The illegible Carolingian army

The title of this post is inspired by the work of James C. Scott, whose book Against the Grain I’ve recently finished. One of Scott’s repeated themes is that states, including ancient ones, are always trying to make their subjects and their lands “legible”. They want complete and standardized descriptions of all the areas and individuals they control for planning purposes. Early states, in particular, wanted to be able to assess resources and people so they could raise tax and levy armies.

It’s at this point that the Carolingianist starts to point out anomalies in which seems overall like a sensible argument. Carolingian rulers didn’t have centralised land-based taxes as part of their revenue. But they did raise considerable armies. So how did they do that when they didn’t have ways in which to make their large territory legible?

Read More »


 The price of good history

I want to discuss a couple of examples I’ve seen recently of the problems of making good-quality history available to the general public. The first comes from the announcement of a special issue (44-1, 2018) of the journal Historical Reflections. The senior editor, Linda E. Mitchell, explains that the special issue:

is devoted to a wide-ranging and erudite critique of the claims of evolutionary psychologist Steven Pinker that (a) the modern world is less violent than the premodern world, and (b) historians are not treating the period of the seventeenth and eighteenth century known as the “Enlightenment” with the respect it deserves. As the pages of this journal have demonstrated over and over, these statements are not just problematic and controversial, they are patently tendentious. Pinker—who is not a historian—belongs to a group of people, often our colleagues, who believe that writing history is easy and that anyone can do it. We know that what makes a good historian is training in the theories, methodologies, and materials of historical study. We know that history is not merely a narrative. Historical writing is labor intensive, often requiring hundreds of hours of sifting through archival materials in many languages and many forms. It also requires a clear understanding of and sensitivity to the contexts of historical inquiry. The 12 articles in this issue demonstrate the best kinds of historical writing in critiquing the methodology and conclusions of Pinker’s best-selling pseudohistorical study.

Read More »

The problems of English education 3: the irrationality of white students

My title, is, inevitably an exaggeration: I’m talking here about undergraduate students at English universities and referring to only one aspect of irrationality: economic irrationality. Nor am I saying that all white students are economically irrationally. But I think there is a good case to be made that white students are less likely to be economically rational when choosing their degree than BME (black and minority ethnic) students and that potentially poses a problem for academics in some subject areas.

To see this irrationality, here is data analysing the earnings of students after graduation graphed against the “whiteness” of the main subject groupings (using the JACS code). The earnings data is from the LEO project (Longitudinal Education Outcomes), specifically the Subject tables from  SFR60/2016. This looks at the earnings of UK domiciled first degree graduates from English HEIs.

In the first graph I show the median annual earnings for the cohort (male and female) who graduated in 2008/09 five years after their graduation, i.e. tax year 2014/15.

In the second graph I show the median annual earnings for male students who graduated in 2003/04 ten years after their graduation, i.e. tax year 2014/15. The data on ethnicity is from HESA Table 8 and is for full-time UK undergraduate enrolments for 2016/17.

Here are the figures:

Problems 3 figure 1
5 year subject earnings v “whiteness”
problems 3 figure 2
10 year male subject earnings v “whiteness”

Read More »

The problems of English education 2: diverse forms of diversity

The recent IPR report on diversity within universities has the great merit of providing some hard data on what is obviously empirically: universities vary greatly in their ethnic diversity. It also looks at this in terms of broad subject categories, which is particularly useful (and challenging) for historians, since history is one of the “whiter” subjects. The disadvantages of the report are firstly, that the main tables are only in pdf format and secondly, that the main measure used to analyse diversity is an entropy index, which isn’t intuitive to non-specialists. But I’ve got some of the data out of the tables and have started playing with it, so here are some of my initial thoughts on overall university diversity. (I’ll say more about subjects in a later post).

Read More »

How archaic were the Ottonians?

I have recently read Gerd Althoff’s book on Otto III and I am still trying to unravel one key statement from it. On page 25, he states: “We speak of an archaic society.” (This isn’t a translation issue: in the original German, the adjective is “archaisch”). So what does it mean for Althoff to call late-tenth century Ottonian society “archaic” and he is right to do so?

Otto III enthroned (image from Gospels of Otto III)


In one sense, Ottonian society can’t be “archaic”, in the sense of “age-old”: a key part of its culture was Christian, and the Saxons had only been Christian for around two centuries at the most by that point. What about other aspects? Althoff rightly states (p. 16) that the Ottonians had fewer state structures than the Carolingians, although, as Tim Reuter pointed out, that comparison is less striking if you compare Carolingian East Francia with the Ottonian kingdom. But it’s not obvious that in the tenth century there is a whole lot more “stateness” in West Francia than the Ottonians have, so if you count the Ottonian kingdom as “archaic” in the sense of “old-fashioned”, then you’d have to rate large chunks of the rest of Western Europe as “archaic” (apart from the Anglo-Saxons, who were forming an up-to-date kingdom like the end of the world was coming).

The Anglo-Saxon kingdom also demonstrates that a culture that has “a wealth of ceremonial and ritual acts and activities, which served to display rule” (as Althoff puts it, p. 16) isn’t necessarily “archaic”. As Charlie Insley’s recent wonderfully-titled article, ‘Ottonians with Pipe Rolls’? Political Culture and Performance in the Kingdom of the English, c.900–c.1050 and other work by Julia Barrow and Levi Roach has shown, we can see similar ritual and ideas of performance of rule visible in Anglo-Saxon sources, especially charters as in Ottonian ones (and in the West Frankish kingdom as well).

But what seems to me key to Althoff’s idea of the Ottonian world as “archaic” is a different aspect: rulers’ supposed inability to plan rationally. He complains (p. 18-19):

modern scholars…postulate a wealth of other areas that supposedly preoccupied medieval “policy makers.” So for the tenth century it is traditional for historians to speak of Italian policies, eastern policies, western policies, and monastic policies. It is also customary to portray rulers and magnates as acting as if political plans had first been developed and then been put into effect in these areas and others. Such portrayals carry the implicit assumption that the rulers took counsel and agreed with their magnates on such plans, which then became guiding principles of a policy applied for a certain period of time or even long-term.

By contrast, the sources of the tenth and other centuries refer to such reasoned policies only very seldom, if at all. This rather forcefully raises the question: do we not deal in anachronisms if we categorize medieval royal rule according to the model of modern government with its plans and policies, especially since we are unable to trace where and how such alleged policies arose? Indeed, one would be right to doubt whether this supposed intense planning is in any way at all in harmony with the conceptual framework and mentality of the central Middle Ages. Plans and policies may have been quite alien in a society whose understanding of politics centered with such certainty on the idea that a God-ordained order must either be guarded or reestablished. Such an understanding also set its stamp on the duties and powers of medieval kings, as the significance of peace and justice among the ruler’s duties attests. But performing such duties did not so much require some sort of future-oriented planning as the employment of the customary usages by which people had always performed these duties.

And towards the end of the book (p. 108), Althoff states that one of the main themes is to consider:

how effective rulership actually was in the tenth century, how wide the sphere of political activity, and whether the king could shape the events of this era. Did he spend his time outlining political strategies in the widely diverse realms of church and monastic policy, the policies of the west, east, and Italy? Did he try to reach accord among the powers concerned? Did he then formulate final plans, and oversee their implementation? Or are these activities anachronisms? Did a tenth-century king understand his duties differently?

So is the concept of planning something alien to Otto III, and early/central medieval kings more generally? Is it anachronistic to think that kings might have policies? There are two fairly obvious counters to this: from the Gospels and one from the Carolingians.

First of all, let’s hear Jesus’ words in Luke 14, 28-32:

Suppose one of you wants to build a tower. Won’t you first sit down and estimate the cost to see if you have enough money to complete it? For if you lay the foundation and are not able to finish it, everyone who sees it will ridicule you, saying, ‘This person began to build and wasn’t able to finish.’

Or suppose a king is about to go to war against another king. Won’t he first sit down and consider whether he is able with ten thousand men to oppose the one coming against him with twenty thousand? If he is not able, he will send a delegation while the other is still a long way off and will ask for terms of peace.

Planning, to Jesus, is a fact of life so obvious that you can use it as an analogy to the costs of discipleship. And the examples he gives are precisely the kind of decision that all medieval kings had to make. Amid all the “customary usages” of Ottonian life, the ruler still had to make such binary choices: do I carry out this particular action now or not? The emphasis on the king “taking council” is precisely because such decisions needed to be made on a regular basis. At least for the Carolingian period we’ve even got a specific image of a council of war; this comes in Ermoldus Nigellus’  In honorem Hludowici, where Louis the Pious’ magnates discuss attacking Barcelona. Given all this, it seems extremely implausible to me that the Ottonians (and Otto III specifically), didn’t carry out at least some “future-oriented planning”.

What about policies? This depends rather more on definitions. If you think of policies as formal statements (often written) which lay out overall principles within which more day-to-day planning takes place, then no, medieval governments didn’t have policies. But if you think more generally of policies being to planning as strategy is to tactics, then I think at least some early medieval rulers had policies. You can sometimes see consistent courses of action, rather than purely short-term reactions to events. You can debate whether or not Charlemagne had a Saxon “policy”, but it’s hard to dispute that he had the long-term aim of conquering the Saxons.

So why aren’t these policies recorded in our sources more? One obvious answer is that a lot of the policies had to do with how to deal with external powers or internal enemies and so it was sensible to keep them secret. It’s not surprising that no Carolingian or Ottonian ruler produces a memo, statement or the like saying “we intend to make sure that no pope hostile to our interests is elected”. That doesn’t stop at least some of them repeatedly interfering in papal elections.

There’s also the point that what medieval authors tend to be writing isn’t histories of the “Policies of the Saxons”, but the “Deeds of the Saxons”. Early medieval traditions of narrative history mean that they don’t tend to focus on long-term policies, but instead short-term sequences of events. So, for example, as Eric Goldberg points out in Struggle for Empire, there’s a good case for seeing Louis the German as having a policy of Drang nach Westen, repeatedly trying to take over parts of the Western Frankish empire, but that’s not directly stated anywhere in his documents or narrative sources of the period.

But Louis’ Drang nach Westen does suggest something distinctive about the quality of early medieval plans: their implementation tended to be more episodic than in later periods. Louis the German may well have spent years wanting to expand into Charles the Bald’s kingdom, but his main moves were in 853/854 and 858, and in between he was focused on his eastern frontier. In a similar way, most discussions of Charlemagne see at least some of his campaigns as “opportunistic”; so does his intervention in a dispute in Lorsch which ended with it becoming a royal monastery.

I think this reflects the fact that early medieval kings (and arguably medieval kings in general) probably were less able to influence events than early modern/modern rulers. In particular, natural disasters (famine, disease, death) could have a major impact in creating opportunities or hindrances to military campaigns and diplomacy (two of the key policy areas). When you add to that the fact that rulers had much less information available to use when assessing foreign policy decisions, in an era of poor communications and without permanent diplomats, it’s not surprising that systematic long-term planning wasn’t as important a priority then compared to say, the early modern period.

Going back to Althoff, I also worry about his linking of planning with the question of the “effectiveness” of tenth-century rulers. Despite the modern 7Ps formula (“Proper Planning and Preparation Prevents Piss Poor Performance”), I’m not sure that you can assume that things going wrong for rulers necessarily resulted from their lack of planning. Carolingian kings and emperors went through a lot of crises, and these often arose when things went wrong in several widely-spaced locations simultaneously (such as happened to Charlemagne in 778 or Charles the Bald in 858). It’s not clear to me how easily any medieval ruler could have planned to prevent such difficulties, given the difficulty of rapid travel.

In the end, the question of whether or not early medieval kings were capable of plans and policies depends on how tightly you define the concept. There was nothing like the impartial and theoretically competent bureaucracy of the modern state to carry out such planning, but given that only really developed from the mid-nineteenth century, that’s not surprising. And it’s interesting how often the word “policy” does creep back into Althoff’s description of Otto III’s activities. In practice, it’s quite hard to deny some policy/planning capacity not just to the Carolingians, but to East Frankish rulers generally.

But it’s now that we get to one final twist, inspired by my favourite description of Otto III. This was uttered many years ago by Susan Reynolds, who said he was “a young man with a lot of fancy ideas, but he didn’t last long”. It’s perfectly possible for the Ottonians as a whole to have been capable of making and implementing plans and policies, but for Otto III not to have used that capacity. Because as we’ve seen in 2016 and 2017, just because you can plan carefully, doesn’t mean that every politician wants to. It may be too tedious for Trump to sit down and estimate if he can afford to build a tower (or a wall) before proclaiming that he will, or for the UK cabinet to consider if David Davis and a couple of civil servants really can emerge victorious from an encounter with an army of Brussels bureaucrats. In other words, if you’re going to label states as “archaic” because of their inability to plan or their lack of effectiveness in achieving their aims, we may now have a few modern states to include on the “archaic” list.

Carolingian honour and its lack

Statue d'Artagnan Paris

(Statue of d’Artagnan by Gustave Doré)

I’ve been worrying away at concepts of Carolingian honour, on and off, for at least fifteen years. It was the theme of a discussion I had the first time I ever met Paul Hyams, for example, many years ago at the International Medieval Congress. He was pointing out that every society has concepts of honour and insults to an individual’s honour. I was trying to put the case, not very coherently, that there was something different about some societies/cultures and their responses to such affronts. The nearest I could get at that point was talking about d’Artagnan at the start of The Three Musketeers, who ends up having to fight duels with Aramis, Athos and Porthos because of him offending each of them in relatively minor ways.

Anyone might be aggrieved by someone knocking into them because they’re in a hurry, or making disrespectful jokes about them, but a fight to the death as a result would seem to most of us an overreaction. More generally, I’d now characterise an honour culture as one in which it is expected by a social group that acts which do not cause physical harm to oneself nevertheless should be responded to with physical violence, sometimes lethal. (This may include responses to physical or verbal attacks on a third party, so that this definition includes codes of chivalry in which a man who fails to protect/revenge physical or verbal “insults” to certain types of women is himself disgraced, as well as cultures which accept or encourage violent reactions to perceived blasphemy).

This isn’t quite the same as a feuding culture, although there is a considerable overlap. I’m taking feuding to mean a situation where an attack on one person is likely to lead to retaliation by the victim’s social group (whether it’s relations, friends, or some other kind of gang/clan etc) against someone from the attacker’s social group and in which this tit-for-tat violence may continue for some time. It’s possible to have an honour culture without feuding: for example, one in which a duel is seen as settling matters once and for all, and there’s no retaliation against a victor, even if he’s killed his opponent. It’s also possible (though I think not common) to have a feuding culture where honour isn’t key, but where instead violence is used largely defensively and not as a first resort.

With these somewhat vague definitions in mind, why am I back to thinking about Carolingian honour? Because I’m reading Gerd Althoff’s biography of Otto III, which takes “honour” for granted as a key to how the king dealt with his circle of confidants. As Althoff puts it (p. 17):

Each ruler had to take the honor of each person in this circle [of confidants] into account. By this is meant the sum of all earned and acquired possessions, offices, abilities and the rank that they conferred. Each ruler had to value each person proportionate to this honor, in other words give preferential treatment, listen, give gifts.

For Althoff, honour is a vital part of the rules of the Ottonian game, reflecting what tenth and early eleventh-century sources say about such matters of etiquette. It’s these sources that mean that Heinrich Fichtenau, in Living in the Tenth Century: Mentalities and Social Orders starts with a long section on ordo and disputes (sometimes violent ones) over status.

The problem for Carolingianists has always been whether the same patterns existed in Carolingian court culture in the period and it’s just that our sources don’t reveal it or whether something different is going on. Coming back to the problem after recently researching political culture, I feel more confident that there is a real difference. Even if eighth and ninth-century narrative sources don’t discuss such moments of offended honour, I’d expect all the moral tracts from the period to say something about the morality of such violent reactions to offence if it was a significant part of Carolingian court life. Why isn’t Dhuoda, for example, telling William about exactly how offended he needs to be if any insults are thrown at him (or his father Bernard of Septimania) at Charles the Bald’s court? Even if she doesn’t want him to get into fights in that way, she should be telling him to “turn the other cheek” if violent reactions to insults are a normal part of court culture.

Yet the tenth-century emphasis on honour obviously wasn’t new to Frankish societies, as you can see by a section on insults lurking in Lex Salica (Pactus Legis Salicae, c. 30). Compensation for specific insults makes it clear that they are “injuries” just as much as knocking out someone’s tooth. So what changed and why?

One possibility is that the refined Merovingian court culture associated in particular with Dagobert I in the early seventh century persisted into later centuries, but for reasons I’ll come to, I don’t think that a culture of restraint would endure that long without continued reinforcement. My suggestion is that it was Charlemagne who reduced the expression of a traditional Frankish honour culture at court.

As support for my hypothesis there’s firstly the fact that Charlemagne had the power if necessary to humble magnates. The most striking example of this is making Tassilo submit to him in 787, but as Stuart Airlie has pointed out, there were a lot of other submissions to Charlemagne in the 780s. In some of these cases, Charlemagne allowed his victims to save face, at least partially, but any magnate must have been aware that it was risky to push his own claim to a “right to respect” (another way of conceptualising honour) too far at Charlemagne’s court.

So the sources suggest that Charlemagne had the power and willingness to affect the culture of honour, even if he didn’t reject it entirely. Secondly, we know that Charlemagne was opposed to feuds and legislated against them. There are obvious questions about how effective his actions were, but it certainly implies an opposition to the kind of “self-help” violence characteristic not only of feuding, but also honour cultures.

Thirdly, Charlemagne clearly didn’t want to be too tightly bound by the existing hierarchy of honour: he wanted to able to bring outsiders into his court. Predominantly these were intellectuals, but part of the integration of newly conquered territories involved rewarding non-Franks who were loyal to Charlemagne (like the Saxons Hitti and Amalung). Honouring everyone “properly” relies on a clearly-established hierarchy, but Charlemagne’s successes were altering this. How do you decide the “proportionate value” given to Alcuin above a Frankish deacon or a newly-sworn Saxon fidelis as opposed to one in the Carolingian heartlands of Neustria? A de-emphasizing of pre-existing hierarchies and honour codes was probably necessary to accommodate such outsiders.

And finally, if you want to see a possible example of the effects of this diminished honour code, look at what Einhard says about Charlemagne in the baths at Aachen (Vita Karoli c. 22). Einhard claims that that there could be over a hundred men in the baths at Aachen at one time: Charlemagne and his sons, his magnates, friends and sometimes even his attendants and bodyguards (satellites, custodes corporis). It’s obviously a status marker to be one of these in the pool, but bathing’s about the least suitable occupation possible for maintaining a status hierarchy within a group. The bathers were probably all naked, so markers of clothing and equipment vanish (a big contrast to hunting, one of the other main Frankish pastimes). And I’m also pretty sure that in a bath with over one hundred men in it, there is going to be horseplay. At some point someone senior is going to get splashed or bumped into and the court culture has to be such that such informality is accepted and doesn’t lead to long-lasting enmity or violence.

Of course, I’m not denying the existence of hierarchy at the Carolingian court. Even if old Frankish hierarchies were affected by Charlemagne’s actions, a new hierarchy promptly developed, but because there was a particularly strong focus on Christian behaviour at court, the disproportionate violence of a normal honour code was inappropriate. As Chris Wickham, The Inheritance of Rome, puts it in the ninth century (p. 413): “Whether or not magnates were governed above all by realpolitik, they felt a strong need to express their political choices in moralized terms”. In the same way, I think that whatever anger secular and ecclesiastical magnates may still have felt about insults to their honour, it had to be expressed within a Christian framework, which limited the options for retaliatory violence. I’ve talked before about how the Vita Gangulfi reveals a conscious attempt by a Carolingian author to oppose one common form of honour-related violence: the killing of adulterous wives. (The one big exception to this limitation on revenging affronted honour was saints: offences to the honour of their church could still be countered with lethal force by dead saints, since such violence could be justified by its pure motives).

I think there are a few other clues from the late eighth century that confirm that it was at Charlemagne’s court that Frankish codes of honour were first muted and rechannelled in this way. One is the emergence of mocking court poetry (from the time of Peter of Pisa and Paul the Deacon onwards). You can have a culture of public mocking along with a strong focus on honour (you can see that in Viking Scandinavia, for example), but it tends to get violent very soon, whereas Carolingian courts mostly didn’t. The argument that this was because it was mostly clerics involved doesn’t bear much weight – Ottonian clerics were extremely touchy about their honour and Carolingian clerics were perfectly capable of using their proxies to carry out violence for them, as seen in the conflict between Theodulf of Orléans and Alcuin over sanctuary. And Carolingian court poetry also occasionally mocked lay magnates, such as Wibod. However much Wibod may have been in on Theodulf’s joke against him, it presupposes a culture in which it is socially acceptable not to fight a man who insults you.

The second point comes from Alcuin’s lay mirror, De virtutibus et vitiis, written around 800 and one of the most successful texts (in terms of surviving manuscripts and translations) from the Carolingian period. That may have been partly to do with the sheer banality of its content, which purveys a Christian morality that I’ve previously described as intended for spiritual couch-potatoes rather than spiritual athletes. One of the passages that made me think that is from Alcuin’s chapter on patience (c. 9), where he says:

We can be martyrs without sword or flames if we truly preserve patience in mind with our neighbours. It is more praiseworthy to avoid injury by being silent than to overcome by responding. He who tolerates evils patiently will merit an eternal crown in the future. (Sine ferro vel flammis martyres esse possumus, si patientiam veraciter in animo servamus cum proximis nostris. Laudabilius est injuriam tacendo declinare, quam respondendo superare. Qui patienter tolerat mala, in futuro coronam merebitur sempiternam).

Looking at this passage again, I now wonder if the seeming hyperbole about the “martyrdom” of true patience is in fact reflecting the difficulties involved in moderating an honour culture.

If you put these pieces of evidence together, it does suggest that Charlemagne was trying to change attitudes towards honour. What’s more, that is how he was remembered at the end of the ninth century. The anecdotes of Notker the Stammerer repeatedly show Charlemagne exalting the humble and humbling the proud. In contrast, Notker’s main anecdote about Pepin III (Notker 2-15) is how he demonstrates to his magnates his superior courage to them by defeating a lion, after they’ve been secretly despising him. Pepin’s violence here is against an animal, rather than his detractors themselves, but otherwise it’s a classic story of challenged honour being regained. Notker, therefore, associates the memory of Pippin with a traditional honour culture, but Charlemagne with a more complex hierarchy, in which virtue counts for more than social status.

Notker shows a memory living on of Charlemagne as rejecting many aspects of honour culture, suggesting that it may well have been part of court culture for several generations of Carolingians. Indeed, it’s only at the end of the ninth century that annals start to include incidents of offended honour, such as the Annales Vedastini from 900 reporting on Robert I leaving Charles the Simple’s court because he’d been insulted by something somebody said about him. There also the intriguing statement of the Fulda Annals from 900 (Tim Reuter’s translation):

Zwentibald….continued to hold onto the Gallic kingdom [Lotharingia] and to attack the lands of the church with immoderate cruelty. His worst crime was to strike Ratbod, archbishop of Trier on the head with his own pastoral staff, contrary to the honour due a bishop. (Zuentipoldus…Gallicanum  regnum secum retinens, res ecclesiarum crudelitate sua inmoderate affectans, maxime crimen eo, quod Ratpodo Treverensi archiepiscopo contra sacerdotalem honorem baculo suo in capite percutiens intulit).

The first sentence is a standard Carolingian denunciation of a bad magnate or king; I can’t think of an eighth or ninth century parallel to the second statement. This isn’t about a brutal physical attack; there’s no suggestion of permanent injury to Ratbod. Zwentibold’s “worst crime” is mainly about disrespect.

As well as this direct evidence of heightened concerns about honour, from the 870s onwards, there are also signs in East Francia of it being harder for kings to control individual aristocrats and feuds becoming more prominent (such as between the Conradines and Babenbergers). What this suggests to me is that in the absence of rulers who are attempting to moderate or oppose cultures of honour and feuding among nobles, these cultures are liable to return. They are, after all, two of the main ways of ensuring protection for oneself and one’s family in non-state societies, by showing a willingness to retaliate, even disproportionately. So I’d now be inclined to say that Carolingian honour was materially different from Ottonian ideas of it and that this was one less-visible aspect of Carolingian reform.

The strangeness of the Carolingian rural state

There’s nothing like reading research far outside your field to give you a new perspective on your own work. In my case it’s making my way through Kent Flannery and Joyce Marcus, The Creation of Inequality that’s brought home to me how strange the Carolingian empire was in one respect.

Flannery and Marcus’ book is an anthropological/archaeological look at a wide range of prehistoric, historical and contemporary societies at varying levels of social complexity. There’s no mention of the Carolingians (or indeed any European societies), but there is a lot on cultures in the Americas, including the Moche, Mayan and Zapotec civilisations.

Panorama of Monte Alban from the South Platform

Panorama of Monte Albán site

Read More »