent position in the early eighth century. From 759, Pippin inexorably ground away at Aquitaine, until, after a series of defeats, the then Duke Waifer (grandson of the Eudo with whom Charles Martel had fought at Poitiers in 732) was assassinated by some of his own men in 768. The struggle was fast approaching its endgame, therefore, when Pippin himself died: unexpectedly, it seems, in September 768. This allowed Waifer’s son Hunoald a brief glimmer of hope. Much more important, a new succession saga was about to erupt which would tie the Carolingian line much more firmly to the papacy and eventually generate its claim to the Western imperial throne.18

Pippin’s death occasioned no challenge whatsoever to the Carolingian line from within Francia, power passing smoothly to his two sons Charles and Carloman. But the familial fraternal tradition ran true to form and, within a year, the two were at loggerheads. After an initial demonstration of solidarity, it was left to Charles to finish off Hunoald of Aquitaine by himself in 769, therefore, while, in the meantime, his mother looked to arrange a marriage alliance between himself and a daughter (probably called Gerperga) of the Lombard king Desiderius. Another of Desiderius’ daughters was also married to the Duke of Bavaria, so this marriage looks as though it was meant to stitch up an alliance network of Charles, the Lombards and the Bavarians, which left out – and hence threatened – Carloman in the middle (Figure 12). The two brothers had the same parents, so it was not just stepmothers like Plectrude, it seems, who played favourites. Later sources present Carloman as peevish, self-pitying and easily flattered, so she may have been moved by clear-headed if rather ruthless analysis of which of her sons was more likely to make the best king (the same kind of analysis which – alongside questions of greed and self-interest – also guided the political choices of key aristocrats), but these characterizations all date long after his death. Nor was Carloman the only one to feel threatened. The marriage negotiations also generated surviving letters from Rome, where Pope Stephen was having kittens at the prospect of a marriage alliance between one of the Frankish kings and the Lombards who still hovered so menacingly over papal territories (and who still had not fulfilled any of the promises laid out in the agreements of the 750s).

Whether – or more likely when – fraternal frostiness would have given way to conflict is unclear, because contingency intervened again in the form of Carloman’s early death on 4 December 771. Charles took immediate control of his brother’s kingdom, and it’s perhaps an indication that Carloman was neither loved nor effective that his magnates showed not the slightest inclination to resist. Charles also terminated the marriage negotiations with the Lombards, which makes it pretty evident that the purpose had been an alliance network against his brother, which the latter’s death had now rendered redundant.

The consequences were fascinating. In an extraordinary volte-face, Carloman’s widow, Gerberga (that’s Gerberga with a ‘b’, not Gerperga with a ‘p’), fled for refuge to the Lombard king: so much did she trust Charles with the fate of his nephews and nieces. Desiderius was not only seriously aggrieved by the breaking off of the marriage negotiations and hence willing to receive her, but also clearly decided that all bets were off with the Franks and started, once again, to roll up the papal possessions in central Italy which he had previously left alone at Frankish insistence. He was probably also encouraged in this by the death of Pope Stephen in January 772 and the accession of Hadrian I. The first months of any new ruler’s reign provide an excellent moment for testing his mettle.

Whether, or how much, Charles would have cared about Desiderius breaking out of the reservation in other circumstances is unclear, but he was pressed by powerful dynastic reasons, and, late in 773, led a large army over the Alps. The Lombard king clearly thought that he was in for a replay of 755 and 756 when Pippin had not had sufficient political confidence to stay away from the Frankish heartland for any length of time. All his predecessor had had to do was hide behind the walls of Pavia, agree to a face-saving treaty, and the Franks would depart with no serious harm done. Charles, however, was confident enough to stay: this again indicates that spreading his authority into his brother’s domain had not generated any potentially dangerous political bitterness. The Frankish army besieged Pavia throughout the winter, and Charles made a lengthy visit to Rome. By Easter 774, Desiderius realized that the situation was hopeless and surrendered. He was sent north to become a monk at the monastery of Corbie, while Carloman’s children were handed over to their uncle and mysteriously disappear. No one records where those bodies were buried.19


At this point, we need to start making an acquaintance with some of the ideas behind one of the most important forgeries of all time: the Donation of Constantine. In full and final form, these original ideas and a few more besides would be embodied in a bogus document – the Constitutum Constantini, to give it its formal title – which claimed to preserve the text of an official grant of the first Christian Roman emperor, Constantine I. In it, the emperor grants to Pope Sylvester I and his successors, as inheritors of St Peter, dominion not only over the city of Rome, but also over the entire territory of the old Western Empire: everything from Hadrian’s Wall to the Atlas Mountains of North Africa. Constantine is making this staggering gift, so the Donation tells us, in joyful response to Sylvester’s Christian teaching and baptism, and also because the Pope had miraculously cured him of leprosy. The Constitutum as we have it would be forged in quite another context, as we shall see in the final chapter, but some of the key ideas incorporated into this later fake were alive and kicking in later eighth-century Rome. They first appear in a letter of 778 which came to Charles from Pope Hadrian I, and which survives because the Franks maintained a file of their papal correspondence, which was later written up and preserved as the Codex Carolinus. The letter comes complete with a famous exhortation:

And just as, in the times of the blessed Roman pontiff Sylvester, God’s holy, catholic and apostolic Roman Church was raised up and exalted by, and through the bounty of, the most pious Constantine of holy memory, great emperor, who deigned to bestow power in these western regions upon it, so also, in these most happy times in which you and we live, may the holy Church of God, that is of St Peter the apostle – burgeon and exult and continue ever more fully exalted.

So, no prizes for verbal economy, but that was not the point.

The sudden appearance of this idea set was closely related on certainly one and probably two levels to how Charles had responded to the surrender of Desiderius. Highly contingent family matters may originally have brought the Frankish king to Italy, but, having destroyed the current Lombard monarchy to resolve them, Charles took the opportunity to advance his royal status to an unprecedented level. Rather than finding himself a pliant puppet prince from somewhere among the higher Lombard nobility, Charles proceeded to do something which no one had done for several hundred years: declare himself king of the state his forces had just conquered. He did not abolish the concept of the Lombard kingdom, or many of the details of its separate operation, but he added the new title to his existing one, and from the summer of 774 onwards became king of both the Franks and the Lombards. Many of the old officials and nobles would be left in place, but Charles made sure, in the years that followed, that this happened on his terms. In both title and fact, he added the Lombard kingdom to the united Frankish kingdom that he had created by eliminating his brother’s line.

In this highly charged context, the story of Constantine’s gifts to Sylvester – although some of it at least was several centuries old at this point: the Pope baptizing and curing the emperor are total nonsense but appear in the sixth-century life of Sylvester in theLiber Pontificalis – acquired a new and entirely immediate significance. Since the Lombard extinction of Byzantine rule over the exarchate of Ravenna and the cities of the Pentapolis, popes had been anxious not only that Rome might be next on the Lombard shopping list, but also, in more positive vein, to secure their hold on any old papally owned landed estates within these regions that now fell under Lombard rule, and establish a papal right to some of their old public revenues (generated by such devices as taxes, tolls, percentages of legal fines). It was precisely on such matters that Pope Stephen had extracted promises from Pippin in the mid-750s, but Pippin had never delivered much in practice.

Once his son had made himself king of the Lombards, however, the situation looked much more promising. Charles was in total control of the disputed territories, and in a position to make substantial grants. Nothing happened immediately, though, and, as a result, Hadrian’s letters from the later 770s are full of references to Charles’ need – for the good of both their souls, of course – to fulfil the promises made by his father. In this context, the claim that, back in the fourth century, the emperor Constantine had made such enormous gifts to Pope Sylvester had a straightforward significance. At this point, Hadrian did not have control of the old Western Empire in mind, but something much more prosaic. Having started by referring back to Constantine in such a grandiose manner, Hadrian’s letter then got down to real business. Charlemagne is directly equated to the old emperor as ‘a new Constantine’, and, by dint of the equation, urged to follow the example of his generosity to the Church of St Peter. The Pope here particularly mentioned ‘possessions in the Tuscan regions, Spoleto and Benevento, Corsica, and the Sabine patrimony’, and sent to the king, along with the letter itself, his representatives armed with the appropriate documentation to prove the Pope’s legal rights over these properties in which he was claiming a financial interest.20

In the end, all the pressure paid off, although, to be fair, it took Charles several years to consolidate his hold on his new Italian territories, and he may always have been intending to make the grants that he eventually did. Either way, in 781 and again in 787 he made two block grants of financial rights to the papacy, which gave Hadrian much of what he had been asking for since 774. The original texts of these grants do not survive, but we do have the confirmation of them granted by his son Louis the Pious in 819, and they were seriously generous. Charles did not give the papacy full sovereignty over the old Byzantine territories of the exarchate and the Pentapolis, which it looks as though his father Pippin might have promised, but he did grant Hadrian certain new rights in these areas as well as returning to him many individual papal estates dispersed across these lands. In separate grants, he also transferred to the Pope full sovereignty over new blocks of territory on every side of the city of Rome, mostly in 781, but with a supplement in 787 after the conquest of Benevento. The island of Corsica was then thrown in for good measure (Figure 13). All in all, the king guaranteed the papacy a massive increase in its annual revenues.21

One thing that is always extremely difficult to judge is the degree of mendacity involved in the creation and transmission of a bogus view of the past. Part of the Constantine and Sylvester story – the baptism and cure – was such an old lie by the time of Pope Hadrian I that he probably just thought it true. Whether this was also true of his letter’s extra dimension, that cure and baptism led to a massive transfer of authority to the see of Rome, is impossible to know. In a context where the papacy was having to reconstitute itself in the face of the sudden explosion into Italy of Frankish power, the new connection was so convenient that you do wonder if a papal adviser dreamed it up as Hadrian was busy dictating his letter, four years after the conquest of the Lombards, to provide that extra bit of rhetorical leverage in the face of Charlemagne’s so far stubborn refusal to hand over the Pope’s share of the spoils. But what survives from papal circles between the sixth and the eighth centuries is so incomplete that even this linkage may have been made long before. Either way, the story of Constantine and Sylvester no doubt helped push things along and, three years later, when the papacy eventually acquired its cut from Charles’ great conquest, Hadrian was no doubt duly grateful. There are very good reasons for thinking that, even after the gifts were sealed, some of the other consequences of Charles’ success were still of concern to the Pope, however, and that, on a second level, the recast story of Constantine and Sylvester may have been deliberately framed to address these concerns as well.

These concerns had their roots in the unprecedented grandeur of the position that Charles was carving out for himself within Latin Christendom. By the time he had added the Lombard kingdom to his existing domains, the totality now broadly coincided with the territories of the old Western Roman Empire. Lombard Italy and the united Frankish kingdom (which already included all the core territories plus Alamannia, Frisia and Thuringia) put Charles in direct control of a territorial area which was far larger than anything achieved by even the greatest Merovingian of the sixth century, or indeed any Frankish predecessor. Nor was 774 the end of the story. Control of the two independent Lombard duchies of Benevento and Spoleto was duly added to the mix, and an entirely unprovoked intrusion into Bavaria in 787–8 then brought the last of the old satellite duchies under Charles’ direct rule. In the meantime, a long and bitter struggle was under way to conquer Saxony and extinguish the paganism which prevailed there. This had begun in earnest in 772, and round one lasted down to the mid-780s including both forced conversions and a massacre of 4,500 opponents in 782. After a brief hiatus, rebellion broke out again in 793, but round two was largely limited to northern Saxony, where it came to an end with a final round of mass deportations in 804. But even this doesn’t quite exhaust the list of conquests, since you also need to add a substantial degree of expansion into northern Spain, and the total destruction of the Avar realm between 791 and 796, whose aftermath saw, as Einhard famously reports, a long line of treasure wagons wending its way back to Francia.22

After this astonishing career of conquest, and really in fact from the moment of Desiderius’ surrender, Charles was indeed Charlemagne: Charles the Great, entirely without rival within Latin Christendom. Constantinople’s reach was confined by this point to some scattered territories in southern Italy, and the line of Visigothic monarchs in Spain had been extinguished by Muslim conquest. The fact that Charlemagne wrote some vaguely respectful letters to King Offa of Mercia in the 790s has led some to consider Offa a genuine peer, but this is nonsense. Charlemagne ruled most of Western and central Europe, whereas Offa’s reach extended to overlordship of the southern counties of England (Figure 12).23 There was just no contest.

What concerned the papacy was not the bare fact of Charlemagne’s continued successes, two of which – the conquests of the Saxons and Avars – anyway advanced the boundaries of Christendom by extinguishing different brands of paganism. The problem lay in the significance that Charlemagne and his circle of advisers – lay and clerical – might attach to them. As the ruler of more than one separate kingdom, and the unifier of pretty much all of Latin Christian Europe, Charlemagne was so obviously much more than a king that the title ‘emperor’ was bound to come into view at some point. When precisely it did so for Charlemagne himself is a tricky question. By the 790s, the evidence is explicit that the title of emperor was being discussed in Charlemagne’s court circle as the only one appropriate for his new status. Both the specific title of emperor and the more general concept of empire crop up in the writings of court intellectuals at this time, but that doesn’t mean that these ideas hadn’t been doing the rounds for some time already.

My own hunch would be that the conquest of the Lombards was the real game-changer. Once king and advisers had time to think about that unprecedented achievement in any kind of seriously reflective manner, the word ‘empire’ would quickly have become unavoidable for an entity which now encompassed more than one kingdom. And even more important than what Charlemagne had created, was a related question: why had he been able to create it? Given early medieval Christian understandings of historical progression, there was absolutely no choice at all when it came to answering this question. Charlemagne had been able to create his multi-kingdom empire because God had willed it to be so; he was God’s agent at work in the world.24

Such ideas had long allowed emperors and kings to interfere in ecclesiastical matters when they wanted to, and, again in the 790s, Charlemagne flexed the muscles of his religious authority in a highly significant way. The opportunity was provided – accidentally – by Constantinople. The steady stream of losses it was suffering at the hands of the Muslims in the first half of the eighth century (the same losses which had allowed the Lombards to walk into Ravenna) eventually generated ideological crisis. If Christianity was the right religion and the Eastern emperor God’s appointee, why was the empire, supposedly supported by an utterly omnipotent Divinity, losing so many battles to a bunch of non-believers? Based on Old Testament example, where the Children of Israel are regularly chastised for falling into false religious practice, there was an obvious answer: the empire was doing something to offend the Lord, who had sent the Muslims as a warning to bring His people back to the paths of righteousness.

What was much more optional was the answer as to what exactly it was that God was finding so offensive. Under the emperors Leo III and his son Constantine V, the Constantinopolitan establishment formally adopted the view, for reasons that still remain obscure, that the problem was the practice of venerating icons – holy pictures. In Eastern Christianity, it was commonly held that icons captured something of the essence of their subjects, and hence could function as a religious hotline to saint, Mother of God, Jesus, or whoever was portrayed. An icon, in other words, was a type of relic and could be used as a portal – a ‘window into Heaven’ – to gain access to the intermediary holy power of its subject matter. As far as we can tell, icon veneration had long been part of Eastern Christianity, but from C.700 the idea took hold that icons were in fact ‘graven images’ as outlawed in the Ten Commandments, and that it was this practice which was making the Almighty so tetchy. Under Leo and Constantine, icons were destroyed and their veneration outlawed, but the doctrine was never accepted in the West, and a succession of popes was happy to condemn the official Constantinopolitan view as heretical. The situation changed again in the later 780s, when a new imperial regime headed by the empress Irene, who had removed and blinded her own son Constantine VI (another example of early medieval royal family love), orchestrated a religious volte-face. A new council was held at Nicaea in 787, and icons – in their full glory as religious powerhouses – came back into fashion. When the news came west, Pope Hadrian declared himself overjoyed and celebrated Constantinople’s return to the religious fold.25

Charlemagne and his churchmen took a different view. After careful preparations, the king called a major synod at Frankfurt, where a different position from that advocated by Hadrian was hammered out, right under the noses of the Pope’s representatives. Religious pictures were perfectly OK, it was agreed, but for instruction only: the Constantinopolitan doctrine of icons as relics of power was rejected. The main target of Charlemagne’s self-assertion here was certainly Constantinople rather than Rome. By this date, talk of empire was rumbling around in Charlemagne’s circle, and, at least at the beginning, the line used to justify a potential reassertion of empire in the West was that Constantinople’s version of it had gone seriously wrong. Religious error was thus an excellent stick to beat the Byzantines with. As the 790s went on, the attack was to be reinforced with the further thought that, given biblical visions of the correct ordering of relations between men and women, then it was clear that a woman – Irene of course – could not possibly legitimately hold the position of God’s emperor: the divinely chosen leader of Christendom. This aggressive stance towards Constantinople demonstrates that the talk of empire at Charlemagne’s court was no mere game being played by his intellectuals. The fact that so much energy went into arguing that the Byzantine version of empire was illegitimate in God’s eyes shows us immediately that Charlemagne himself had his eyes firmly on the imperial prize.26

From the papal point of view, all this encompassed one major problem. As Charlemagne’s behaviour underlined, the only concept of empire available in the late eighth century carried strong connotations of religious authority, by dint of the idea of divine appointment. And, as we shall see in Chapter 7, the king was already exercising authority in the religious sphere, not only at Frankfurt, but more generally. Charlemagne’s imperial pretensions thus meant that the latest heir of St Peter was facing up to a formidable rival for leadership within the Latin, Western Church, and Frankfurt showed that the king had no compunctions about asserting his own religious authority at the expense of that of Rome. This all came into the open in the 790s, but the fundamental early medieval linkage between worldly power and God’s will means that the ideas would have been there from the moment Desiderius surrendered. In the years after 774, therefore, for all that the Roman republic had benefited so materially from his grants of 781 and 787, the papacy cannot but have perceived Charlemagne as a threat to its religious authority, and we have good evidence of a natural determination on its part to impose limits on the king’s pretensions.

One dimension of ideological response to Charlemagne’s challenge is already apparent in the preserved papal letters of the Codex Carolinus from the 770s. Where Frankish sources unanimously attribute their king’s successes to God, the papal letters attribute them to St Peter’s intercession, the payback for Charles’ loyalty to Rome. This line of attack was adopted in part, no doubt, to help extract the best possible deal from Charlemagne after the conquest of the Lombards, since the implication was that, should the king not fulfil his promises, St Peter could take victory away again. But it also implicitly limited Charlemagne’s part in God’s plans by suggesting that the Almighty was not acting directly through Charlemagne but only via the Pope and St Peter. God’s central purpose, in other words, was not to put Charlemagne at the head of Western Christendom, but actually to strengthen the Church of Rome.27

In this context, the ideas that would eventually be written up in the Donation of Constantine have a second dimension of importance, beyond their obvious relevance to papal land-grabbing. Quite explicitly, they also made the case that there was now no justification for having an emperor in the West at all, since Constantine had left dominion over the entire Western Empire to Pope Sylvester when he left for Constantinople. Since Charlemagne had started to look like an emperor as soon as he conquered the Lombard kingdom and soon came to be interested in acquiring the title, and since emperors could by right of divine appointment exercise religious authority on a very large scale, this dimension of the papal argument can hardly be a coincidence. Although the papacy had benefited massively from Charlemagne’s victories, it was clearly also, at the same time, doing its level best to keep his successes from getting out of hand. The Constantine and Sylvester story, together with the determined attempt to attribute everything to the intercession of St Peter, show that Charlemagne’s successes had quickly set alarm bells ringing in Hadrian’s HQ.

We are now, finally, within a whisker of understanding the historical dynamic behind Christmas Day 800. From the moment when he decided to pick a fight over the end of iconoclasm at the absolute latest, and I would suspect considerably earlier, Charlemagne had his eye on the imperial title, with a full understanding of its traditional Roman meaning (one entirely preserved in contemporary Constantinople) as the divinely appointed supreme authority – secular and religious – within Christendom. The papacy had been happy enough to benefit from some of the fruits of his conquests, but did not want any imperial significance to be attached to Charlemagne’s career. This leaves us with two final puzzles. Given that the last thing the papacy wanted was to revive the imperial title, why do we find Pope Leo crowning Charlemagne emperor? And why, despite the overwhelming evidence to the contrary, does Einhard maintain that Charlemagne would never have entered St Peter’s if he had known what was going to happen?


The answer to the second question is straightforward. The only legitimate emperor, according to all the official definitions of imperial propaganda, was one who was divinely appointed. The trick, of course, was how to be sure that any particular individual was divinely appointed, because theory allowed for the contingency that illegitimate, man-made emperors might insert themselves into the role from time to time.

Generally speaking, two telltale signs were used. First, an emperor who was truly there by divine appointment would be successful, since God was indeed an omnipotent deity. And military victory was usually held up as the sign par excellence of true, God-supported legitimacy. Second, an individual truly chosen by divine will could not help but become emperor, whatever they or anyone else did, because God willed it. By straightforward inversion, overweening ambition was a clear sign that the individual in question was entirely unworthy of the job. Thus the emperor Julian famously muttered ‘Purple death has seized me’ when appointed Caesar by his cousin the Augustus Constantius II, a Homeric tag that could be used to signal resistance to unwanted promotion. For the same reason, Julian was extremely careful to establish some more or less plausible deniability when his troops later hailed him as Augustus – in a direct challenge to Constantius’ authority – when all the indications are that he not only knew that his troops were going to proclaim him, but had in fact had his people organize it. The same trope was also popular in the Christian context where it soon became fashionable for any seriously holy individual to be recorded in his Life as having done his damnedest to avoid episcopal office, before finally giving in because there was ultimately no choice.28 And that was, of course, the perfect mark of true divine appointment: if God willed it, there would, in the end, just be no avoiding it.

Einhard was well read in the classics, modelling his biography of Charlemagne quite substantially on Suetonius and displaying knowledge of a wide variety of other ancient texts besides. He certainly knew plenty, as did many of his fellow intellectuals at court, about how one might set about identifying true divine appointment. In other words, Charlemagne and his advisers knew perfectly well that if they were caught with their hands too obviously in the imperial till, then the presentation of their man as divinely appointed would start to ring hollow. This, then, was Charlemagne’s dilemma in the face of a Western imperial tradition that had been broken in the fifth century. As soon as he had definitively smashed the Lombard kingdom, he had fulfilled the criterion of military victory in spades, and the rest of his reign just went on to confirm the fact of God’s support, as one territory after another fell into his lap at the point of a sword. But if, by this means, God was making it clear that Charlemagne was chosen to be His emperor, how could that fact finally be proclaimed out loud if neither Charlemagne nor his advisers were allowed to do it themselves?

They were clearly worrying about this at least from the early 790s onwards, and, as we shall see in more detail in the final section of the book, Charlemagne’s reform programme was by then showing an industrial level of care for the Church, as also befitted the received job description of the divinely chosen Christian emperor. In other words, Charlemagne not only achieved the kind of battlefield success that screamed the word ‘emperor’ to anyone who knew their political theory, but his attention to Christian reform (which I’m sure was entirely genuine, lest I sound too cynical) also made the same point under a different heading. When faced with the evidence that Charlemagne and his advisers had their eyes on the imperial title for some years before 800, it has sometimes been objected that, when he actually received the title, his behaviour did not change in any fundamental ways (except that he ceased travelling so far: a fact that is really attributable to his advancing age). That is to miss the point. Charlemagne had been deliberately behaving like an emperor for at least one decade (and quite possibly two) before Christmas Day 800.

The solution to the other, final problem fell obligingly into the imperial lap in the person of Pope Leo III. Leo succeeded Pope Hadrian I in 795 and soon ran into trouble with factions of the Roman nobility hostile to his appointment. Matters came to a head on 25 April 799 when Leo was seized by his opponents and accused of perjury, simony (the selling of ecclesiastical office) and sexual impropriety. The story put about later was that they had attempted to gouge out his eyes and cut out his tongue, and independent sources confirm that he was at least wounded, but Leo was certainly able to see and speak on Christmas Day 800. Papal sources attribute this to miraculous intervention, but you cannot see without eyes and blinding was not that difficult an action to perform, so there’s more than a whiff of spin in the air. But Leo was certainly attacked and imprisoned, and only got away by climbing a monastery wall to escape into the protection of one of Charlemagne’s representatives in the city. He was then sent north to meet Charlemagne himself at Paderborn later in the same year.

What exactly happened when they met, no source records, and both sides subsequently mounted carefully organized plans of disinformation. But when you realize that Charlemagne clearly wanted the imperial title, whereas Leo would rather have eaten his own liver than see a resurrection of empire in the West, the sequence of events is extremely suggestive. In consultation with his churchmen, Charlemagne declared that no one had the authority to pass judgement on the Pope because he was the apostolicus: the direct descendant of St Peter. He then sent Leo back to Rome with enough military force to put down the opposition, and, just over a year later, Charlemagne himself entered the city in November 800. At the start of December, a council was held which reiterated the position that no one had the authority to judge a pope. Nonetheless, on 23 December, supposedly of his own volition, Leo swore a solemn oath that he was innocent of all charges, and two days later proclaimed Charlemagne emperor.29

At this point, and in the face of the deafening silence about what was said at Paderborn, I’d like to introduce you to a principle of one of my doctoral supervisors. He is an extremely distinguished ancient historian, who ended up studying the late Roman Empire but with that suspicion of Christianity that is shared by many (though not all) of those who study the classics. He once famously said, ‘I may not know much about Christianity, but I do know the smell of rotting fish.’ The meeting at Paderborn is a case in point. We will never know how Charlemagne’s offer was worded, and who made it (whether it was direct, or whether one of the king’s people went for a walk in the woods with one of Leo’s). Nor will we ever know if Leo agonized long and hard, or quickly accepted. But there is not an iota of doubt in my soul that, one way or another, Charlemagne effectively said, ‘Make me emperor and I’ll put you back on the papal throne, no questions asked.’30

From Charlemagne’s perspective, there could be no better solution. What could be more perfect than getting the Pope to declare out loud what Charlemagne’s staggering list of military successes and assiduous care for the Church had been demonstrating for the last twenty years: that he was – without the slightest doubt – God’s chosen vessel for the restoration of Christian empire? No one had any doubt that a particular authority attached to the papal see, because its occupants were the descendants of St Peter, even if there was some disagreement about what that authority might mean. So the Pope was a good solution to Charlemagne’s problem, and one that was made even neater by the claims laid out in the ideas that would be written up later in the Donation of Constantine. Its key claim – that Constantine had given imperial authority in the West to Pope Sylvester when he caught the chariot for Constantinople – had already featured in some of Hadrian’s correspondence to Charlemagne. Totally delicious! Who better to proclaim Charlemagne emperor than the latest occupant of the throne to whom imperial power had been left back in the fourth century?

As is always the case with shotgun marriages of the diplomatic kind, both sides quickly publicized their own versions of events. Leo, famously, erected a mosaic in the great hall of the Lateran Palace (at that point papal HQ rather than the Vatican) which depicted Charlemagne and Leo kneeling side by side at the feet of St Peter. The original was destroyed but a copy was made and can now be seen on the exterior of St Peter’s (Plate 17). This iconography explicitly denied that Charlemagne had any greater authority than the Pope, and was exactly of a piece with the papal insistence in the letters preserved of the Codex Carolinus that Charlemagne’s victories came at the behest of St Peter; that is, not from God.

Charlemagne’s take was rather different. All the Pope had done on Christmas Day 800 was declare the crushingly obvious: God (the main man) not St Peter (his deputy) had chosen Charlemagne for unique imperial authority, because Charlemagne’s merits were clearly worth it. Part of the delay between the meeting at Paderborn in the summer of 799, and the coronation which took place over a year later, was to allow time to generate the appropriate level of diplomatic fanfare (although Charlemagne also needed to arrange affairs in the north with a view to his lengthy impending absence). Amongst other exotica, Charlemagne arranged that a deputation from the Patriarch of Jerusalem should just happen to be in Rome when he arrived, to present him with the keys to the city of Jerusalem, the holiest Christian city of them all. When, in due course, he considered the time ripe to pass on the imperial torch, Charlemagne himself performed the ceremony for Louis, his son and heir, entirely by himself, with not even a walk-on part for the Pope, since a sitting God-appointed emperor could certainly, without any imputation of hubris, designate his successor.31 What happened to this empire, on which Leo’s problems had set the final seal, and how Louis and Charlemagne’s other successors fared, is the subject of the next chapter.

If you find an error please notify us in the comments. Thank you!