Charlemagne’s Sons and the Problems of Royal Succession

The Coronation of Charlemagne, fresco by Raphael at the Apostolic Palace, c.1516 / Wikimedia Commons

The problem of Early Medieval royal succession in different circumstances.

By Dr. Elke Ohnacker
Professor of History
Universität Konstanz


The article is concerned with the problem of Early Medieval royal succession in different circumstances: the death of two of Charlemagne’s designated heirs in 810 and 811, the succession of Louis the Pious in 814 and the conflicts between Louis and his sons resulting in Louis’s deposition in 833 and the division of the Carolingian Empire. Counterfactuals are employed in the interpretation of the events surrounding and leading up to the central political and legal problems of royal and imperial succession. Asking questions like “What if … event x would not have taken place?” and – if possible – developing likely and less likely scenarios proves to be a valuable tool of historic research, especially with regard to the Early Middle Ages’ grave lack of written sources. The overall effect of a methodic use of counterfactuals in this form is a deconstruction or what still may be seen as a “logical succession” of events.


Louis the Pious / Wikimedia Commons

On January 28th 814, Charlemagne died at the age of 72. His son Louis the Pious succeeded his father into kingship and empire. Long before Louis died in 840, the kingship had been contested amongst his sons. This conflict peaked in the deposition of Louis 833 and, later, in the division of the Frankish empire formulated in the treaty of Verdun, in 843. But, taken the collective mentality of Early Medieval Elites, Louis the Pious’s sons were neither particularly greedy nor exceedingly belligerent. Rivalries between brothers, fathers and sons, the offspring of women married to or allied with the same king, etc. were omnipresent in these times. Conflicts were violent and frequently resulted in war, political assassination, the disfigurement of opponents, banishment into monasteries and exile, etc.[1]

Amongst Charlemagne’s sons, four were treated as potential successors, Pippin the Hunchback, Karl the Younger, Karlmann/Pippin[2] and Louis. Apart from Pippin the Hunchback’s revolt, the written sources do not mention conflicts between these brothers, them and Charlemagne’s other sons, or between father and sons. But the written sources we have got are few and often biased. In historical situations which are but sparsely documented by sources of any kind, it is necessary to analyse these sources closely. Comparing the issues in question with what happened before and after can lead to a more detailed view. Furthermore, questions which are capable to break open the often fragmentary chronologies of events which are typical for the Early Middle Ages, like the question “What if … event x would not have taken place?” or: “… would have taken place at an earlier or a later date?” offer a wide range of possibilities to analyse certain phenomena more closely. The development of alternative scenarios can as well reduce the risk of being trapped in teleological interpretations suggested by a chronological succession of events. In the perspective of the historian it is imperative to keep in mind that a given succession of events need not be the result of causality. Thus, in order to answer the question “What if … Karl and Karlmann/Pippin had survived their father?” we have to include two preliminary steps: analysing the characteristics of the written sources and the information they give and comparing characteristical problems of kings, their sons, and the conflicts about royal succession reaching back into Merovingian times.

Sources and Methods

Alcuin (pictured center), was one of the leading scholars of the Carolingian Renaissance. / Wikimedia Commons

Research on the history of Western Europe in the Early Middle Ages is subject to a crucial problem: the lack of writing. During the rule of Charlemagne and his successors, the developments in culture and administration, later summarised under the term “Carolingian Renaissance”, resulted in the production of more written sources. This does not mean many. Research on events in high politics suffers gravely from this problem, let alone the history of the common people who were still further away from the focus of these sources. Interpretation is made still more difficult by the fact that an overwhelming majority of their authors were clerics, dealing out harsh criticism on the lay society surrounding them, which did seldom conform to Christian norms and ethics. If a ruler like Charlemagne was interested in introducing different forms of written documents as new techniques of rulership and administration, it is to be expected that these written sources – be it records or narrations – were produced by men who worked and wrote in the ruler’s sphere of interest and often owed him their splendid careers. Modern categories like “objectivity” and “plurality” (of interpretations) are not to be looked for in these texts.

Thus, a critical methodological approach towards such sources is imperative. The long-term development of techniques of interpreting medieval sources comprises historiographical as well as semantic and semiotic methods, interdisciplinary approaches, etc. Many of these methods, like the Annales-School’s integration of interdisciplinary concepts into mediaevistic research, have fiercely been disputed in their time (Raphael 1994, 59 ff., 393 ff.). However, counterfactuals and counterfactual thinking are rarely perceived as scientific tools enlarging the stock of refined and interdisciplinary methods. Although the question “What would have happened, if… ?” might occasionally come up in informal circles, it is rarely asked
seriously, especially not in published research in a market where such an approach might be used to question the author’s scientific credibility. In spite of a wide-spread reluctance, counterfactual thinking can be a valuable tool of mediaevistic research, especially for the Early Middle Ages.

Marc Bloch encouraged the historian to always be attentive to what might be hidden under the seductively evident surface of sources, “facts” and “common knowledge” (Bloch 1974, 151 ff.). Counterfactuals can invert what we think to know and thus improve our insight into historical phenomena. How- ever, to avoid “wild” speculation leading to doubtful results, counterfactual thinking requires a critical approach. Questions like “What would have happened in the case of an alien attack on Christmas 800?” (when Charlemagne was about to be crowned as emperor) are only slightly less to the point than asking: “Was Charlemagne French or German?”, a topic raised during the nationalist euphoria in 19th century France and Germany[3] and reaching another infamous peak in Nazi Germany (Hampe, 1935). Charlemagne, who lived in an age when extraterrestrial activity was strictly limited to agents of the Christian god, and when the modern states of “France” and “Germany” were not yet existing, would not have understood either question.

The condition for a historian asking counterfactual questions and, at the same time, avoiding anachronisms and ethnocentrisms resulting from a modern point of view, is the thorough knowledge of the historical society in question, its structural characteristics, social practices, norms, forms of knowledge and understanding etc. The characteristics and many-layered problems of royal succession in Early Carolingian and Merovingian society form a field where counterfactual questions can lead to interesting insights. The kingdoms of Early Medieval Western Europe, different as they may have been in many aspects of their development, show certain common characteristics concerning rulership and royal succession. Legitimate as well as illegitimate sons might succeed to the throne (Hagn, 2006, 192 ff.). Having several kings and dividing kingdoms was a widespread phenomenon, especially in the Merovingian realm (Schneider 1972, 73 ff.). Having designated successors acclaimed as (sub-) kings, once they came of age (at the age of c. 15), and giving them military commands of their own was common practice. The king’s sons tended to be in a privileged position concerning royal succession but were not the only candidates possible. Marriages and re-marriages[4] amongst the elites, as well as endemic poligamy, frequently led to relatively large groups of potential candidates, each of whom was claiming his superior right to the throne. The chaotic feuding which could result from such constellations is vivid in Gregory of Tours’ and the Pseudo-Fredegar’s descriptions of Merovingian society in the 6th, 7th and 8th centuries. It is easy for the modern observer to interpret such a description as destructive, as typical for the “dark ages”, or even as agonal (Scheibelreiter 1999, 236-237). Here it is important to overcome the rhetoric of Roman civility displayed by a clerical member of the Gallo-Roman aristocracy who, like Gregory, described the new barbarian lords’ politics. The catalogue of political deficits of a degenerating Merovingian kingship that the author of the Pseudo-Fredegar’s continuationes emphasises[5] draws a different, but equally biased, picture. But, violent as these conflicts might have been, having several candidates for the throne served an important social purpose: supplying a surplus of qualified potential successors in a time when people died young and when a king’s offspring might not survive him.

There were several strategies of limiting down the number of candidates for the throne. Recourse to violence was frequent. Chlovis I., the first Catholic king of the Franks, considerably expanded the Merovingian realm and secured his single kingship, as well as the division of the realm amongst his sons according to his will by killing off whoever might have stood in his way.[6] Apart from assassinating competitors, exiling them or banning them into monasteries and cutting their hair, thus depriving them of their symbol of rulership, the crinis,[7] were frequent methods of doing away with less successful candidates. Another efficient corrective concerning the number and qualification of candidates was to be found outside the King’s family (stirps regia). Even powerful kings were not capable to rule alone but, to a considerable extent, depended on aristocratic co-operation and on their ability to create consent between them and the aristocracy. Claiming the kingship depended on aristocratic support. Candidates who did not succeed in pulling a significant part of the kingdom’s magnates on their side were out of the game.

Pippin III., Charlemagne and Their Unsuccessful Brothers

Pippin III / Wikimedia Commons

When Pippin III. deposed the last Merovingian king Childeric III. in 751, he did away with a dynasty, not with practices of ruling and ensuring power. Political opposition was eliminated by the usual methods: forced conversion to monasticism, imprisonment (often in out-of-the-way monasteries), exile, blinding and murder. In Pippin’s family, the office of mayor of the palace was contested between the different heirs. Karl Martell had three sons: the brothers Pippin and Karlmann, sharing their father’s office of mayor of the palace, and Grifo, a younger brother who had inherited Aquitaine. In 741 Pippin and Karlmann joined forces to deprive Grifo of his inheritance (Fredegar, Cont.a.741). By banning his brother Karlmann (who died in 754 under suspicious circumstances (Nelson 2004, 103)) into a monastery at Sorakte near Rome, and later Montecassino, he secured the succession of his sons Karl and Karlmann. After Pippin’s death in 768, Charlemagne and his brother Karlmann found themselves in the same situation. The rivalry between the brothers seemed to have been deeply rooted and omnipresent,[8] only ending with the death of Karlmann (771). Given the ruthlessness of Charlemagne’s actions against aristocratic opposition, it might be guessed that Karlmann was spared his uncle’s fate by early death.

These were not the only similarities between father and son. Charlemagne as well as Pippin III. were bent on improving their status and political power: Pippin rose from a high royal office to the kingship, Charlemagne was crowned emperor in Rome (Christmas 800). Given the social and political climate of the time, heirs might be expected to try and get more than their share and not be too scrupulous about it. But, would the events of 751 and 800 have been possible if both Pipppin III. and Charlemagne had not first ensured that they held their position alone? Given the problem of aristocratic support, the legal aspect that the emperor’s title could not be as easily divided as a kingdom, and, last but not least, the considerable risk of either action, both events would probably not have taken place had the respective brothers kept their share of the inheritance.

Charlemagne had eight wives, numerous concubines and surviving children from most of these alliances. Eginhard describes his great attachment towards his children, especially towards his daughters, in terms exceeding the topical formulae reserved for the ideal prince. The education of the royal offspring, legitimate or illegitimate, seems to have been thorough. Unlike their father and most members of the lay elites, sons and daughters were taught to read and write and were educated in the liberal arts. Quite a-typical, Charlemagne did not wish his daughters, excepting the eldest Rothrud,[9] to marry and had them stay in his household (Eginhard Vita, 19). Apart from the incestuous insinuations this statement might awake, this information can also be explained differently: not allowing his daughters to marry and keeping them in his household, can also be seen as a means of narrowing down the number of potential successors – another question for which counterfactual thinking might prove new insights. The question “What if … Charlemagne’s sons-in-law and their off- spring had interfered in the politics of the realm?” clearly evokes even more rivalries, conflicts and political problems. To take this question further, one might well contemplate what would have happened if the planned marriage of Rothrud and the Byzantine Emperor had taken place. On the one hand, such an alliance might have been advantageous for both emperors, but, on the other hand, for this very reason such a match would probably be bitterly opposed by most of the Byzantine as well as the Frankish elites. This marriage would have happened before 800[10] when Charlemagne was crowned emperor and deeply disturbed the diplomatic relations to Byzance. Would he, after this marriage, have claimed the emperor’s title as well? And if so, what would have been the politic consequences? The answers to these questions tend to be very speculative, but an event which is seen as a crucial turning-point of European History (and which is still far from being sufficiently analysed)[11] might not have taken place. The resulting question “What would have happened, if the translatio imperii of 800 had not taken place?” is too complex and far-reaching to be answered even in an essay about counterfactuals.

The following sons of Charlemagne are mentioned in the written sources: Pippin the Hunchback, Charlemagne’s oldest illegitimate son, Karl the Younger, Karlmann/Pippin and Louis[12] (sons of his second wife Hildegard), as well as three other natural sons, Drogo, Hugo, and Theoderic. Despite their illegitimate birth, the natural sons also carried dynastic names,[13] an indicator for their not being principally excluded from succession. Out of these sons, Pippin the Hunchback was the only one to revolt outright against his father in 781/82. This revolt illustrates clearly the deficits of contemporary sources, the lack of our knowledge and the wide potential of “educated guesses” which have to be employed to achieve some more insight. Eginhard tells us that Pippin, being seduced by “idle promises of the kingship” (Eginhard, Vita, c. 20), conspired with several unnamed magnates to depose his father. The conspiracy was discovered, its leaders punished and Pippin, due to his father clemently revoking capital punishment, was sent away to the monastery of Prüm (Eginhard Vita, c. 20). Like so often, we do not know what became of him. Only one relatively late source mentions him giving advice to his father about how to deal with other conspirators (Notker, Gesta II, 12). Charlemagne’s clemency in revoking capital punishment cannot be seen as a singular event. Suspending verdicts of capital punishment and clemently receiving back most of the perpetrators into royal grace was a frequently practised ritual in medieval politics (Althoff 1997, 99 ff.).

When it comes to conspiracies against the king and the parties involved therein, showing clemency and sparing most of the participants might also have been a political necessity given the power and influence of the magnates involved. What was clothed in the de-legitimising terms of rebellio, coniuratio, conspiratio, etc. (Brunner 1972, 14 ff.) could have been more seriously challenging the ruler’s status than the sources want to make us believe. This is
especially important with regard to contemporary Carolingian sources written in the Carolingian kings’ chancellery and thus being far from neutral. Also, in these cases, the question might be asked: “What would have happened, if the king had not shown the expected clemency?” Could this have resulted in more uprisings amongst the magnates? Who the magnates of Pippin’s party were, what became of them after having been punished or received again into Charlemagne’s graces, what exactly their expectations in the case of Pippin’s success might have been, is not known. Charlemagne’s motives of excluding a son from succession whom he had already been treating as a successor (Kasten 1997, 139 ff.) are far from clear, too. That Pippin had been treated as a potential successor is proved by several donations which are consistent with what we know about how successors were provided for. Thus, Pippin’s position seemed to have been based on more than “idle promises”. Kasten suggests that a disease resulting in the deformation of Pippin’s physical appearance at a later age might have been responsible for his exclusion from succession (Kasten 1997, 139 ff.). Physical health and beauty were a basic condition for a ruler’s idoneity, a physical handicap thus might have been a reason for exclusion.

However, on a praxeological level, Pippin does seem to have found allies amongst the magnates. Would this have been the case if his ability to become king had been principally scrutinised by his handicap? An indicator supporting this opinion is found in Eginhard’s Vita: Pippin is described as hunchbacked but having a beautiful face (Eginhard Vita, c. 20) – obviously Pippin was not completely devoid of the physical attributes of kingship. In this case, the question “What if…?” might lead to a more detailed evaluation of the gap between
theory and practice of rulership. “What if … Pippin had not had this physical handicap?” Presumably, Pippin might have been treated as a successor in line with his three younger brothers. This would, in turn, have resulted in an increased number of rivals. Four potential successors might, furthermore, have brought about even more violent conflicts because the common and convenient practice on such cases, i. e. forming an alliance with one rival to overpower the third party would not have worked any more. “What if … Pippin and his party had been successful?” is a still more relevant question. Could this have resulted in the deposition of Charlemagne or in forcing him into a compromise concerning his succession? Another possibility is that successful action against Charlemagne might have had a domino effect on other groups of the realm who were dissatisfied with their ruler. That these groups existed is beyond doubt.[14]

The Sons of Hildegard

Hildegard from Promptuarii Iconum Insigniorum / Wikimedia Commons

Charlemagne wished his three sons by Hildegard (Karl the Younger, Karlmann/Pippin and Louis) to jointly inherit the Frankish Empire. That the division of the Empire during his life time could severely threaten his own power, he would have known only too well, given his own family history. Trying to make this division as water-proof as possible hardly comes as a surprise. According to Frankish legal tradition, the division was proclaimed and accepted publicly in 806. Furthermore, a written version of the proceedings (the divisio regnorum) was compiled and one copy sent to the Pope. Before this date, the three brothers had already independently participated in politics, especially Karl who, since 784, had held his own military commands, mostly against the Danish invaders (Kasten 1997, 50, 55 ff.). In 806 all three were
proclaimed sub-kings: Louis of Aquitaine, Karlmann/Pippin of Italy, Karl receiving the lands between Loire and Elbe. But the divisio regnorum avoids the central question. What was to become of the emperor’s title? Although the divisio speaks of empire and kingdom (imperium vel regnum) (Divisio regnorum, c. 5, 128), no solution of this problem is brought forth. There was no precedent to this central legal issue and, moreover, the conflict with Byzance stemming from Charlemagne’s coronation 800 was still pending.[15] Counterfactual thinking concerning this difficult political and legal issue might involve questions concerning “What if … there had been a reaction of Byzance?” In the case of the Byzantine emperor denying Charlemagne the right to the title, what would have happened? In such a case, the legal experts of both sides would have been busy, each denying the other side’s legitimacy. Solving the problem by war would have been risky for both parties, given their respective size and power. Furthermore, such as far-reaching conflict might, as a side-effect, been utilised by different parties for their own ends. A more practical solution might have been to simply ignore the Byzantine emperor’s votum. Also, a third power would have to be considered – the Caliphate. Not in vain had Charlemagne sent a legacy to the court of Harun al-Rashid before 800. Would Harun-al Rashid have stayed indifferent, given the case of a hypothetical war, or have sided with either the Eastern or the Western emperor? Both alliances imply a considerable risk for all participants. But, staying neutral in the case of a war between the Eastern and Western emperor would have been difficult, especially with regard to all party’s economic interests in the Mediterranean region being at stake.

The second set of counterfactuals is focussed on the problem “What if…there had been a legal solution concerning the emperor’s title (nomen imperii)?” There might have been two possible solutions: giving the title to one of the emperor’s sons, thus proclaiming one privileged heir, or, forfeiting the title. Giving away an instrument of power once acquired hardly fits in with the collective mentality of Medieval rulers, although there have been serious doubts concerning the impact of the concept on the Frankish elites (cf. Boshof 1996, 86, 88; Fichtenau 1949, 55ff.). Thus, in spite of all legal concepts, the nomen imperatoris might have been divided. Although the pragmatic adaptation of new concepts into existing political practice is a common feature of Medieval society, this would be a far-reaching speculation. Raising one of the three sons into a privileged position would have been dangerous and quite likely have ended in an alliance of the two excluded sons against their brother and/or their father. Dissatisfied factions of the empire’s magnates would have utilised the conflict for their own ends. This would have meant even more military conflicts in a time when the frontiers of the Frankish empire were not uncontested.[16] Leaving the question open might have been the most sensible course of action.

Crisis and Decline

The years before Charlemagne’s death in 814 somehow blemish the image of the “great ruler”. Apart from personal losses, much within and outside the realm went amiss. In the narrative sources, an atmosphere of impeding change (or even brooding disaster) is created by the mention of an extraordinate number of supernatural signs. The Vita Karoli Magni contains a moving but nonetheless gloomy description of an old king shortly before his death, losing the
symbols of kingship during a fall from his battle-horse.[17] The question if astronomical phenomena like eclipses or comets were, at the time, seen as foreboding unlucky times cannot be attended to here. But there were in fact grave problems which could lead to gloomy narrations. In 807, due to unfavourable climatic conditions, a famine broke out which exceeded the boundaries of commonplace local misharvests. While regional famines and diseases were
common enough and did not bring about political change, overregional events of this sort might provoke problems and push over the established order (Schubert 2006, 38 ff.). Groups or individuals in conflict with the ruler might seek advantage therein. Furthermore, the Danish invasions reached another peak.[18]


The divisio regnorum was rendered invalid by the deaths of Karl the Younger and Karlmann/Pippin in 811 and 810.[19] Given the importance of securing royal and imperial succession, it is surprising that Charlemagne did not directly react to the changed conditions. Only in September 813, a group of magnates led by Eginhard convinced him to have his son Louis crowned emperor (Boshof 1996, 86 ff.). Charlemagne, having been ill for some time, died shortly afterwards on January 28th, 814. The reasons for this temporal lapse remain unclear. Albeit the Western Emperor’s title was not acknowledged by Byzance until 812, when Charlemagne was proclaimed Basileus by a Byzantine embassy (Classen 1986, 602 ff.), waiting for the acceptance of Byzance does not seem to be the main reason. Given the usual dynamics of his politics, it might be guessed that Charlemagne would not have let 12 years pass without pressing the political solution of a problem he deeply cared about.

The coronation followed the Byzantine rite (Boshof, 1996, 88), the son being crowned by his father. The lay magnates and high clergy present conceded and acclaimed the new emperor according to Frankish legal tradition. After the coronation, the participants attended mass together. There is no mention of big festivities afterwards, as might be expected (Becher 2000, 116). In such a case, pomp and ostentation accompanying political acts cannot be seen with the eyes of the modern observer for whom they are rather unspecified and unclear “symbols of status” with no legal meaning of their own. Legitimacy, especially in the important political contexts of royal succession, did not, like today, base on the written word as the vehicle of “right” procedure. Legitimacy was produced by the “right way” of doing something, in this case an inauguration. Even if most ritual details of an inauguration in the 9th century remain unknown (Schramm 1954, 311), the greatest ostentation and the largest audience possible would have to be expected. Early medieval ostentation had a crucial legitimising function. Sparse pomp could easily mean sparse legitimacy. That the lack of ostentation could be due to Charlemagne’s poor health thus seems highly improbable. But, could pompous festivities have provoked those who were dissatisfied with the course of events? If so, the fear of powerful oppositional groups must have been significant.

Another aspect deviates strikingly from the ordinary: After the ceremony, Charlemagne sent his son back to Aquitaine despite being gravely ill. The expected behaviour in such a case would have been to stay close and be present at his father’s deathbed. This has been seen indicating a strained relationship between father and son (Boshof 1996, 86 ff.). There are more hints that Louis’s and his father’s characters and likings were quite different and that they did not like each other too much.[20] Again, the analysis of these a-typical events in an issue of central political importance can be furthered by employing counterfactuals. “What if … the coronation had taken place as was to be expected and Louis had stayed at his father’s side until he died?” Both aspects hint at deeply-rooted problems concerning Louis’s succession. Even as the contemporary sources, as usual, stress the harmony of all parties involved, the issues they leave out are telltale enough. What happened to the followers of Karl the Younger and Karlmann/Pippin after 810/811? Where were Charlemagne’s natural sons Drogo, Hugo and Theoderic?, and: which were the interests of Charlemagne’s powerful advisors, his cousins Adalhard of Corbie and Wala? The answers to these questions can only be guessed. Moreover, given the contemporary social functions of ostentation, conspicuous consumption and the
production of political consent: might the ostentation to be expected in a “usual” ceremony have led to problems with one of those groups? Might the invitation not have been followed by all too many of the realm’s magnates, who, when in discord with the ruler and his policy, would have recurred to the general practice of showing dissent: staying away (Althoff 1997, 60 ff.)?

Charlemagne’s Later Years: The Discrepancy of Research

Charlemagne’s sarcophagus / Wikimedia Commons

Louis the Pious succeeded his father in 814. But, although Thegan claims that the succession was uncontested (Thegan Vita, c. 8), the Astronomus’ Vita hints at a dangerous situation when the loyalty of Wala and others after the death of Charlemagne seemed far from certain.[21] The description of the events the Astronomus gives us is a typical example for the style of the written sources of the time. Here, counterfactual questions are a valuable tool of analysis. The
counterfactuals to be employed here are “What if … Wala had succeeded in pulling the majority or a significant part of the Frankish aristocracy to his side?”, or even: “What if … the magnates had not accepted Louis by granting him the ceremony of adventus, one of the central symbols of a ruler’s legiti-macy, or, if they had ignored Louis’s orders to wait for him where they were?” Most likely, a crisis of state would have emerged, which would have been all the more severe, as the designated ruler had not yet been inaugurated. This conclusion is consistent with the harsh measures Louis took against Adalhard of Corbie, Wala and his natural brothers.[22]

The rule of Louis the Pious has long been seen as an age of decay. Charlemagne’s life work, the Carolingian Empire, broke apart under an unfit ruler and his feuding sons. Louis, after practically having been deposed by being left by his troops on the battlefield against his son Lothar in 833, was subjected to penance (thus being deprived of his idoneity) by his clergy. The infighting of his sons and their followers broke out because of his second wife Judith and her role in overthrowing the Ordinatio imperii of 817[23] to include her son, the later Charles the Bald, as a successor.[24] Recent research (Boshof 1996, 256 ff.), as well as the older works of Ganshof (Ganshof 1971, 240 ff.) and Fichtenau (Fichtenau 1949, 211 ff.) have contributed to the rehabilitation of Louis. There is an increasing number of historians[25] who, in their analysis of Charlemagne, differ from the image of the “Beacon of Europe” Pharus Europae (Poeta Saxo, 226). Obviously, there was a lot amiss in his father’s realm when Louis succeeded in 814. Here, the central question is: why did Charlemagne’s sons not revolt against their father like his grandsons did against theirs, or, like grown-up sons were known to do? Of course, the Carolingian sources play down each and every action against or dissatisfaction with the emperor. This does not necessarily mean that the coexistence between the three brothers Karl, Karlmann/Pippin and Louis has been harmonic. Moreover, there are only contemporary biographies of Louis the Pious which (despite all formalisms employed therein) outline the personality of their object. Due to the lack of sources, his brothers do not surface as individuals. Were Karl and Karlmann/Pippin too involved in warfare in their respective subkingdoms to undergo the military risk of revolting against a father who had proved extraordinarily successful in this field? Did the wars in Italy and against the Danes provide them with enough prestige, so they could quietly await the demise of their old father? The considerable risk of revolting against an emperor known for his ruthless politics could be avoided by simply waiting him out. Formulating counterfactual questions from the angle of Louis the Pious’s sons illuminates another aspect. “What if … Louis the Pious’s sons had had only a short time to wait for the demise of an old king?” They might or might not have taken action against him. And – reversing the point of view once more: “What if … the sons of Charlemagne and Hildegard had suddenly been faced by a new-born brother whose powerful and politically well-connected mother was fixed on throwing over the existing testament for her offspring?” They would have been far from content and in all probability would have tried to reinstate the status quo ante. Obviously, there was much more at stake for the three older sons of Louis the Pious than for his brothers before 810/811.

Again: “What if…Charlemagne’s other two sons had survived their father?”

Charlemagne was powerful enough to ensure the Divisio regnorum of 806 to be respected. But, had Karl and Karlmann/Pippin, with or without Louis, decided to utilise their father’s old age and declining health to deposit him, they might have stood a chance. In such a scenario, the magnates could be reasonably expected to side with the powers-to-be. The risk of such an action could have been minimised by a joint course of at least two of the brothers. Individual
action of either son would rather have resulted in infighting. But most conflicts emerged after a king’s death, when alliances were overthrown and many loyalties questioned.[26] At this point, several scenarios seem possible or even probable. Karl and Karlmann/Pippin, like their ancestors Pippin III. and Karlmann might have joined forces to deprive Louis of Aquitaine, only to transfer the latent structural conflict between the two of them to a later date. The outcome of this hypothetical conflict between the remaining two brothers would have depended on how many powerful magnates each of them would have been able to pull over to their side – and keep them there.

The probability of scenarios depends on profound analysis of political practices in the eighth and ninth centuries. Before this background, some scenarios seem quite impossible. Despite the threat the Danish invasions posed at the end of Charlemagne’s reign, a Danish conquest and the resulting re-distribution of the political chances of the aristocracy appears highly improbable. Also, direct interference of Byzance in the questions of royal succession, like under 6th and 7th century Merovingian kings, seems out of the question despite the undoubted affront of Charlemagne’s claim to the title of Roman Emperor in 800. This case, like the scenario above, would lack the conditio sine qua non of Early Medieval rulership: it had to be based on the consent of a significant part of the magnates, be they clerics or laymen. On the other hand, hard conflicts about the succession between brothers are highly probable. This conflict might either result in eliminating one rival and his family, or even in breaking apart a hugely expanded empire which had already become increasingly difficult to rule under a charismatic leader, let alone his sons. What the role of other relatives, for instance Charlemagne’s powerful cousins Wala and Adalhard, might have been in these conflicts, can only be speculated upon. Leaving aside all further guesses, if Charlemagne’s two other sons had survived him, bets are solid that they might have been the ones bringing about the division of the Carolingian empire.

The example of Charlemagne’s succession has shown clearly that, in addition to other (interdisciplinary) instruments of mediaevistic research, working with counterfactuals can drastically improve scientific results. The counterfactual questions asked, often reached a point where, due to the lack of sources, definite answers cannot be given. But, developing probable or less probable scenarios, or even formulating counterfactual questions from different point of views, has changed the interpretation of events. The formulation of counterfactual questions, like the formulation of any research question without extensive knowledge of the subject, can, however, easily lead into anachronistic speculation. This happens when the researcher takes modern categories for granted and intemporal. As a result, Charlemagne is integrated into a national background which had not existed in the 8th and 9th centuries by scientists who take nationalist categories for being self-evident. On the other hand, the very absurdity of scenarios based on clearly anachronistical questions, can serve as a safe- guard against biased thinking.

This might be illustrated by yet another example comparing Charlemagne and his son Louis the Pious: the use of writing and written administration. In retrospect, Charlemagne was far more successful a politician than his son. He mastered quite a few highly critical situations and succeeded in fixing his succession according to his wishes. The latter was accomplished during a phase of duress and frequent invasions by an old ruler, all in itself de-stabilising factors. Amongst all the possible interpretations of these problems listed above, written administration, a standard modern category of successful rulership has not yet been mentioned. Modern perception sees literacy as a basic condition for the access to social chances and also of successful administration. Max Weber’s theory of rationalisation would not work without written administration. Charlemagne, unlike his predecessors, furthered the employment of writing and surrounded himself with experts who employed new techniques like capitula, written legal texts.[27] The later term “Carolingian Renaissance” is funded on the novelties in the technique of written administration developed by the experts in Charlemagne’s chancellery and schools.[28] Everything just seems to fit in fine with the modern perception of the value of written administration. But – the output of legal and administrative texts, like capitula, distinctly increased under Louis’s rule. The production of written documents in general, of beautifully illuminated texts, of works of art in general, as well as the ruler’s building activity reached a peak from 814 until 829 (Boshof 1996, 267). On the contrary, even Charlemagne’s friend and panegyrist Eginhard had to concede that Charlemagne’s capitula were “few and incomplete” (pauca et incompleta) (Eginhard Vita c. 29). With regard to the different research opinions concerning the extent and political value of Carolingian writing,[29] counterfactual questions might be formulated like: “Would Charlemagne have suffered severe shortcomings in royal power, if he had not employed written administration?” Here, the answer would rather be in the negative. Although both Charlemagne and Louis tried to establish these innovative forms of government and administration, written documents only could serve as instruments of power as long as their enforcement was guaranteed by other means: the achievement of a consensus between the actors concerned, or the rigorous and violent exaction of the ruler’s will. Upending this question produces a similar answer: “Would Louis have not been deposed in 833, if he had enforced the production of capitula even more?” Again, the answer is no, no text could have made his sons change their behaviour, were it not enforced by direct action. Despite strong efforts to establish written techniques, the impact of writing on the practices of power and rulership in the Early Middle Ages was somewhat limited.


Counterfactual thinking can form a valuable tool for the medievalist, not only in the case of Charlemagne’s succession. Of course, given the little we know about Early Medieval society, frequently there might not be enough information to even formulate speculative answers for these questions. But asking continually “What if … the course of even ts had not taken place like this, or: … other actors had been involved?” etc. can break open what is taken for granted and thus produce new scientific results. Also, counterfactuals are quite a good instrument to prevent ethnocentrisms and anachronisms from distorting hy-pothesis and scientific results. On this level, they can even be successfully employed to evaluate research beginning with research questions, concepts and categories. Apart from these possible uses of counterfactuals for the historian in general, counterfactual questions have been proved as a valid instrument for research on Early Medieval history on two levels. First, the sparse and distorted information the contemporary sources give, needs to be questioned closely. Asking the question “What would have happened if …?”, in addition to other methods, is a scientific tool to effectively break open the sources’ surface of social harmony and largely uncontested kingship. On this level, counterfactuals also can refine other methods of analysis. On a second level, by implying counterfactuals into the analysis of the primary sources, the images historiographic research constructs upon historical society can more closely be evaluated. Counterfactuals become a tool for the research of the History of Science. In the example concerned, the deconstruction of many a thesis of a ruler’s “strength”, which are still to a large extent determining the popular image of Medieval kingship, will be furnished with an additional set of instruments. But, what is more, counterfactuals might also be employed to evaluate the frequent conflicts of research opinions. In this field of research, the value of counterfactuals is still to be explored.


  1. Charlemagne himself was confronted with this problem in his own family history. Not only
    did he have a powerful rival in the person of his brother Karlmann. His uncle, who carried
    the same name, tried to contest his father Pippin III.’s position until he died in 754.
  2. Charlemagne’s second son from Hildegard was first named Karlmann. After the unsuccessful revolt of his elder step-brother Pippin (781-82) he was given the dynastic name of Pippin.
  3. Although the image of Charlemagne in French and German nationalist discourse has become a subject of research in itself (cf. Erkens 1999), popular perception does not seem to have changed.
  4. Divorces seem to have been as common as the existence of concubines or explicit bigamy in spite of religious norms, albeit the clerical authors of our sources, for obvious reasons, tend to play down this part of the social life. The common practice of marrying the divorced or widowed former spouse of a king or royal successor who already had children from this alliance(s) and having children with her makes it appear even more chaotic for the modern observer.
  5. The continationes were commissioned by duke Childebrand, a half-brother of Karl Martell
    and an uncle of Pippin III., and his son Nibe lung. Their strong emphasis on the legitimacy of the power to come is not surprising.
  6. Gregory wrote that Clovis had ordered other kings as well as his close kin to be killed to
    prevent them from competing his status. The latter even ordered a treacherous search for
    surviving kinsmen, under the pretext of being alone and helpless, to make sure that nobody survived (Gregory, Ten Books of Histories/Historiarum Libri Decem, II, 42).
  7. The crinis is a hair-knot worn on the forehead, distinguishing the Frankish king from his
    people. The Liber Historiae Francorum claims that, from the time of the election of the
    (mythical) first Frankish king, the Franks were ruled by reges criniti. (Liber Historiae
    Francorum, 344).
  8. Eginhard, Charlemagne’s biographer, avoids the acknowledgement of a deep-rooted rivalry between the brothers, so unfitting for a panegyric. For him, this conflict was due to the efforts of Karlmann and his followers to disturb peace and unity (Eginhard Vita, c. 3).
  9. Rothrud first was chosen to marry the Byzantine Emperor Constantine VI., a match which
    did not take place. Around 800 she married Rorico, the later Count of Maine (Boshof 1996, 25).
  10. Cf. the previous note.
  11. Cf. Godman et. al. 2002.
  12. Louis being the surviving part of the twins Hildegard gave birth to in 778. The other brother, Lothar, presumably died in 799 (Boshof, 1996, 24).
  13. Cf. Drogo and Theoderic. On dynastic names, cf. Geuenich, 2006.
  14. For example, the deposition of the Bavarian duke Tassilo III. in 788 and the exilation of his
    family enlarged the circles of those who bore a grudge against the king. Also, Charlemagne’s father Pippin III. had made himself a name for ruthlessly getting rid of anybody who stood in his way. The threat arising from dissatisfied magnates must have been considerable, given the rash action of any ruler to suppress what the written sources call rebelliones or coniurationes. The oath of fidelity (789, renewed 802) Charlemagne demanded from all Frankish nobles and freemen after overthrowing the Thuringian magnate Hardrad’s conspiracy (786) only a few years after Pippin the hunchback’s revolt, can be seen as indicators that Charlemagne’s lordship might not have been uncontested.
  15. The question was and is disputed. Whereas Hägermann (Hägermann 1975, 287 ff.) claims that the Emperor’s title was left out because the understanding between the three brothers was not to be endangered, Boshof gives the interpretation that Charlemagne left this problem open to be capable of changing the divisio, if necessary (Boshof 1996, 86).
  16. For instance, apart from the Danish raids, the last victory over the Saxons was in 840,
    meaning that – given the length of the conflict – there would have been oppositional groups waiting for a chance to strike back.
  17. According to Eginhard, there were signs in the sky, a building at Aachen collapsed, a bridge over the Rhine Charlemagne had built burned down. The narrative deals explicitly with the loss of symbols of the kingship: before his last expedition against the Danes, Charlemagne was thrown off his horse and lost the fibulae of his cloak and his lance; his swordbelt was torn. The old man had to be helped up by his servants. Furthermore, his palace at Aachen was subject to several earthquakes, the beams of the buildings he was staying in creaked incessantly, etc. Several months before he died, the inscription KAROLUS PRINCEPS (The Prince Karl) in his church at Aachen began to fade, the word
    PRINCEPS became completely illegible (Eginhard, Vita, c. 33).
  18. For a summary of the problems, see Ganshof 1971, 240 ff.
  19. In spite of the suspicions which easily come to mind in such a situation, there is no hint at
    foul play.
  20. For instance, whereas Charlemagne enjoyed the Frankish oral tradition and ordered it to be written down (Eginhard, Vita, c. 29), according to Thegan, Louis despised them (Thegan, Vita, c. 19). The latter also was not overfond of the festivities he was obliged to hold and did not even laugh at the jokes of the artists and jesters (ibid.), whereas his father enjoyed these diversions but – a-typically for the time – frowned on inebriety going along with such occasions. (Eginhard, Vita, c. 24).
  21. The Astronomus’s narrative tells of wide-spread fears of Wala, who had been very close to Charlemagne, acting against Louis. Wala journeyed to Louis and commended himself before Louis even reached Aachen His example was followed by many Frankish nobles (Astronomus Vita, c. 21). Given the usual style of contemporary written sources, this episode might hint at a potential crisis of state. Wala would have been aware of the suspicion nursed against him. The only chance he stood was to react as he did. The behaviour of the nobles leaves open several options: Did they follow Wala’s example only after having waited if Louis’ss succession was to be uncontested?, or: “What if .. Wala had stayed put?”
  22. Adalhard was banned to Noirmoutiers, Wala left politics and entered the monastery of
    Corbie, Drogo, Hugo and Theoderic were kept under close surveillance at their brother’s court (Boshof 1996, 94).
  23. The Ordinatio imperii specified the inheritances of Louis’s sons. Although the realm was
    divided up according to Frankish legal tradition, the nomen imperii was to go to the eldest
    son Lothar.
  24. The wicked image of Judith that some sources draw has recently been corrected, cf. Koch 2005.
  25. Cf. the works of Peter Godman, Janet Nelson, Michael Richter and Matthias Becher.
  26. The first phase of succession could be dangerous because the new king had to summon all his people to confirm existing privileges and renew the oaths of fidelity and the commendations which had been given – not to the abstract office of king – but to the person of his predecessor.
  27. Although two capitula had already been issued in the name of Pippin III., the use of this
    type of legal writ started to be systematically employed during the rule of Charlemagne and his successors.
  28. For a complete evaluation of these novelties as well as the different research opinions on the “Carolingian renaissance” cf. McKitterick, 1994, Pohl, 2002.
  29. Cf. the previous note.


Althoff, Gerd. 1997. Spielregeln Althoff, Gerd: Spielregeln der Politik im Mittelalter. Kommunikation in Friede und Fehde, Darmstadt: Primus.

Becher, Matthias. 2000. Karl der Große, Müchnen: Beck.

Becher, Matthias and Jörg Jarnut (eds.). 2004. Der Dynastiewechsel von 751. Vorgeschichte, Legitimierungsst rategien und Erinnerungen, Münster: Scriptorium.

Becher, Matthias. 1993. Eid und Herrschaft: Untersuchungen zum Herrschaftsethos Karls des Großen, Sigmaringen: Thorbecke.

Bloch, Marc. 1974. Apologie der Geschichte oder der Beruf des Historikers, Stuttgart: Klett-Cotta.

Boshof, Egon. 1996. Ludwig der Fromme, Darmstadt: Wiss. Buchgesellschaft.

Brunner, Karl. 1979. Oppositionelle Gruppen im Karolingerreich, Wien, Köln, Graz: Böhlau.

Classen, Peter. 1964. Karl der Große, das Papsttum und Byzanz, Düsseldorf: Schwann.

Erkens, Franz Reiner (ed.). 1999. Karl der Große in Renaissance und Moderne. Zur
Rezeptionsgeschichte und Instrumentalisierung eines Herrscherbildes. Berlin: Akademie-Verlag.

Godman, Peter, Jörg Jarnut, Peter Johanek. (eds.). 2002. Am Vorabend der Kaiser-
krönung. Das Epos Karolus Magnus et Leo Papa und der Papstbesuch in Paderborn 799, Berlin: Akademie-Verlag.

Haude, Rüdiger. 2000. „Kaiseridee“ oder„Schicksalsgemeinsch aft“. Geschichtspolitik beim Projekt „Aachener Königsausstellung 1915“ und bei der „Jahrtausend-ausstellung Aachen 1925“, Aachen: Verlag des Aachener Geschichtsvereins.

Fichtenau, Heinrich. 1949. Das karolingische Imperium. Soziale und geistige Problematik eines Großreichs, Zürich: Fretz & Wasmuth.

Ganshof, Francois L. 1971. The Last Period of Charlemagne’s Reign. A Study in Decomposition in The Carolingians and the Frankish Monarchy. Studies in Carolingian History, London: Longman.

Geuenich, Dieter and Ingo Runde (eds.). 2006. Name und Gesellschaft im Frühmittelalter. Personennamen als Indikatoren für sprachliche, ethnische, soziale und kulturelle Gruppenzugehörigkeiten ihrer Träger, Hildesheim, Zürich, New York:Olms.

Hägermann, Dieter. 1975. Reichseinheit und Reichsteilung. Bemerkungen zur Divisio regnorum von 806 und zur Ordinatio imperii von 817, HJb 95: 278-307.

Hagn, Hans. 2006. Illegitimität und Thronfolge. Zur Thronfolgeproblematik illegitimer Merowinger, Karolinger und Ottonen. Neuried, ars una.

Hampe, Karl (ed.). 1935. Karl der Große oder Charlemagne? Acht Antworten deutscher Geschichtsforscher, Berlin: Mittler.

Kasten, Brigitte. 1997. Königssöhne und Königsherrschaft. Untersuchungen zur Teilhabe am Reich in der Merowinger-und Karolingerzeit, Hannover: Hahn Koch, Armin. 2005. Kaiserin Judith. Eine politische Biographie. Husum: Matthiesen.

McKitterick, Rosamund, 1994. Carolingian Culture. Emulation and Innovation, Cambridge, Cambridge University Press.

Nelson, Janet. 2004. Betrada, in Becher, Matthias and Jörg Jarnut eds. Der Dynastiewechsel von 751. Vorgeschichte. Legiti mierungsstrategien und Erinnerung, Münster: Scriptorium.

Pohl, Walter. 2002. Vom Nutzen des Schreibens. Soziales Gedächtnis, Herrschaft und Besitz im Mittelalter, Wien, Verlag der österr. Akademie der Wissenschaften.

Raphael, Lutz. 1994. Die Erben von Bloch und Febvre. Annales-Geschichtsschreibung und nouvelle histoire in Frankreich. Stuttgart: Klett-Cotta.

Scheibelreiter, Georg. 1999. Die barbarische Gesellschaft. Mentalitätsgeschichte der europäischen Achsenzeit, Darmstadt: Wiss. Buchgesellschaft.

Schneider, Reinhard. 1972. Königswahl und Königserhebung im Frühmittelalter. Untersuchungen zur Herrschaftsnachfolge bei den Langobarden und Merowingern, Stuttgart: Monographien zur Geschichte des Mittelalters, 3.

Schubert, Ernst. 2006. Essen und Trinken im Mittelalter, Darmstadt: Wiss. Buchgesellschaft.

Schramm, Percy Ernst. 1954-1956. Herrsch aftszeichen und Staat ssymbolik. Beiträge zu ihrer Geschichte vom dritten biszum sechzehnten Jahrhundert, Stuttgart:MGH Schriften XIII vol. 1-3.

Weinrich, Lorenz (ed.). 1963. Wala, Graf, Mönch und Rebell, die Biographie eines Karolingers, Lübeck, Hamburg: Matthiesen.

Originally published by Historical Social Research 34:2 (184-202), 2009, republished by SSOAR Open Access Repository under the terms of a Creative Commons Attribution-NonCommercial 4.0 International license.