Plz be respecting feudalizm: further opinions from Chris Wickham

I figure in these hard times Geoffrey Chaucer needs all the promotion he can get, you know? Though he seems, despite his absence from the Internet, to be writing in to the Guardian, or at least his agent is. But, you will be shocked to hear, this wasn’t the whole point of my post.

The 1999 Spoleto conference on feudalism that I mentioned before had its keynote speech given by Chris Wickham, and he set out to rehabilitate the word and its meanings. If you’ve been through a course in medieval history that went before 1200, you are probably aware that in 1974 Elizabeth A. R. Brown went thoroughly into the uses of the word ‘feudalism’ to describe the society of the high Middle Ages in an article in the American Historical Review, found them basically lacking because of lack of consensus about what it meant and lack of basis for the various explanations that there were, and asked all historians to please stop teaching it as soon as possible.1 Not satisfied with the results, Susan Reynolds in 1994 wrote a thick book, about ten times the length of Brown’s article, in which she railed at considerable length at how silly all the people were who were still using this broken and misleading concept and requested them at length to acquire clues and similar.2 Reynolds’s book was sufficiently vituperative and powerfully-phrased, and her career sufficiently autumnal, that while those who loved controversy lapped it up, those who had no intention of changing were able to more or less dismiss it as the fulminations of a crank. This was unfair, because it says important things, but it doesn’t, admittedly, leave one very clear on what Reynolds actually thinks did organise medieval society, at least in part because she’d written that at length elsewhere anyway.3 Reynolds’s work can’t be ignored, but nonetheless, because it’s older and therefore better cited, because it’s (slightly) less polemical, because it’s of US origin and therefore attracts itself to more people who are aware of the historical community on the Internet perhaps, and mostly I suspect because it’s shorter, Brown’s article remains a bigger influence on people than Reynolds’s book. Another Damned Medievalist has written about this, and so I’ll refer you there and get on with the point. If you do want more, the Reynolds debate is one of the rich patches in the Internet Medieval Sourcebook and Melissa Snell’s blog on tackled the whole issue not so very long ago too, so that should be all the HTML any reasonable person can eat and beyond that you’ll actually have to read some print.

Roland receiving the sword Durandal from Charlemagne in exchange for homage, from Wikimedia Commons

Roland receiving the sword Durandal from Charlemagne in exchange for homage, from Wikimedia Commons

Anyway, the fine art of leaving Susan Reynolds by herself and arguing with the (easier) Brown was on good form in Chris’s hands in 1999.4 He carefully stays more or less clear of Fiefs and Vassals, arguing that he doesn’t have time to do a detailed critique of the sources, and instead goes for the theoretical throat, which is Brown. Yes, he admits, ‘feudalism’ is an ideal type, and thus probably didn’t really exist anywhere as we imagine it. But we can’t get very far without ideal types, he argues, because, lacking 100% information, we have to construct in order to imagine what our subject population was really like and doing. Ideal types are a thing that powers the imagination, and powers discussion about the respects that they don’t match. What he’s essentially arguing here is that we need, if we’re to understand medieval society to our own contentment, to be able to work not in terms of slow assembly, but in terms of bold-but-wrong thesis, careful and destructive antithesis, and synthesis, repeated each step closer to the objective reality and so on, you know, the way we actually do history.5 None of us do history without any basis in the area at all, from first principles, so the real issue with ideal types is whether one can identify the ones one uses or not! Do you know why you think what you think, can you argue it? and so on.

Having thus cleared some ground in which a concept of feudalism might stand, Chris argues that in fact there are three, and that really most of the trouble has come from people confusing them and using one’s inapplicability to disprove another. He separates them out as follows:

  1. Marxist feudalism (because this is Chris after all): the feudal mode of production in which peasants own and operate the means of production but lords control it. The advantage of this is that it brings the 80% of medieval society who didn’t write or ride chargers into the picture and reminds us that it all comes down to who feeds whom at some level; the downside is that it really doesn’t cover much else and frequently doesn’t describe how people are actually related when for many other reasons we’d want to call their society feudal. For example, a Catalan free peasant paying rent to a lord but otherwise independent in the 1030s is in a society where feudal oaths are the organisational basis of pretty much all power relations; but for a Marxist he isn’t himself embedded in the feudal mode, so that breaks quite easily when you try and make it do more than it can cover.
  2. Exactly opposite, the juridical sense which says that really, the only thing you can call feudal is relations centred around fiefs, that is what others might usually call ‘feudo-vassalic relations’ so as to separate it cleanly from the whole big shebang; a lord and a follower in a relation of mutual benefit where the follower gets to hold land from the lord ‘in fief’ or ‘in fee’ in exchange for service and loyalty. The problem with this is self-evidently the opposite one, that it misses out the peasants and anything not to do with the military organisation of society, and thus misses really rather a lot, though many would argue that the rest isn’t ‘feudal’, thus setting themselves at complete cross-purposes with the Marxists…
  3. Finally, the Marc Bloch version, what Georges Duby called, translated from the French, ‘the feudal imaginary’. This is the big-picture version of Poly & Bournazel’s Feudal Transformation and also of course of Bloch’s own Feudal Society, in which everything is built up from the juridical version to extend through the whole of medieval society, its literary production, its self-image, its self-justifications and legends, its economy and culture, its religion (Jesus as Lord and lord) and the whole lot really.6 The obvious problem with this is that it is, again, an ideal type; I’ve argued elsewhere and many have before me too that this just never happened all at once, and if this is the real definition nowhere was ever ‘feudal’. Chris, having already defended ideal types, says that that doesn’t matter because it gives a big picture view that encompasses both (1) and (2) and leaves you feeling as if you can see the society better. We know it’s wrong, in detail, but it’s helpful, even if by now mainly to critique.

I am usually persuaded by Chris to an extent, except mainly where he justifies an argument with something like, “because aristocrats/peasants are like that”. (And indeed, one of the great beauties of the Settimane series is that they also publish some version of the discussion, so you get Chris having a very high-level exchange of compliments with Dominique Barthélemy that includes him saying, “As you know well, some historians cannot argue their way out of a paper bag” in print where it can be found and cited (p. 50), which amuses me.) So here I think he quite rightly distinguishes between some concepts which have been wrongly identified together, and makes an entertaining and persuasive pitch for continuing to talk about feudalism, which can basically be summarised as ‘we already are, so we may as well admit it’. And I do use the word freely in a way that would cause some of my colleagues to start uneasily, mainly because in my area I can point you at upwards of a hundred feudal oaths and contracts which are, among other things, almost the earliest written examples of a Romance language but that also make a pretty good case for admitting that in Catalonia there was a feudal system, even if it didn’t necessarily reach everywhere and include everything.7 But, in the end, the basic problem is this: it being a good thing to think with doesn’t alter the fact that it makes you think in terms of one thing (or here, three that are one; Chris draws the Trinitarian analogy himself in the discussion). And what was going on wasn’t one thing, but lots all at once. So although Chris may not be, I remain confused and unconvinced, but now by Brown and him alike.

1. Elizabeth A. R. Brown, “The Tyranny of a Construct: Feudalism and Historians of Medieval Europe” in American Historical Review Vol. 79 (Washington DC 1974), pp. 1063-1088, repr. in Lester K. Little & Barbara H. Rosenwein (eds), Debating the Middle Ages: issues and readings (Oxford 1998), pp. 148-169.

2. Susan Reynolds, Fiefs and Vassals: the medieval evidence reinterpreted (Oxford 1994).

3. Eadem, Kings and Communities in Western Europe, 900-1300 (Oxford 1984, 2nd edn. 1997).

4. Chris Wickham, “Le forme del feudalesimo” 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. 15-46 with discussion pp. 47-51.

5. That is, assuming that you accept that there was an objective reality, which I do even if everyone in it perceived it differently. Let’s not have that argument again just now.

6. Referring to Marc Bloch, La Société Féodale (Paris 1949), transl. L. A. Manyon as Feudal Society (Chicago 1961); to Georges Duby, Les Trois Ordres, ou, l’imaginaire du féodalisme (Paris 1980), transl. Arthur Goldhammer as The Three Orders: feudal society imagined (Chicago 1980); and to Jean-Pierre Poly & Eric Bournazel, La Mutation Féodale, Xe-XIIe siècles, transl. Caroline Higgit as The Feudal Transformation, 900-1200 (New York City 1983), 2nd edn. in French 1991.

7. Now covered at length in English by Adam J. Kosto, Making Agreements in Medieval Catalonia: power, order and the written word, 1000-1200, Cambridge Studies in Medieval Life and Thought 4th series 51 (Cambridge 2001), which does not entirely replace Michel Zimmermann, “Aux origines de Catalogne féodale : les serments non datés du règne de Ramon Berenguer Ier” in Jaume Portella i Comas (ed.), La Formació i Expansió del Feudalisme Català: actes del col·loqui organitzat pel Col·legi Universitari de Girona (8-11 de gener de 1985). Homenatge a Santiago Sobrequés i Vidal, Estudi General: revista del Col·legi Universitari de Girona, Universitat Autònoma de Barcelona nos. 5-6 (Girona 1986), pp. 109-151, with Catalan summary p. 523, Castilian summary p. 531 & English summary p. 557. On the linguistic importance, you can see Joan Bastardas, “El català vers l’any 1000” in Immaculada Ollich i Castanyer (ed.), Actes del Congrés Internacional Gerbert d’Orlhac i el seu Temps: Catalunya i Europa a la Fi del 1r Mil·lenni, Vic-Ripoll, 10-13 de Novembre de 1999 (Vic 1999), pp. 495-513 with Catalan résumé p. 495 & French & Provencal résumés & English abstract p. 514.

18 responses to “Plz be respecting feudalizm: further opinions from Chris Wickham

  1. What serendipity. I forgot I’d written that, and dammit, I’m teaching that lesson today! Thanks — you’ve just helped me focus what it already shaping up to be a day of hellish unpreparedness — 5 hours of lecture, and an hour of meetings, and I feel barely prepped.

  2. Oh — and Reynolds and Brown are right, Wickham is wrong. Except that Reynolds and Brown are not entirely right. Because Bloch and Ganshof aren’t entirely wrong. I’m all for the “for the sake of convenience, and with many caveats, here is how we can use the damned word — it’s very limited, and only relates to the general types of mutual obligation we can see in charters like these.”

    We shall not speak of Japan.

  3. This is of course the simple answer to how I have time to write this stuff, is I’m not teaching most of the livelong day. This year, anyway. Happy to have been what help I’ve been, although faintly worried to be your before-class reading…

    As for the argument, I’m not sure Chris is wrong, but I’m not sure he’s helpful, and the same could be said for Brown. I’m sure Susan’s right, but only in terms of what other people think and what words are safe to use. Ah dear: if a lord and a knight agree a grant of land in return for military service, and no-one’s there to write it down, does it still make a feudo-vassalic agreement? etc.

  4. I think so — writing down is good, but mutual recognition and custom are fine. After all, do we know that all those chaps in Annales Fuldensis who switch sides between Louis and Charles had written contracts? And would it have mattered if they had?

    Fortunately, MA isn’t till I’ve already taught 4 other classes and gone to a faculty meeting. 3 down, 1 to go. Sanity? Long gone.

  5. Oh – and Wickham — it’s not an ideal, it’s a model by another name. I don’t trust models; they reek of social science, and I rest in the humanities, dammmit!

  6. Ah, no, Chris has convinced me of that much, not just here but in other work; we need models, to take from areas we’re certain about to other places where we haven’t the evidence and see if it might explain things. And as I say, we all have some kind of model in our heads of how it was, don’t we? The question, and it will be taken up in another post I’m about to write, is whether this particular model is a help or a hindrance.

    Carolingian military loyalty and personal bonds is something that someone else beyond Barthélemy should look at, though he has a really good paper on it in Le Jan’s Royauté et les Élites. I think only he would try and argue that the guys at Strasbourg are functionally the same as Duby’s Mâonnais milites. The technology of war has changed and the prospects of employment as a warrior are much better a century later, and I think that’s important. But as to the importance of writing, I actually agree with you, it’s not important. But I’m not sure that the oath and the kneeling are either, you know? Somewhere in my thesis I suggest that the difference between me and another pair of historians working on Spanish power relations is that, if that dialogue were being conducted by telephone, they’d be looking for the line and the installers, whereas I’d be trying to reconstruct the conversation. Here too, the form is less important than the content, and that’s unhelpfully variable if what you want is a Europe-wide explanatory system. If you can accept that there isn’t one, of course…

  7. I think they are equally important and unimportant in the same way. And, like Reynolds, I’m obviously leery of models. I absolutely agree that we compare based on what we find in one area, and then may extend our inferences to a general picture, with local variants. But the ‘model’ or ‘ideal’ seem to me to be taking it all a step further than where we need to go. I’m ok with the fuzzy grey areas. But for me, building a model implies that there is a model that can be built — and from which we then will, whether we want to or not, end up comparing to the model, rather than to other realities, if you will.

    Either that, or it’s all semantics, but I don’t think so.

  8. I think the form only matters if it itself has a social effect, and I’m not very sure that it does. I realise that homage is a very solemn and splendid thing hung about with sacral importance, but like coronation, it doesn’t prevent it being overturned at times. So the fact of there being a homage ceremony is to me always less important than what the result of that or its background may turn out to be. I agree with you about models here though; I think it’s just a matter of quite how far along the comparative line we’re respectively prepared to sit. I think Susan’s great and her work very necessary but I think I’ve got more use, myself, out of Chris’s ideas. I guess you might say the opposite?

  9. As an old teacher of mine said, before either Brown or Reynolds had written, what makes an oath a “feudal” oath?

    The word “feudal” has its uses, but it is perfectly possible to teach medieval history for 25+ years without using the word “feudalism.” I decided after I read the Brown article that all the time I would otherwise spend explaining what exactly I meant by the word could be devoted to discussing things that happened in the Middle Ages.

    I think “feudalism,” or at least the “feudal regime” is what the National Assembly abolished in August, 1789. It had a lot to do with dovecotes.

  10. As to your question, I think the only answer is: one dealing with a fief. But then we get into the question of whether such-and-such a benefice given for service is a fief, and whether when they use the term feodum in such-and-such a document we’re yet talking fiefs or if it still only means ‘fisc’… and you could do the same argument again two centuries before about fiscs, argh. So your abstinence sets an example to us all, sir. Which leads me to ask, can I borrow that last paragraph for the next time I have to teach this?

  11. Borrow away! Showing a few pro-feudalism types the August decree and having them relate it to what they study might be fun.

  12. Carolingian military loyalty and personal bonds is something that someone else beyond Barthélemy should look at

    Shameless self-promotion, and it’s only on the ideology, not the practise, but the recent Jinty Nelson Festschrift includes my paper comparing Frankish and Anglo-Saxon ideologies of lordship.

    As for Carolingian practice of lordship, if there are substantial sources out there I’ve never found them (and nor have Ganshof, Matthew Innes or Guy Halsall, who’ve all written stuff on military service).

    I think Chris’ embrace of ideal types was originally used usefully by him to look at questions of continuity and change, and when the feudal revolution/mutation/something might have taken place.., by asking: Does the situation in 800 or 900 or 1000 look more like society type A or type B? (And he couldn’t have done ‘Framing the Early Middle Ages’ without using ideal types, to show what he’s trying to compare with what).

    But I don’t see how it’s useful to combine two ideal types (Marxist feudalism and lord-man relationships) into one, as he seems to be doing here, because that blurs the question of whether, like love and marriage, you can have one without the other. If you can, it’s not analytically helpful to combine the two concepts. (Though I seem to remember Trish Skinner in the same Spoleto volume worries whether you can separate the two concepts out nicely in the first place.)

  13. The Skinner paper is a real breath of fresh air, and will get some coverage before long. I was very glad to reach it after an awful lot of over-analytical German. Similarly, I shall be interested to see that paper and indeed the rest of the Festschrift when time permits me. I think Matthew’s come a step closer, in his own particular way, with his new paper in McCormick & Davis’s The Long Morning of Medieval Europe, which I would recommend especially as it deals with Hincmar (unless I’m thinking of his TRHS paper instead).

    As to Chris, well, I think the point of his paper was to separate the two types, but I agree that the Bloch type does rather combine them once more. At worst, though, that’s not Chris’s fault. I think the weak point of his argument is certainly there though, that the Bloch picture still has value even as something to disagree with. He may find it so, but many of us won’t any more. It’s quite picturesque of course…

  14. Trying to think back to what I’ve heard Chris say about ideal types, I think one of the other problems is that he’s argued that you really need ideal types in pairs. So you can say, does this look more like ideal type A or more like ideal type B? (Rather like opticians do in eye tests: Is the feudal model clearer or is the Roman model clearer?). This works OK for Marxist feudalism where you have the alternative of other modes of production. It works OK for feudal-vassalic relations, because you have the warband on one side and the salaried soldier on the other. But what is the alternative to the feudal society?

    If you look closely, the alternative tends to be Carolingian ideology, not actual Carolingian society. So you’re comparing an actual society that you’ve simplified (normally the Maconnaise) with a society that never existed in the first place, which is daft. You can usefully say whether something looks more like a cow or a horse because those are real things, even if you’re working with a simplified mental image of them. But how can you usefully say whether something looks more like a horse or a unicorn? It depends how you imagined a unicorn in the first place.

  15. But feudal society as object doesn’t exist either; it’s an ideal type same as the Carolingian one. But what is the Carolingian ideal? Public justice, a top-down-imposed authority structure, office, appeal to the king… ? Bloch gave us a picture of ‘feudal’ society; but the corresponding ‘ideal before’ has never had its Bloch (although between them Jinty and Rosamond make quite a lot of one, and Louis Halphen and Ferdinand Lot between them were at least half of one (the French half… )). And I’m not sure that’s a bad thing because that ideal would be no use either. But the fact that we find it so hard to come up with one ought to lead us to question the one that Bloch did build all the more.

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

  17. Pingback: In praise of Marc Bloch’s Feudal Society | A Corner of Tenth-Century Europe

  18. Pingback: Seminar CCLXXI: feudalism beats capitalism for most of history | A Corner of Tenth-Century Europe

Leave a Reply

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

You are commenting using your 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.