I've wanted for a long time to do a conference paper on this topic, but it would probably not be taken in the right sense. And I don't think there's quite enough material to make a best-selling book out of it. So here, in blog form, are

The management secrets of Charlemagne

Discussions of how Charlemagne’s realm was organized and governed have often started, at least implicitly, from a modern model: present day or perhaps more accurately nineteenth century bureacratic government. Some authors have accepted this model and then argued whether or not Charlemagne’s government succeeds in these terms. Those who reject this model, stress, instead, how unlike any modern concepts of administration Carolingian government is. I want to argue, however, the possibility of a different modern analogy for the Carolingian state, which can provide another view of it and different criteria by which to assess its effectiveness. That model is the company, and its organization and effectiveness as seen in current management thinking.

I do not, of course, want to claim that the idea of a company was one that would have made any sense to the Carolingians. But what if we do look at Carolingian government ‘as if’ it is Francia plc? How effective does it look then?

The role of paperwork

Political historians (certainly of the medieval period) remain one of the outposts of ‘big government’: the size and complexity of the governmental bureaucracy is the key to its success, The serious government can be recognized by the size of its archive. Current management theory, however, has turned against the bureaucracy, an attitude that is now affecting government departments as well as businesses. This view is summed up in the acronym KISS: Keep It Simple, Stupid. In particular, records and information should not be collected unless they are useful and should be retained only as long as they are still needed. Collecting information for no good reason is a waste of resources.

This is a frequent problem in the modern civil service. A common response to parliamentary questions is the comment that ‘this information is not available except at excessive expense’. The urge to collect such information has to be resisted, for the sake of efficiency. The Department of Health has to impose restrictions on the questionnaires that civil servants may send to GPs, since otherwise they’d never have time for their actual work. Similarly, there are a few Carolingian ‘questionnaires’ that look excessively complicated, where simple questions must have led to hours of work for the respondents. We do not have the replies to many of these demands for reports. Does this mean, as is sometimes suggested, that these were not carried out? Or is it a reflection that such information was not seen as needed long-term and hence not preserved? Did a Carolingian ruler need to keep a record of which monasteries had been found not following the Benedictine rule sufficiently closely at the last inspection? Or was the key thing what was happening now, not what had been the case several years ago?

Similarly, there is a tendency to take too bureaucratic a view of the efficiency of medieval government communication. On this view, Carolingian government is ineffective because it relied on oral communication and its lack of standardised channels for circulation of capitularies and other materials was unable to keep the regions properly informed of its wishes. This presupposes a view of the effectiveness of written communication that modern bureaucracies hardly support. Even if you send out a vast stream of paperwork, people either don’t get it, don’t read it or lose it. The most heavily used section of the Department of Health website by the public when I was there, was the texts of DoH circulars - i.e. the material that had already in theory been circulated to the NHS and other organisations.

The use of assemblies and other forms of oral government also tends to be seen as a sign of a more primitive state. But in one big company I was in, whenever they had something particularly important to announce it was done via senior management standing up in general meetings and talking. They were even prepared to travel round the country or to video this and replay it elsewhere. Modern organisations have writing, e-mail etc, but it’s still felt that for some things face-to-face communication is necessary.

Delegation

Linked with the modern managerial urge for simple systems is the principle of delegation. The organization should be ‘flat’, with the minimal numbers of layers of management and the maximum amount of authority delegated to those ‘on the ground’, dealing directly with those outside the organization. They, then, can react to local conditions and problems in a speedy and flexible manner, and this is seen as more effective as the standardised response of a centralised, top-down organisation.

On this view, the Carolingian method of control via counts has to score relatively highly. Indeed, even in today’s bureaucratic government, there are times when such an approach is preferred. The modern civil service sends out routine information from the centre. But they also have a ‘cascade’ system for rapid dissemination of information and instructions in emergencies. In this, key figures are contacted, who must then pass on the information to others in their region. The mobilization system that Charlemagne developed for the army in later years, might almost serve as a model for this.

This delegation of responsibility also contributed to keeping systems simple. Ganshof argued that the local raising of armies meant that the king could not know in advance how many troops would appear for the annual campaign. I want to ask a basic question: did he need to know exactly how many troops he would get? If none or very low numbers appeared, of course, that was a problem, but provided a reasonable number were present, exact details of how many didn’t matter. The idea lurking in Ganshof’s mind is a modern army, where a complex deployment of regiments in a pre-set plan is supported by equally complex and centrally-organized logistics. Carolingian warfare, by delegation, eliminated most of these problems. Men brought their own rations: if they were inadequate, that was their look out.

Audit and corporate culture

One obvious problem with delegation is that the centre cannot easily keep control, and in particular that those to whom power is delegated may misuse it. Most organisations have some form of audit/control mechanisms. I would argue that the royal missi look in some way like an OFCOUNT (an office for ensuring the standards of counts), on the lines of OFSTED. Modern scholars have seen a fatal flaw in the use of the missi: the missi were themselves part of the ruling class and so complicit in the abuses. This sounds like a rerun of the debate about OFSTED as opposed to the previous system of school inspectors. Do you have outsiders as the auditors, who are able to be more objective, or does that result in ignorant comments, when a better analysis can be made by those insiders who know what it’s really like at the sharp end?

More generally, it’s interesting to consider modern views of Carolingian officials in the light of Douglas McGregor’s Theory X and Theory Y ideas of employees. Historians tend (not surprisingly given the extant evidence) to have a Theory X view of counts: they are going to get away with whatever they can and they cannot be trusted an inch. I’d argue that the Carolingian system, however (like many successful companies), was built on Theory Y ideas: staff can perform well without micromanagement. It is reasonable to trust your staff to do their job. Why? Because you’ve got good staff. This is why Carolingian mirrors are obsessed with the need for king to choose good men as advisors and officials, because it’s a method of organisation that depends crucially on the quality of personnel within the system, not the formal perfection of the system. This is the difference between a company that believes that ‘people are our most important resource’ and a civil service model which is supposed to carry on functioning effectively even if there’s no government to direct it.

One of the things which a company needs to work in this way, of course, is a strong corporate culture, so employees instinctively know how to do things ‘the company way’. And methods to strengthen this are precisely what we can see the Carolingian rulers employ, whether it’s the bonding rituals of shared hunting/warfare/religious rituals etc, the social role of the assemblies or the attempts to create an ethos of ‘secular sanctity’ among the lay nobility. (And what do the eager young men at Carolingian courts remind you of so much as management trainees?)

This also explains what seems to modern eyes one of the biggest weaknesses of Carolingian government: the tendency for comital and other office to become hereditary. To the modern rationalist, this seems obviously absurd, but there are parallels in business practice. I once worked for a company that recruited largely by word of mouth: most of the people coming in were friends of the current staff. Although this would now probably run foul of discrimination legislation, there was a rational business purpose. It was seen as lowering both the cost of recruitment and the risks: you employed someone who was a ‘known’ quantity, the right sort of person for the job, rather than possibly getting someone totally unsuitable. A very similar attitude is seen in the few explicit comments in Carolingian sources on how you might choose successors to counts: the idea is the virtuous sons of previous officials. Heredity determines the pool of applicants: these are men who will have had the right upbringing and knowledge; competence (in the sense of virtue, which was the main par of the job description for a count) determines the actual choice. Such recruitment in one’s own image in the long-term is likely to damage an enterprise, through lack of diversity; in the short-term, it’s an easy way of avoiding the truly bad official. (The Carolingian use of non-Franks and the fact that not all the officials came from the Reichsadel also suggests there were attempts to make a slightly wider pool available).

Conclusion

Much discussion of Carolingian government still seems to start from two opposing positions. One is the assumption that all good government should be a bureaucracy, much along the lines of 19th and 20th German, French and English ones. The other is that early medieval government is fundamentally different from modern ways of thinking. What I’ve tried to show is there are other modern models of organising enterprises that can give us insights. (As to why Carolingian government look a bit like some modern companies, there’s an obvious answer: they’re both trying to run a large-scale organisation as cheaply and with as few people as possible, and they’re not fundamentally worried about ‘fairness’ of outcomes).

I’m not sure how effective Carolingian government actually was and we may not actually have the sources to know. Our picture of Carolingian administration, after all, is largely based on legislative and administrative orders, a few court cases and political polemic. If you judged modern government on the basis of the statute book, the Times law reports and tabloid headlines, you’d probably have a very odd view of that as well. What I am sure of is that much of the criticism of the effectiveness of Carolingian government has been based on some fairly dodgy assumptions, and in particular an ideal of the modern state that seems very unrealistic. Perhaps the first step to a valid assessment of Carolingian effectiveness is to look most critically at the effectiveness or lack of it of current organizations: whether the civil service, private-sector companies or even, perhaps, universities.

I’ve just started turning my thesis into a book on Carolingian masculinity and morality and in the introduction face again answering some of the objections I faced to my work. I think I’ve got a reasonable argument for why it makes sense to study masculinity, but in discussing why it makes sense to study morality, there is one big problem. Did all the vast amount of Carolingian moral texts (sometimes it seems that the Franks, like A.P Herbert’s English, never did anything except for a moral reason) actually make a blind bit of difference?

This is only secondarily a problem about the scarcity of early medieval sources (or even about texts and realities). It’s really a problem about another gap: between attitudes and actions. How do you assess (even today) the views of a society (or a group within society) on a particular moral norm? One take on this is that actions speak louder than words. So you look at whether there’s adherence to a moral norm. The problem is, does adherence necessarily correlate with attitudes? As one interesting example, if you judged by abortion rates, the people in the US are far more positive towards abortion than in the Netherlands. Similarly, you might conclude from the higher US murder rates that murder is taken less seriously there. As this shows, even if you can get accurate statistics on behaviour (and my examples are some of the relatively few acts concerning sex or violence where you can), the interpretation of them is tricky. Given the complete lack of statistical data from the early Middle Ages, adherence doesn’t look very helpful as a measure of the impact of moral norms.

What does legislation on a particular moral topic tell us about wider views on it? It’s perfectly possible to have a wide gap between the legislation passed and majority views, even in democratic societies: for example, capital punishment in Britain was abolished by MPs on a free vote long before public opinion as whole was opposed to it. In the case of early medieval law-making, carried out by the king with the supposed consensus of the ruling elite, it does seem reasonable to assume that the political class as a whole accept the moral norms conveyed, at least nominally. A trickier question is how much emphasis should be put on enforcement of legislation. There’s an argument that a lot of early medieval legislation is in fact largely symbolic (or in the term that’s preferred ‘programmatic’); it doesn’t make much practical difference. But increasingly, I can’t help feeling that a lot of modern legislation is mainly symbolic as well: from Section 28 of the Local Government Act via the Dangerous Dogs Act to all the tinkering round with Criminal Justice Acts. That doesn’t mean that people don’t feel strongly about such matters. On the contrary, much of this kind of legislation is passed precisely because people are very concerned about a moral issue (such as the growing acceptability of homosexuality) but there’s no practical way to stop particular behaviour/changes. Similarly, repeated legislation tends to be a sign of a feeling that ‘something must be done’: it doesn’t necessarily correlate with the actual scale of the problem. (For example, violent crime in the UK has actually been decreasing since the mid-1990s, but that doesn’t stop constant demands for more laws and more determined enforcement of them). There’s also the problem that the amount of legislation (or discussion) on a particular moral issue doesn’t necessarily reflect the actual importance of the issue: as has become notorious, there was far more parliamentary time discussing and legislation on banning hunting with dogs than on starting the Iraq war.

How else might you find out about moral attitudes? Nowadays you can use opinion polls, but aside from issues about leading questions and the unwillingness to tell the truth, that’s not much help for most of historical research. The other main way is by listening to what people say: either at a micro-level or via reflections of ‘popular opinion’ such as the mass media. There are always issues here: is what you hear representative, how much are people saying what they think they ought to say rather than what they really think, etc? Discourse of this kind, however, does seem to me a reasonable way of assessing attitudes, because so often the way the debate is framed is crucial. (It’s extremely difficult, for example, to talk positively about homosexuality if you start from a term like ‘sodomite’, just as the unofficial renaming of the UK community charge as the ‘poll tax’ and US inheritance taxes as ‘death taxes’ were politically devastating moves).

So in my assessment of how seriously moral norms were taken I’ve tended to focus on discourses and also collective legislation. I’ve paid less attention to evidence on adherence to or enforcement of norms and I haven’t assumed that repeated legislation is always a sign of ineffectiveness. These are all, as I hope I’ve shown, logical decisions. But taken together, I now find myself worrying: am I putting my thumb on the scales? Have I rigged my criteria so I’m bound to find that early medieval noblemen took moral norms seriously?

If I look at my underlying assumptions, I think I do presume generally that people normally pay attention to ‘rules’ (notice, this is not the same necessarily as obeying them). I suspect that some other historians I know start from the assumption that of course people don’t pay any attention to rules (possibly our different views start from differences either in our personal temperaments, political outlook or religiosity). There’s no way to tell which is a more accurate assessment of human nature. So, to approach the problem from a different angle, what would it take to persuade me that particular moral norms weren’t seen as important by laypeople? (If there isn’t evidence that would change my mind, I’m in the realms of faith-based history, which is always suspect).

Well, in some cases, we get told that explicitly by those promoting the norms. Jonas of Orleans says that men who are told they shouldn’t sleep with their pregnant wives should not laugh at the suggestion. And that demand only turns up in one council in the Carolingian period (one at which Jonas was particularly influential). So that moral precept can certainly be marked down as non-normative. Otherwise, I think you can only attempt to judge whether there seem to be wide-scale and blatant disregarding of moral demands. I don’t think, for example, that you can see the church’s ban on tournaments as taken at all seriously by laymen; not only do they carry on with them, there’s no attempt to excuse them. However, if you compare some of the eleventh century marriage disputes that Georges Duby looks at in ‘The knight, the lady and the priest’ and Carolingian period ones, there does seem to me a difference. There are relatively few ninth-century cases where the church’s rules on marriage and divorce are simply ignored (as happens e.g. with the Capetian kings Robert the Pious and Philip I). Instead, what you see is attempts to bend the rules. The aims are the same - to get a desired marriage, to get out of one no longer desired, but the tactics are different. It seems to me that you can therefore say that moral norms about indissoluble marriage are more important to Carolingian noblemen than later on. On the other hand, when ninth-century Catalonian counts are marrying their first cousins, you have to say that they at least don’t care two hoots about 100 years of incest regulations. In the end, the evidence on most moral topics isn’t clear-cut. All I can do is try and set out my assumptions and data reasonably clearly; at least that way someone with a different view can see my justification for my conclusions, even if they’re not convinced by it.

I’m just back from a colloquium in Cambridge on ‘La Culture du haut moyen age: une question d’elites?’ (programme here), which had a group of scholars (mainly Paris and Cambridge-based, but with a few outsiders like me as well) considering many aspects of elites and culture between late antiquity and the eleventh century. It was an interesting but exhausting time and I’m only now trying to connect together some of the themes raised.

One of the things that did emerge was how different late antiquity (here taken to be the period up to about 600) and the early (Western) Middle Ages are with respect to culture. There was a long classical Roman tradition in which the political/social/economic/religious elite are either the cultural/intellectual elite themselves or at least their patrons. Those at the top are supposed to be educated in high culture and appreciate it and will lose social prestige if they’re revealed to be too ignorant. (This is actually a common phenomenon in many European and non-European societies for most of history. It’s only really in the twentieth century that you can find much of a cultural/intellectual elite that doesn’t rely on upper class patronage for support, but can make a living in other ways).

This Roman cultural system survived a number of shocks to it: the fall of the Roman Republic, the conversion of the Empire to Christianity and even (for a while) the political fall of the Western Empire. Men (and women) were still being educated in a traditional Roman frame of reference, even if the numbers were increasingly dwindling. Cassiodorus and some of his circle still seem very much senators in outlook, even if not in practice.

Culturally, the real changes from late antiquity are when a) the traditional schools disappear and b) literary education becomes optional for the lay elite. Education doesn’t cease with the end of the Roman school, but a shared educational tradition is vital for perpetuating an intellectual elite and the new monastic and courtly routes to education probably didn’t provide a critical mass of students in the sixth to eighth centuries. It’s very difficult looking at what happens to lay culture in the post-Roman or sub-Roman world from the sixth century onwards, essentially due to a lack of sources. Guy Halsall, 'Gender and the end of empire', Journal of medieval and early modern studies 34 (2004), 17-39 argues for the adoption of barbarian cultural traditions by ‘Roman’ men, based largely on archaeological evidence. Other than that, we’d reduced to saying ‘Oh, you can use Beowulf to study lay culture’, which is a bit like trying to reconstruct US society from Hollywood films. (I do use heroic literature in my own research, but largely to explore moral attitudes, which I hope is less problematic).

So that’s the end of late antiquity. At the other end of the process, institutional structures for monastic and higher-level cleric education essentially get re-established towards the end of the Carolingian period, with the development of networks of clerical and monastic schools, which are the basis for tenth and eleventh century Latin culture. Elite lay people reappear as audiences and sometimes authors of written cultural traditions in the eleventh and twelfth centuries (at least in Francia - I’m aware of the problem that I’m doing what one of the speakers of the conference did and treating Italy, Byzantium, England and Spain all as exceptions).

What this suggests is how much is up for grabs culturally in between these two periods: essentially in the late Merovingian/Carolingian era. Most of the papers at the conference were about this period and that actually makes quite a lot of sense: it’s when the relationship between elites and culture is being reworked. It’s one of the few times when there is a question-mark about whether culture is really a matter of elites, and what/who those elites are going to be. As I’ve already said, new monastic and clerical educational structures get put into place. I argued in my paper that part of the Carolingian reforms were an attempt to create a new educated lay elite that was also a moral elite, but that that was abandoned in the 840s. (This effectively meant that literary education remained optional for lay noblemen for another couple of hundred years). Meanwhile, the entire basis of literary culture was on shifting ground because of the evolution of proto-French and other proto-Romance languages. I’m not yet sure how you can fit together these different developments (or indeed if you can), but it does suggest a key role for the Carolingian period in the development of medieval culture.

I’ve finally had the chance to finish reading the Harry Potter series (or rather, I’ve shamelessly neglected my research in order to do so), so I wanted to blog a bit about a subject that interests me: J. K. Rowling’s linking of morality and outcomes. Every creator of narrative fiction (whether a novelist or a screenwriter/director) is faced with deciding such questions: do characters get what they ‘deserve’? In the most blatant form (such as many stories for small children and rather too many for adults), the nasty suffer and the good are triumphant. Those who are basically good but with some moral flaw are redeemed, though whether this means surviving or not varies. (One of the main moral differences between Charles Dickens and Wilkie Collins is whether or not fallen women have to be killed off). There are normally also some innocent victims, although how central they are and how much they suffer is also a variable.

A writer who wants to express a strong moral point of view in their story (like Rowling) has to decide how much to adhere to this morally determined set of outcomes and how much to admit the reality of chance: that the good, too, die, sometimes in ways that don’t make sense. To ignore random, meaningless suffering is not to be true to the realities of life that growing children have to realise. To make too much of randomness encourages the idea that it doesn’t matter what moral choices you make. I think Rowling in fact makes some quite clever choices in this, within a basic scheme of good conquering evil.

Rowling’s strength, it seems to me, is in deciding how the many characters with some good in them are treated and this is particularly reflected in whether or not they survive and the manner of their deaths. There are a number of innocent victims of evil and she is sufficiently clear-eyed to show that the deaths of some of them have no redeeming or heroic quality. (It’s noticeable that it is only Lily Potter’s sacrifice of herself for Harry that creates magical protection, not his father’s). There has been criticism by some adults that she didn’t kill off enough central characters in the last book (an unusual criticism of a children’s novel), but in fact I think the death of Fred Weasley (after the earlier maiming of George) is quite a blow. The twins represent much of what a certain kind of teenage hero is normally like; what Harry himself could have been if it hadn’t been for the burden he carries. And she does prove herself tougher than Tolkien in one respect: Dumbledore, unlike Gandalf, is truly dead.

The most interesting bit of Rowling’s attitude is what happens to those who (whether or not basically good) make a really serious moral mistake, choose at a key moment to do the wrong thing. Here there seems to be an age divide: teenagers who make such mistakes are normally allowed to survive; adults are likely to die. For example, Sirius Black’s lack of concern for house-elves kills him, while Ron is able finally to overcome such feelings. Of the two characters ‘redeemed’ by Dumbledore, Draco survives, but Snape doesn’t. It also seems somehow appropriate for Snape that he doesn’t get a big heroic death (e.g. saving Harry); it’s only a while after his death that his courage is realised. One of the most morally-charged revelations of the book is that Dumbledore didn’t actually die because of an impossibly noble sacrifice. He was going to die anyway, before he brought about his own death slightly prematurely. And that was because he’d make the fatal mistake of putting on a cursed ring: he got distracted from what was truly important, combating Voldemort. Harry makes a similar mistake and gets captured; but it is because of this that he is able to take Draco’s wand and so defeat Voldemort at the end. In other words, at times Harry has sheer dumb luck. This may be unrealistic (a lot of teenagers in real life do reckless things that kill them), but it does provide hope for teenager readers who know the limits of their own good moral judgement.