Feudal Transformations VII: Michel Bur and the motte-and-bailey castle

I don’t plan to talk about each of the papers in that Spoleto volume individually, but Michel Bur’s contribution is a very good example of the kind of explanation of the changes in European society around the year 1000 that annoyed me enough to draw my famous diagram.1 His argument, very basically, is that at the end of the tenth century new motte-and-bailey castles proliferate all over northern France, and that this must have fundamentally affected the way that areas are run, bringing the collection of surplus much closer and, well, transforming society. And certainly this has some weight, but it’s only part of an argument. There are two ways to tackle this, what after one of the earlier posts we might call the Bedos-Rezak way, that is invalidate it by logical and theoretical argument, and the Jarrett way, which is to punch it full of exceptions until it deflates. First let’s do the theory.

Model of motte-and-bailey castle, from Wikimedia Commons

Model of motte-and-bailey castle, from Wikimedia Commons

It is certainly the case that one of the changes in society that characterises this period is the way castles pop up everywhere like plague. In Italy, where the phenomenon was first really studied, they call it Incastellamento and the effects on society of a new local and unassailable structure of domination are well worked out.2 But it’s not as if the technology to achieve these sites is unthinkable before this time. The Romans could throw up fortified camps many times this size as part of an afternoon’s work, and if you follow the campaigns against the Vikings in the Anglo-Saxon Chronicle or the Annals of St-Bertin you’ll find each side putting up forts against each other as if it were no big thing.3 So, why isn’t everyone doing it before this? What’s changed? What, in fact, explains the castles?

Dépiction de l'action d'une bêche intellectuelle d'un archéologue, comme vue par Jonathan Jarrett

Dépiction de l'action d'une bêche intellectuelle d'un archéologue, comme vue par Jonathan Jarrett

The diagram above illustrates what’s going on here. Castles are part of a larger cycle of causation in this process, and because M. le Prof. Bur is an archaeologist as well as a historian this is what he knows best of all and sees in most detail. Above therefore we see the intellectual spade of M. Bur exposing the castles to the historical view, leaving the fact that they in turn were permitted to arise by other phenomena buried. So, what about these other phenomena? Time for the Jarrett view.

This is one of those things where Susan Reynolds has a good grip on the problem of explaining it all. We know that later kings police the building of castles very tightly, because it is by then well understood that just this process of local Incastellamento is downright detrimental to their command of an area. If their armies can be diverted or resisted they have no recourse against their vassals, and so on. So it is assumed that the Carolingians and their ilk were just as avid against local castle-building when in fact this is very hard to show. Certainly there are Carolingian cases of rebels’ castles being demolished, but these aren’t adulterine castles in the later English sense, these are just castles that were in the wrong hands. Out in my area, in the supposed cauldron of the Feudal Transformation, pretty much the first thing a lord does when he opens up a new area is put a tower up in it.4 You’ve seen some of these towers in the blog already:

The Castell de Tona, Osona, Catalunya, photo by Jonathan Jarrett

The Castell de Tona, Osona, Catalunya, photo by Jonathan Jarrett

Tona here boasts a tower whose foundation date is unknown but which was certainly not a refuge for the townsmen; you can barely get two people inside. And Òdena, founded by Sal·la of Bages as I told you a while ago, is not so very much bigger. I could tell you more, but I don’t have the pictures to hand to make the point.5

What remains of the Castell d'Òdena, founded by Sal·la and Unifred

What remains of the Castell d'Òdena, founded by Sal·la and Unifred

Now certainly Northern France was not a frontier area like Catalonia, and the obvious need for defensive towers and guardposts is rather less. All the same, there was no fear on the part of the counts of these castellans-in-the-making. The counts made a good few of them, they wanted this area castled, because castles brought them power. How? They could demand military service from these places, they could use them as bases, they could if they had to maintain garrisons out there but they’d far rather have had someone else paid for that.6 And perhaps this was a dangerous game to play, gambling on the biddableness of their fideles in time of trouble, but the prize was arguably worth the gamble, because the prize was military domination of an ever-increasing area and a considerable increase of surplus as the newly-protected areas around the new castles filled out and become productive. You can actually, with Catalonia, put castles into a different chain of causation with land clearance and therefore economic growth, though which caused which would be a bit more of an argument.7

So Michel Bur is surely right to stress the importance of castles, and when Dominique Barthélemy tried to dismiss his findings in questions by saying that the ‘wave’ of castles at issue were spread over most of a century, Bur was quite easily able to answer with “no they’re not, I’ve dug them and you haven’t” or words to that effect.8 But Barthélemy’s argument is the wrong one to use. The castles were important, and not just in Northern France either; but there are so many other things that are important too happening at the same time that this is not the answer we need, if it even exists.


1 Michel Bur, “Le féodalisme dans le royaume franc jusqu’à l’an mil: la seigneurie” in Il Feudalesimo nell’Alto Medioevo (8-12 aprile 1999), Settimane di Studio del Centro Italiano di Studi sull’Alto Medioevo Vol. 47 (Spoleto 2000), pp. 53-78 with discussion pp. 79-83.

2. Originated, I believe, with Pierre Toubert in (or rather translated from) his Les structures du Latium médiéval: Le Latium méridional et la Sabine du IXe siècle à la fin du XIIe siècle (Rome 1973); international findings gathered by him and Miquel Barceló in L’« Incastellamento »: Actes des Rencontres de Gérone (26-27 Novembre 1992) et de Rome (5-7 Mai 1994) (Rome 1998).

3. Easily accessible in M. Swanton (ed./transl.), The Anglo-Saxon Chronicle (London 1996), and in the older translation of J. A. Giles online here, and in Janet L. Nelson (transl.), The Annals of St-Bertin (Manchester 1991), online here for those with subscriptions.

4. So, for example, the Torre dels Fils de Guadamir, as it’s briefly known, seen in Ramon Ordeig i Mata (ed.), Catalunya Carolíingia IV: els comtats d’Osona i Manresa (Barcelona 1998), doc. no. 1472; see Jonathan Jarrett, “Pathways of Power in late-Carolingian Catalonia”, unpublished Ph.D. thesis, University of London 2005, pp. 183-184. The tower was apparently a family venture, perhaps by the sons of an estate manager of the Cathedral of Vic. It seems to have rapidly disappeared or else, more likely, changed its name.

5. Tona was certainly standing by 889: see Ordeig, Catalunya Carolíngia IV, doc. 9. For Òdena’s connection to Sal·la see ibid. doc. no. 769.

6. The military service more or less assumed, see now Ramon d’Abadal & Josep María Font i Rius, “El regímen político carolingio” in J. M. Jover Zamora (ed.), Historia de España Menéndez Pidal, tomo VII: la España cristiana de los siglos VIII al XI, volumen II. Los Nucleos Pirenaicos (718-1035): Navarra, Aragón, Cataluña, ed. M. Riu i Riu (Madrid 1999), pp. 427-577 at pp. 467-503. Some form of comital control over castles is evidenced from the apparent resumption of the castle of Maians near Bages by Count Borrell II between 966 and 972; given to the monastery of Sant Benet de Bages in its 966 endowment (Ordeig, Catalunya Carolíingia IV, doc. no. 995A) it was gone by the monastery’s church consecration in 972 (ibid. doc. no. 1127), when the scribe used the monastery’s lack of military property as a moral high ground, and Unifred, Sal·la’s son, later held it from Borrell (ibid. doc. no. 1238). On all this see Jarrett, “Pathways”, pp. 38-46.

7. So, for example, economy driving lordship in Pierre Bonnassie, La Catalogne du milieu du Xe à la fin du XIe siècle: croissance et mutations d’une société (Toulouse 1975-1976), 2 vols; the other way round in Gaspar Feliu i Montfort, “Societat i econòmia” in Federico Udina i Martorell (ed.), Symposium Internacional sobre els Orígens de Catalunya (segles VIII-XI) (Barcelona 1991-1992), 2 vols, also published as Memorias de le Real Academia de Buenas Letras de Barcelona Vols. 23 & 24 (Barcelona 1991 & 1992), I pp. 81-115.

8. Again the delight of having the discussion printed! After Barthélemy’s harangue, Bur begins his response with (I translate): “Thankyou for that bracing statement of a case. It permits me to be clear in my turn.” In academese of course that’s tantamount to rolling up sleeves and asking him to step outside. Let no-one ever say that good manners prevent you being rude.

27 responses to “Feudal Transformations VII: Michel Bur and the motte-and-bailey castle

  1. This blog is turning into quite an historigraphical resource. I can imagine grad students’ delirious joy at your sensible explanations of problems and arguments.

  2. If you can call hand-sketched thought diagrams with illegible French labels sensible… but thankyou, I’m glad it reads as if it might be useful.

  3. Speaking for grad students, of which I am one, I know several who actually print your posts to study in-depth as time permits. They don’t always speak exactly to our particular projects, but they are well written, informative, and give us a more informal way of exploring ideas and areas we wouldn’t normally.

  4. Thanks for the very good post, but can I also suggest that we don’t forget the Reuter way to tackle these questions, and ask: ‘Were they doing the same thing in Germany?’ If they are (and as far as I know they were building castles there too), that scuppers that as THE cause of the F word.

    Of course, medieval Germany is Different, which is a Bad Thing (as opposed to medieval England being Different, which is a Good Thing, and medieval France being Typical, which is a Good Thing), but it’s still a very handy rule of thumb, just as any suggestions on the causes of the ‘Fall of the Rome’ have to pass the ‘But did it happen in the East as well?’ benchmark.

  5. Michael: that’s marvellous but horrifying. For once I’m actually glad I make myself do the footnotes, just so I know that I have some basis for what I’m saying…

    Magistra, it is the key question, I agree. Incastellamento plainly isn’t the root cause of the supposed transformation anyway, as I think I demonstrated, though obviously if your local lord sets up a castle just over your village it will make a difference to what you can do to resist his demands! But the question is who allows him to? In France the king loses the ability to demand his rights from castles, and we slowly get this idea of adulterine castles and permission developing. In Germany, by contrast, the king may not control the castles, but the big nobility more or less do, and the kings are the big nobility in Saxony for a while; one of the interesting things about the Henry IV situation is that you can see him trying to hold this supposed heartland and it slipping like sand through his fingers. If we really understood why that happened, and Leyser’s article about the crisis of the nobility is a big step towards it, we’d really have a good demonstration case for the loss of royal power elsewhere. All the same, I don’t think the Reuter argument helps here, because we can say how Germany is different, and it’s that the big nobility keeps control of castles in a way that can’t be maintained in the West. One of the things that I’m getting from this volume is an idea that fiefs are what you make to keep people in your service when you haven’t got much clout left otherwise. In Germany they just don’t become necessary till much later if ever, although some people have reason to use them, whereas the argument is that in the West they’re general. I think that’s probably questionable but I think it is true, especially in Catalonia, that higher or older public structures stop being sufficient and effective in a way that doesn’t occur in Germany, and that one of those public rights is access to and use of fortifications.

  6. you page is very inter esting.

  7. this is realy boring by my self.

  8. IT COULD BE IMPROVED VERY VERY GOOD

  9. Wow! This is a neat concept. I must say as a grad student in tech history this is invaluable for giving me a more current sense of scholarship on the subject. I enjoyed DeVries Medieval Military Technology, but it is 18 years old. Time to go find articles to corroborate what I’ve read here….

  10. Pingback: I should have read this the moment I bought it, III | A Corner of Tenth-Century Europe

  11. Pingback: Feudal Transformations VIII: two ways of confusing the issue | A Corner of Tenth-Century Europe

  12. Pingback: A Black Country Motte-and-Bailey | A Corner of Tenth-Century Europe

  13. As I see it, incastellation outside frontier societies (i.e. Carolingian Catalonia)/ conquered regions (i.e. Norman England) is mainly a matter of changing political culture. I think that provides the best explanation for why aristocrats in, say, ninth century Carolingian Francia, late Anglo-Saxon England and to some degree Ottonian and Salian Germany (pre-Henry IV) aren’t throwing up castles everywhere; because they have other ways to display their status and establish/ maintain local power and because there isn’t a culture of endemic private warfare. Yet, as always, it remains that there isn’t a one size fits all explanation. That’s why I think your historical magimix is such a helpful way of thinking about the feudal transformation – in each region its always a different combination (in varying orders) of a shared pool of factors that lead them all, ultimately, to very similar social setups.

    • I suppose I agree with that, but I feel like the argument is a bit of machinery I need to take out and examine in case it’s backwards. The reason for that feeling is that there’s actually a lot of aristocratic rebellion and violence in all those areas; you have the Brüderkrieg and the fragmentation of Carolingian authority (and all the rebellions against Charles the Bald especially), the notorious weakness of Æthelred II’s enforcement of his judgements and the fraternal and other rebellions against the Ottonians and then much more so against the Salians… These look like contexts in which you would expect local bases of power to be set up and defended. I suppose, though, that the difference in all those cases is that what was being fought with and resisted was still royal authority, rather than royal authority having become so irrelevant that the principle lordly horizon was whoever was next-door, the persistence of the relevance of Königsnähe rather than Königsfern. And maybe if the king is coming to get you, you have to hole up in something bigger than a motte-and-bailey… But I still don’t buy it so easily for Æthelred’s reign, because as far as we can see he never campaigned against recalcitrant subjects like Otto I against brother Henry or similar. There a good old-fashioned burh could apparently be enough to forestall justice for years… For my money it’s just that they don’t ‘count’ as castles (and the fact that the one such site that’s been identified, Goltho, probably isn’t that thing, doesn’t help us see the phenomenon) which has led to incastellamento having to wait for the Normans.

      • I’d completely agree with you there – those are some much needed points of nuance to my initial tentative argument.
        From what I’ve noticed, the fundamental pattern with Carolingian and Ottonian rebellion is that the aristocracy always have to rally around a disaffected member of the royal house or an external royal with a decent claim to rule there. We can see this in the 850s with Charles the Bald and Aquitaine – the Aquitanian magnates unhappy with Charles’ rule, but not keen on the washed-up Pepin II either, decide to rally around Charles’ East Frankish nephew, Louis the Younger, and make him their king instead. Sometimes, of course, the king to rally around is within their midst, as the Provencal and Burgundian magnates find with Boso in 879 or the East Frankish magnates with Arnulf in 887.

        Then in the 890s, when many of the West Frankish magnates decide they don’t like Odo’s heavy-handed rule. they rally around Charles the Simple. The reverse then happens again in the 920s with Charles the Simple and Robert of Neustria. Meanwhile, Charles’ Lotharingian subjects decide to switch their allegiances to Henry the Fowler, and Duke Gislebert decides to really go in for the konigsnahe by marrying Henry’s daughter, Gerberga.

        In the Ottonian kingdom we see this with how the Saxon magnates, unhappy with Otto the Great’s more elevated, authoritarian style of kingship in contrast to the more subtle, diplomatic methods of his father, Henry the Fowler, decide to get behind Otto’s disaffected, disinherited brother Thankmar. Then after Thankmar and the rebels in Saxony are defeated in 938, the Lotharingian magnates, specifically Duke Gislebert of Lotharingia, Count Dirk of Holland and Count Isaac of Cambrai, decide to rally around Louis IV, given that he’s a Carolingian and his father was king in Lotharingia, in hope that he’ll better serve their interests instead. Meanwhile, Eberhard of Franconia and Eberhard of Bavaria rally around Henry, Otto’s younger brother born in aula regis, to try and get more favourable kingship from him – seriously, Otto the Great was one of the luckiest monarchs in early medieval history in that these three rebellions across the realm in the span of just two years didn’t end his reign then and there. We see a sort of repeat of this again in the 950s, with Henry up to his tricks again and this time with Otto’s son Liudolf being thrown into equation.

        And back in West Francia, Herbert of Vermandois is able to rebel against King Raoul because he’s got two dynastic trump cards up his sleeve – his own direct descent from Charlemagne and (until 929) the captive Charles the Simple. What goes on with Louis IV and Hugh the Great is a bit more of an outlier in that respect, but even there its all about who should really exercise effective control over the royal court + the crucially important archbishopric of Rheims.

        The fundamental dynamic of Carolingian and Ottonian political culture is, as you suggest, that the royal court remains the main forum for political action, to the point that no one can go without it for long, and kingship is still nigh-on universally exalted, with everyone looking to the king/ a claimant to the throne to solve their problems. With this outlook, castles are not at all necessary. If you want an easily defensible hideaway when your straits are looking dire, its much easier to just go and hide in old Roman fortified cities, which are a nightmare to siege unless you have a sufficiently large army, as Flodoard and Richer make clear in their accounts of the events of Louis IV’s reign.

        But when it comes to eleventh century rebellions, certainly in West Francia, follow a very different pattern. When aristocrats take up arms against kings, they no longer require a royal figurehead to legitimise them. And there are just so many aristocratic quarrels, rebellions and private wars that have no reference to kingship at all – where is King Robert the Pious in the Conventum Hugonis! Aristocrats are now content instead with pursuing local domination of their own coherent territories and protecting them from their neighbours, with castles being the central instruments for both aims.

        But yet the big question is, why should this shift happen? Is it because the royal court finally loses pulling power (the Matthew Innes view)? Is it because of the accumulation of local resources (the Georges Duby view)? Is it because kings, somewhere down the line, lose nerve and cop-out, leaving regional aristocracies on their own? Or does kingship somehow get ideologically discredited? Or a combination of all four? A lot of this then raises further questions of how castles relate to it all in terms of causation, with the possibility that my argument gets invalidated and we’re back to square one. But I think its right to see the political outlook of the aristocracy as the key ingredient in whether incastellation does or does not happen.

        As for Anglo-Saxon England, whether burhs count as castles or not, I’d say I’m very much on the public fortress/ fortified towns side of the debate. And yeah, Goltho and Sulgrave definitely don’t count as castles – I remember John Blair saying to me that the whole idea that there was a ninth century fortified site there was “pure fantasy”, though with the early eleventh century excavations he suggested that it might have been built for one of Aethelred’s ministerial thegns (so not exactly a “private” fortification) in the wake of the renewed Danish invasions after 1002. And while teh Anglo-Saxon state might not have been as strong as some argue it was (thanks so much for making me aware of Aethelred’s difficulties in enforcing royal justice), I think Campbell was definitely right in arguing that there was no Peace and Truce of God in eleventh century England because there simply wasn’t that kind of culture of private warfare in Anglo-Saxon England. Aristocratic rivalries in Anglo-Saxon England, as far as I know, seem to have mainly ended in the murder or exile of one of the parties, and like with Carolingian and Ottonian ones they revolved around offices/ position at court.

        • Well, indeed! I actually think I just said much of that, just a lot shorter :-) John Blair is also my source for the discredited Goltho burh, though I did find it a convincing rebuttal (and don’t know if he has actually published it to be cited). The main nuance I’d add to your nuance is that there is some difference between rallying round whatever Carolingian is available – very often a preferred option, I completely agree – and deciding that Odo’s ten years in power makes the Robertians royalty-in-being enough, especially when you then start to add the ties by marriage and other closenesses that you need to make your picture work in Germany too. Something is happening there in which Carolingian blood is no longer the main qualification for kingship, and – if I can borrow a point from Fraser McNair here – one of the proofs of it is that no-one at the time mentions or even seems to know about that Vermandois direct male-line descent from Charlemagne. No-one tries to make the Vermandois into kings. why not? They’re theoretically superbly qualified. But actually, that isn’t a qualification; it’s something more immediately throne-worthy than just descent, and for me the first step towards that is everyone getting fed up with Charles the Fat in 887.

          As for Æthelred, while I’d cautiously stress disorder and disloyalty as a theme of the reign, I would tend to agree that it isn’t, Eadric Streona aside, of the level that disables as effective a state system as Campbell has taught us to see (and Molyneaux has nicely boxed off) in tenth-century England. But Eadric is quite a big aside, and I still think that the relative fixity of Æthelred’s court personnel shows a élite which was partly already opting for Königsfern, for now, over Königsnähe. In fact, that’s an interesting thought case, because Cnut seems to reverse the trend (while relying only even fewer major nobles). How, and why couldn’t the Carolingians? Resource shortage shouldn’t be the problem here… (Sorry, I’m crossing threads here…)

          • Apologies for the late reply – been very busy lately with work then I went on holiday to Whitby (very nice area).

            I think Fraser is very much on the money with that remark, as he is with a lot of things, which is also a reminder that I need to read Stuart Airlie’s “Making and Unmaking the Carolingians” at some point. The significance of 888 can be overplayed, but at some point around then the rules of the game really do change – Boso of Provence in 879 is probably the first step – and being a Carolingian is no longer what you need to qualify for kingship. But, and this is where Fraser is being really insightful, we need to invert the familiar narrative, going back to Regino of Prum and Notker the Stammerer, somewhat. Its not just that so many men with equally good claims all qualify for the kingship, but some men with superior claims mysteriously don’t. Herbert of Vermandois is obviously the case in point, but he’s not the only one. Why didn’t Baldwin II and Arnulf I of Flanders get involved in the West Frankish succession in 888, 893 – 897, 922 – 923 or 936? With their descent from Charlemagne (via the female line) surely they were as qualified to be kings in West Francia as, say, Guy and Lambert of Spoleto, Louis the Blind (also ruling in Provence) and Berengar of Friuli were for Italy? Either that’s just not the game they want to play, or for some reason (like with the House of Vermandois) no one considers them throneworthy. Instead, why did the West Frankish magnates opt for Odo, who really did have not a drop of Carolingian blood in him, and whose family (whatever the real origins of the Robertians were) were relative parvenus among the West Frankish nobility? And then why do they rally round his brother again in 922? Obviously it takes a very long time for a discourse of Robertian/ Capetian royal legitimacy to properly develop, certainly long past 888 or 987, but something must have changed in determining what makes you supremely qualified for kingship.

            But linking in the point about Aethelred and wrapping up the broader thread, I think that ultimately what makes early medieval kingdoms/ states (I do think Carolingian Francia, Ottonian Germany and late Anglo-Saxon England were all states, despite the protests of some to the contrary for each and all of the three) successful and enduring is the degree to which the elites buy into them. We can see it across the Middle Ages. Like (as Paul Fouracre et al have shown) the Merovingian realm didn’t completely fall apart in the late seventh century despite being ruled by lots of weak, short-lived kings. And, at the other end of the medieval period, the general consensus among historians of fifteenth century England is that the Wars of the Roses weren’t “a return to the old feudal amenities of sackage, carnage and wreckage” (the traditional view ingeniously parodied by Sellars and Yeatman), but were mostly about getting the system to work again after the disastrous kingship of Henry VI. Carolingian Francia, Ottonian Germany and late Anglo-Saxon England seem to be no exceptions to this.

            It is here that I introduce my theory that there are essentially three alert levels for royal authority/ the coherence of medieval kingdoms. The first is disorder and simple defiance – in an age of premodern communications, no ruler , however personally able, is going to have their authority felt equally everywhere in their realm and this is true across the Middle Ages. And when much of the state’s coercive power rests on the co-operation of militarised landed elites, then that gives them a lot of leverage. The next level is crisis of a regime, when a particular dynasty or ruler loses grip on the reigns of power due to widespread dissatisfaction with their rule, but the elites still buy into the system so they try and get in a new king/ dynasty – this is undoubtedly the stage Aethelred got to in the end; Charles the Bald went there in the 850s; Charles the Simple in the 920s; Otto the Great in the late 930s, we can go on. The third stage is system breakdown, when the elites stop buying into the system and decide its best to go their own way – this to me seems to be where West Francia ends up between 987 and 1031, and this is accompanied by a proliferation of castles. Meanwhile, the English elite still buy into the system and want it to work, so they ditch Aethelred for Cnut, and I’d argue that the late Anglo-Saxon was probably at its most powerful under him. I’m afraid I can’t think of an answer to your question as to why West Francia couldn’t have its Cnut at whatever point, but I think its undoubtedly the question we need to be asking …

            • I’m not sure, myself, that there was a system of monarchy in operation which could explain all these things at once. Odo’s succession, I put down to his being a successful commander and perhaps no. 1 magnate at a time of military desperation; and the fact that that was temporisation brought about by lack of a competent adult Carolingian I think is clear from the way he conceded the succession to Charles without much struggle. But his having been king then raised the possibility for brother Robert when it became clear that Charles was adult but maybe not competent (a long-term subject of disagreement between myself and Fraser). The counts of Flanders’ apparent disinterest in kingship, however, I put down to two different things, one being their security in their own position versus the gamble of Frankish over-rule with their limited resource base and the other being the position it would have put their core, frontier, territory with respect to the Ottonians. That is to say, I think they thought they couldn’t get away with it and it might not have been worth it.

              I am happier buying the three alert levels as a thinking tool, though I’d note that a Jinty Nelson piece I was re-reading only this morning, her “Rulers and Government” in the New Cambridge Medieval History III, stresses towards the end that the proliferation of castles was already a problem for Charles the Bald, but that it had a different value then, and I’d likewise note that my counts of Barcelona built castles all over everywhere and it was a security bonus and tool of control for them, until it wasn’t. And even at the third level, I think the point of my old Königsfern idea was that even a magnate who’d gone his or her own way (Matilda of Tuscany always there to remind us it was possible) still thought there should be a king somewhere, just not where they could affect said magnate. I would agree, however, that it’s all about buy-in; but perhaps the key we’re looking for in your first paragraph there is that by the tenth century, or even the late ninth, that buy-in had become conditional on performance in a way that doesn’t, as far as Gregory of Tours lets us see, seem to have affected the Merovingians. Now, I wonder if that was a possible result of Louis the Pious and even Charlemagne having had, perhaps because of their shaky ancestral claim, to codify the responsibilities of kingship, with their successors then unable to live up to what had now been set as the ‘industry standard’…

              • I think I’d agree with your explanation as to why they chose Odo. That’s the explanation Richer of Rheims opted for as well – Richer noted that Odo was a vigorous warrior (strenuus miles). I would also note that Odo probably had the most secure powerbase of any of the candidates – as Margrave of the Breton march, Count of Paris, Tours and Angers and Lay Abbot of more than half a dozen really ancient and prestigious monasteries in the Seine basin and the Loire valley, he really does seem to have been the premier magnate in the kingdom. The West Frankish nobility definitely didn’t want a weak king, and probably thought the realm would be in safest hands with Odo – perhaps they envisioned him as a “caretaker king.” Its tempting to see parallels between the election of Odo and that of Harold Godwinson in 1066, and one wonders whether Edgar Atheling would have ended up doing the same as Charles the Simple if Hastings had gone a different way, but maybe that’s pushing a parallel way too far. I’d also say I’m completely convinced by your explanation of why the counts of Flanders stayed out.

                Regarding the Janet Nelson piece, I have to say I really liked that chapter in the NCMH Volume III, and it really helped me get in the right mindset for thinking about tenth century kingship when I was preparing for my masters’ thesis. I’ve noticed that Nelson is very sceptical about the feudal revolution generally. She seems to make quite similar arguments to Dominique Barthelemy, not just on castles but on other things i.e. see her review in Speculum 69, Volume 1 (1994) to Thomas Head and Richard Landes (eds) “The Peace of God” (1992), where she argues that the “evil customs” of the early eleventh century are really no different to the abuses complained about in ninth century Carolingian capitularies. This is of course where someone like Charles West comes in – as he’d argue (and I’d agree) of course there were many Carolingian officials who took more than the state required from the peasantry, or simply took from them with no apparent justification at all, but never in a regular, formalised way like with the eleventh century exactions. She also argues that Duby, Fossier et al get most of their chronology of economic and social change wrong, and that a lot of the motors of local power were taking off in the ninth – I’d certainly agree that a gloomy, subsistence view of the Carolingian rural economy is unsustainable. Elsewhere, she argues for discontinuity between the Carolingian period and the High Middle Ages being overstated i.e. “Ninth Century Knighthood: The Evidence of Nithard” reprinted in “The Frankish World, 750 – 900” (1996), which is very much on the same wavelength as Barthelemy. I think her sympathies are with the anti-mutationist camp, mainly out of the spirit of rigorous scepticism and “can we find examples of phenomenon x happening earlier?” (both of which are good things), but unlike Dominque Barthelemy, Stephen D White, Richard E Barton etc she has a much more positive view of the Carolingian reform effort and doesn’t push the mutation documentaire. Incidentally, she also expresses quite strong scepticism about the maximalist view of the late Anglo-Saxon state in that chapter.

                For more about Charles the Bald and “castles”, here’s a recent article by Simon MacLean (https://research-repository.st-andrews.ac.uk/bitstream/handle/10023/20962/MacLean_TRHS_accepted_manuscript.pdf?sequence=1&isAllowed=y) that also gives a look at the origins of the castle from a Carolingianist standpoint – its really good, and gives a really thorough and contextualised reconsideration of the Edict of Pitres. And I think I broadly agree with his conclusions – yes, Carolingian aristocrats did have fortified residences, but like with vassalage and other stuff they really weren’t that important to how the Frankish aristocracy operated and saw themselves, and that seismic changes regarding the significance of castles really did take place in the tenth to twelfth centuries. I think something that also factors into this is changing perceptions of honour, which Rachel Stone discusses here https://magistraetmater.wordpress.com/2020/11/12/the-carolingian-preudomme/ – Carolingian aristocrats see honour as overwhelmingly coming vertically from the royal court and state service. Meanwhile, as Richard Kaeuper argues in “Medieval Chivalry” (2015), twelfth to fifteenth century warrior aristocrats saw honour as something that emanated from themselves and the approbation of their peers – service to the state could be honourable (as in the case of William Marshall, Geoffroi de Charny or Pero Nino in their different ways), but it was by no means essential.

                I also think I’m totally with you on that final point, about Charlemagne and Louis the Pious creating standards of kingship that would lead to their successors being held to account and deposed when they failed to live up to them. I think its possible to see parallels there with the West Saxons, the Salians and, even more so, the Angevins – by exalting and elevating the responsibilities of kingship, you also heighten standards by which the individual office holders can be judged, “the higher the pedestal, the harder the fall” as they say.

  14. Pingback: Seminars CCLIII-CCLVI: Friends and the Famous Speaking at Leeds | A Corner of Tenth-Century Europe

  15. Joseph Brown

    Here’s a longer exposition of my thoughts on the origins of castles, taking the form of a debate with myself. I basically run through all the possible hypotheses on their origins, and end up with the conclusion “look guys, its all really complicated, but monocausal explanations definitely aren’t gonna work.” https://carolingiansarecool.blogspot.com/2022/07/getting-glimpse-of-origins-of-castles.html

    • That is quite the round-up! I did attempt a reply on site, but it seems I’d have to have a Blogger ID to do it, and it looks as if Blogger no longer supports OpenID, so that’s not an option. If you don’t mind having the conversation here, therefore, I’d say first that I haven’t yet read Simon’s article, and you make it very clear that I must; anything I say here is obviously vulnerable to his already having refuted it. With that possibility open, therefore, I have always mainly favoured the aristocratic competition model of all of the ones you roll out, but with the proviso, based on my patch in Catalonia as much as what you say about Charles the Bald trying to defend against Vikings, that there are times in history when the top-level powers in medieval polities really rather wanted a proliferation of small local fortifications (and one could add Byzantine-then-Lombard Friuli to the count there as well as many more, I’m sure). The problem for them all was that those fortifications could become power centres to be used against them, so that the crucial question is maybe less what all these castles were for and more who got to decide that. The game these rulers had to win was, as ever (here comes my Innes training), remaining a better prospect for their local agents than simply embedding locally. Whereas no one local agent could really go rogue for very long successfully (though in some cases it might be worth doing it briefly by way of establishing a basis for renegotiation of his or her relationship with the ruler), a ruler who demanded too much and/or delivered too little might risk losing any consensus basis to act against such rebels. I suppose at its most brutal the logic of power involved is whether it’s better to have your people widely armed, to act against external threats, or thinly armed, to stop them acting against you…

      This begs a question that I think reveals something else, however. The question is: why therefore do we not have this proliferation of local fortifications already in the age of Alfred? Why instead are the wars of Wessex and the Viking Midlands fought from town to town? And one answer might be, because the Anglo-Saxon Chronicle wasn’t interested in telling us about anything smaller, but the Burghal Hidage too suggests that town-sized and -spaced forts really was where it was at in C9th-10th England. Of course, England’s urban fabric was basically missing compared to France’s at that point (I use both names carelessly, since neither England nor France had yet existed, but you know what I mean), which must make some difference, but I think that the other important difference is the absence of authority. I don’t mean by this that pre-unification England was some kind of ungoverned anarchy, but that it wasn’t clear to many people that the kings of the West Saxons had any say in much of the land where they were now asserting themselves over who built what with their labour. The extent of reluctance to comply even in Wessex with the royal fortification programme is clear in Asser’s Deeds of Alfred; how much less traction did he have in Mercia, where the Danish rulers might have seemed as legitimate and probably more interested successors to Mercian authority? Whereas in France, as you say, there might have been an argument about what Charles could demand, but there was relatively little argument, outside of Aquitaine anyway, that it was him who could demand it if anyone could. And what happens in all the cases where castles really proliferated, it seems to me, meaning Carolingian West Francia, Ottonian East Francia, Norman England, Italy at various phases and Catalonia in the C11th, is that a strong institutional basis for the labour provision that underpins being able to fortify remained in the power of what we may as well call ‘the state’ for a while, and then that ‘state’ lost the Innes game and the owners of those castles and their competitors had to turn to local resources and outdoing each other. That is, it’s not just that a strong institutional state could tolerate a lot of castles in some circumstances; I’m not even sure one could build them without access to such a set of institutions. I’m least sure that this applies in England, though the Anarchy would count, but what I’m getting at is that lots of castles might be a mark of a strong central power, rather than a failed one – until it suddenly wasn’t.

      Lesser points: the quote you want is actually, “It is often said that the spade cannot lie; but it owes this merit in chief to the fact that it cannot speak.” (Grierson, ‘Commerce in the Dark Ages’, first couple of pages). John Blair would argue with the chronology of Goltho and thinks the earliest fortified phase is actually Norman; without it, I don’t know if we have any plausible pre-Norman fortified homesteads, though there would still be Oxford and Dover castles as a sign of something else. Crop yields: our figures for the Roman period are pretty terrible but actually suggest rather better yields then, as far as it goes, than twelfth-century ones, though we are quite unlikely to be comparing like with like here. Fiscal inheritances: I completely agree that these were not just sub-Roman remnants, as anything claimed by, left to or even bought by the king could be added to it, meaning that while ancient and very long-term fiscal complexes are definitely not to be denied, the fisc as a whole was actually a good degree more fluid than we sometimes think, like a bottle of drink that had been in the freezer and now part defrosted, with its Roman and Merovingian core still floating around solid in the middle. Lastly, I twitch at calling ‘feudal’ anything other than actual exchange of property entitled ‘fief’ for military or other services, as you know, but I really think we should stop trying to make Japan fit the word, as its warrior caste structure really didn’t run like the Western one or with the same relationships to the state; see now Karl Friday, “The Futile Paradigm: In Quest of Feudalism in Early Medieval Japan” in History Compass Vol. 8 (Oxford 2010), pp. 179–96, DOI:&nbsp10.1111/j.1478-0542.2009.00664.x.

      (Numerous typo edits followed, sorry.)

      • Joseph Brown

        Thank you so much for the detailed, critical response.

        1. Simon’s article is definitely a really important read and I would highly recommend it to you. Simon is of course, as I’m sure you probably know much better than I do from what he’s written/ said elsewhere, reacting against the old school view of the Carolingian state, which sees it as both rigidly institutionalised but ultimately doomed to ossification and then failure in the face of a selfish, power-hungry aristocracy, as soon as the government machine runs into serious difficulties. Instead, he’s trying to suggest that actually the legal and institutional framework of the Carolingian state was much less formalised, and thus the idea that a state monopoly on fortifications that then got privatised by evil aristocrats doesn’t work. But nonetheless the Carolingian state could accomplish a lot of things through lots of heavy-lifting, creative propaganda and consensus-building and that ultimately what mattered was that the aristocracy really bought into the whole Carolingian reform project and saw it as a source of legitimacy for them as well as the monarchy. I’m not sure what he’d say about your view on incastellamento, as he never once mentions Catalonia and isn’t actually trying to explain why it happened – his main aim is to debunk the idea that the Edict of Pitres is the skeleton key to unlocking the history of the medieval castle.

        2. I agree with your points comparing Alfred the Great to Charles the Bald – I think you’re probably bang on the money there. I never did British 1 at Oxford (or British 2 for that matter, which made doing the Norman Conquest special subject an interesting experience, to say the least on many accounts) but I’m not a subscriber to the Alfredian revolution in government theory. Perhaps unconventionally for a Continentalist, I support the maximalist view of the Anglo-Saxon state but with all the important modifications, caveats and nuances that the post-Campbell generation have added. Therefore I’d ask, what about the unified English state under Edgar and his successors? That, we know, was able to command the labour of its free subjects to build roads, bridges and fortresses. And as for Goltho, I recall from personal conversation with John Blair that he said his position on Goltho had softened – that he thinks that there was a fortified phase from 1000 – 1050 but that it must have been inhabited by some kind of government agent of Aethelred Unraed/ Cnut. In which case, to speculate wildly, I might suggest that England may have conformed to your theory of incastellamento from a somewhat earlier date. But let’s not get too hasty.

        3. I’m broadly convinced by your theory of incastellamento as a deliberate government policy that then goes wrong when the centre enters a power vacuum, and I don’t think I explored it systematically enough in the post (I did mention Catalonia though). I think that theory of incastellamento definitely works for all the places you’ve mentioned, plus a few others – Flanders from the mid-tenth century up to 1127 is another good candidate. I think it works perfectly for England once we count the Anarchy. For Wales the second step was just unnecessary because that was essentially part of the government policy from the beginning until Edward I in effect tries to nationalise what had hitherto basically been a private baronial enterprise. I think in the German case does slightly differ from the others. Henry IV tries to make Saxony his territorial powerbase in 1066 – 1073, and builds a much denser network of castles than has ever existed in the Reich before. This process of incastellamento proves too oppressive (“evil customs” and the like) and the Saxon nobles and peasants join forces in rebellion and destroy many of the castles. Henry IV comes out on top by 1075, yet we all know what happens next and once the Salian theocracy is smashed and the German Reich has erupted into civil war, castles start proliferating like crazy. I also think this theory might have worked for the the Chateau de Lourdon. I think it was probably a Carolingian public fortress that the public power (Duke-Count William the Pious) gave to the monks of Cluny, who made it a private castle, before the general breakdown of public government took place in the Maconnais after 980 (assuming Duby and Bois’ chronology of social and political transformation there is correct).

        4. Thanks for the Grierson quote – absolute lifesaver there!

        5. In terms of the crop yields thing, what is the ancient Roman evidence actually like and from what bits of the Roman world does it actually come from? And when would you say the best Roman yields were finally exceeded?

        6. Thanks also for the Karl Friday paper reference. And I’d probably agree with you about the use of the term feudal. On a somewhat unrelated note, I’m really looking forward to what you have to say about Chris’ latest article on the “Feudal Economy.” I have to say that I couldn’t help but be impressed by the theoretical coherence and Chris’ inimitable grasp of the historiography of so many different regions (especially since the bulk of the article concerned periods much later than those he normally works on), but I didn’t like where he was going with it either. I was especially disappointed to see that he’d abandoned the whole idea of a transition from antiquity to feudalism in favour of the idea that basically every complex agrarian society with a state system and a class-based mode of surplus extraction is essentially feudal – for me, that left a rather nasty taste in the mouth.

        • Thanks as ever for the feedback; I wasn’t meaning to develop a wholesale theory, but let’s see how it goes. On 1, yes, I saw it come through my door more or less literally with the relevant issue of Transactions and shamefully just haven’t made the time because I haven’t taught my Carolingians module or had to write anything about castles for ages. But when I get back to my military service article some day (currently only two things in the queue ahead of it but one is a monograph…), I will.

          On 2 and 3, on looking back I could have been clearer. I wasn’t meaning to say that agency for a proliferation of castles necessarily always lay with the top-level power. It seems pretty clear that lots of local aristocrats wanted an impregnable base of operations and that many kings and rulers saw that as a problem they would like to prevent, albeit sometimes only after it had happened. But I do think that (a) sometimes the interests of aristocrats and their superiors could coincide in wanting that proliferation to happen, and (b) that it probably couldn’t (or wasn’t needed) where no institutional apparatus existed for the labour extraction. You need a pretty rigid authority structure to enforce that.

          Therefore, on 2, I think with what I now think of as Molyneausian England, we have a situation where that authority structure had been created, by a sequence of idea-to-realisation running from Alfred to Edgar, but one also where most of that time had been spent destroying enemy fortresses of local rulers and setting up governmental ones which can’t easily be monopolised by a single lord. Edgar and successors knew the danger that proliferation of small fortresses could present for them (roughly the ones that Æthelred II ran into when his writ stopped running smoothly), and had no need of them anyway. That created defensibility problems for England in the Second Viking Age, as it turned out, and then the pressures of the Norman Conquest and experience with a different governmental model made William I throw England into the proliferation model anyway when he took over. And eventually Henry II wound up with castles somewhere like Ramon Berenguer I did with fiefs: they’re no longer public, OK, but I do darn well control them nonetheless. And that compromise proved just about viable till I suppose the English Civil War, except for the revolt against John, the Barons’ War, the Percys’ various risings, the Wars of the Roses…

          On 3, I was thinking of Saxony as somewhere that did fit the model and this makes me wonder if the mistake here is to try to take Germany as a unit, rather than as a set of separate or separating polities for historical reasons sharing a king to very uneven effect. In other words, Saxony might fit but Bavaria, where I have the impression something more old-fashionedly Frankish with big urban fortifications and not many small ones was going on, would probably not, and I don’t know how many other variations there might be. But if we’re seeking to explain a proliferation of small castles, and somewhere didn’t have one, that doesn’t of itself require that our theory account for it!

          4, no problem; I love that article and have often used the quote. On 5, it’s lousy, basically anecdotal from Cicero’s letters and Columella’s treatise. There are actual documentary figures from Egypt, but few and hard to interpret and anywhere nowhere else farmed like Egypt; there are also some Byzantine figures from Palestine but it’s still wet agriculture versus dry so not really comparable. As with anything about crop yields, everything I know, including those cites, is in my “Outgrowing the Dark Ages: Agrarian Productivity in Carolingian Europe Re-Evaluated” in Agricultural History Review Vol. 67 (Reading 2019), pp. 1–28, which is now technically beyond the journal’s three-year moving wall so should be online for free any day now.

          Lastly 6, yes, I will be writing about that. Up till this point I haven’t read that article either but I do have a better excuse this time, being that I saw him deliver it as a Hobsbawm Lecture at Birkbeck so do know roughly what he’s arguing. I have to say that I quite like it; following John Haldon, I was already halfway to seeing ‘feudal’ as a meaningless modal category anyway and classing everything in that general bracket as ‘tributary’, but was struggling to articulate the difference between ancient tribute and modern tax, and for me Chris has untied that knot by making that broad ‘tributary’ occupy the name ‘feudal’. I still won’t use the word if I can get away with it but I find Chris’s take coherent. But perhaps we should save that conversation for a post which is actually about it!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.