Different sorts of rulers on the edges

View over the Universitat de Girona

View over the Universitat de Girona taken earlier today by your author

Hullo again! I actually write this from a hotel in Girona, where I was kindly invited to give a guest lecture, because when I would ordinarily write the week’s post I’ll be travelling back by the dubious offices of Ryanair, so things are going on to which, let’s be optimistic, I will some day soon catch up. Right now, however, the post I promised you was about the culmination, for now, in 2018, of my network project for ‘rethinking the medieval frontier’. Now, I was more or less set last week for the need this week to write up a report on that conference, and then while writing last week’s, I was reminded as I linked to the project blog that I actually already did so, there rather than here, within literal months of the conference actually happening. So the first point of this post is to point you at that account, which is here:

Report on 1st Conference

There are photos and everything, and also links to others’ reports should you (rightly) think that something I put on a project blog might seek to emphasise the positive out of all proportion. But what, of course, that post has fairly little of, except in phrasing, is me, and what, as I have often said is the point of a blog except to give the Internet more of yourself? So secondarily in this post I want to talk a bit more about where my paper came from, where it was and is intended to lead, and why, in fact, I was even reading up on ‘Abd al-Rahmān ibn Marwān al-Ŷillīqī. Since that is kind of gratuitous, though hopefully interesting, I’ll stick it behind a cut even though it’s not very long, and encourage you to go read about other interesting people and their thoughts via that link first.

But OK then, you’re apparently in for the rest; thankyou! So then. The point of my frontiers project has always been comparison. I know medieval Iberia best, but I know it better for occasionally looking at somewhere else and realising what is and isn’t distinctive about it. When in the throes of a project whose grand aim is actually to come up with transportable generalisations about frontier situations, of course, what is generalisable might be more important than what’s not, but the effort of comparison still has to be made. Also, you may remember from my write-up of the project’s first workshop that there was some scepticism among my collaborators as to whether we medievalists could really come up with capital-t Theory on frontiers, rather than just models for people to use. Now, I’m not sure those things need be different, but at any rate I thought that if I was going to give myself the last paper—which I’m afraid I did—then I had at least to make some kind of effort to show what it is that I think we can do with these opportunities for comparison and testing with each other.

Castell de Llordà, Isona, Catalunya

The current state of the Castell de Llordà (image from Viquipèdia), populated for Borrell II by a monastic subsidiary

So! Three years before, I had taken my first step in this direction with a paper at the Leeds International Medieval Congress, before I even worked at Leeds, rather pompously called ‘De Administrandis Marcis: the 10th-Century frontier with Islam, seen from Barcelona and Byzantium’. The point of this was expressly to compare the frontier management I know best of all, that of Borrell II of Barcelona (about which I am now learning more still—more on that in due course as well) with another near-contemporary about whom I had also then been learning, Emperor Constantine VII (some of which also made it here). While stressing the many differences between the two rulers, especially of relative scale of power and resource, I thought that there were still similarities, firstly in the way that ideologically they reached back to semi-imaginary pasts with which no-one could effectively argue to justify their present position and policies—though perhaps every smart medieval person did that—but secondly in the way that when they had to deal with zones at the edges of their control, both of them usually preferred to find a trustworthy subordinate and then give him an absolutely free hand, even if it meant that that man—so far always a man, but I keep looking—became an unexpectedly major player. For Borrell this was people like Ennegó Bonfill son of Sendred, whom I wrote about long ago and who got so powerful so quickly that later sources assumed he must be a count, and for Constantine it was Melias, tourmarchon of Melitene. Now, this is actually medieval difference talking to us. For many a ninteenth-century historian and a good few twentieth-century ones, one of the things that medieval, and especially early medieval rulers, did wrong was to hand out too much power when in difficulty, making people mighty whom they subsequently could not control.1 I don’t mean to say that this is always wrong, but the danger must have been as obvious to those medieval rulers as it is to us, and they still made their choices against what ‘we’ think hindsight should have advised. I used this paper to suggest that the alternative was worse, because somebody who could be fully controlled would not be as effective as someone who could, if they chose, be independent; the trick for the central ruler was therefore to keep them from needing or wanting to take that independence. There is probably more I could say on the basis of what work I’ve already got out on this subject about why that might be, but that was the point I made.2 And then the next year I gave a paper about the Andalusi warlord family of the Banū Qāsī, which took up some of the same points from the point of view of the frontiersmen, stressing how they were able repeatedly to escape central control but actually did better when they were able to get what they wanted in cooperation with it.

So at the end of this I had, not a theory, but a component structure within some future theory. If I were to express that now, it would be something like: (1) the medieval frontier was controllable principally by local domination, whether because of its simple proximity or because of roots and connexions on which it was able to draw which gave its authority traction that an outside could not maintain. (2) Because of this, it was to a centre’s advantage to recruit such locally powerful persons, and it was worth giving what they wanted in the short-term, even at the cost of subsequent difficulty in controlling them, to advance central control in the moment. (3) In the other direction, the centre was usually able to offer these lords something in terms of support, legitimacy, or range of available resources, that made it better to work with the centre than without it. For such frontier lords, it was necessary for there to be a centre that could recognise and enhance their positions, and they would often seek its attention out when it did not provide that recognition and enhancement. (4) But to maintain their importance, they also had to maintain their position as local power-brokers, this often being reliant on their frontier position and privileged ability to deal with those across the notional border compared to less well-rooted plants of the centre. If they lost that, they also lost any privilege of their position. And of course, if this project is to succeed, pretty much any such synthesis I reach will have to be followed with a loud stagey (5), “PERHAPS THIS IS ALSO TRUE IN OTHER TIMES AND PLACES…”, and I will then have to wait to receive reactions from the metaphorical crowd.

Anyway, I carried this idea into the preparation for my own 2018 conference, but by then I had met some examples of a different form of frontier-centre relationship, and foremost among these was yer boy al-Ŷillīqī. His position seemed to me to be inverse to that of the Banū Qāsī. They got to be powerful by representing al-Andalus in areas it could not otherwise fully control, something they sometimes demonstrated by withdrawing their obedience; but they were at their most powerful in cooperation, not in opposition. Al-Ŷillīqī, on the other hand, drew his strength from opposition; if he had rejoined the régime he might have got governance of a city somewhere but strictly on the centre’s terms, whereas as it was, his frontier location and the weakness of the centre there allowed him to achieve an effectively impregnable position by threatening the state and offering it, as an incentive to give him what he wanted, not his obedience but a simple agreement not to attack it. Both of them were parasites on the Andalusi state in one way or another, but at least the Banū Qāsī were in some way symbiotic; in this body-type metaphor, al-Ŷillīqī was more like a persistent itch which one mustn’t scratch in case it gets worse. Furthermore, all of my examples in the first two papers were people who had been raised to their positions by the centres with whom they had these relationships: they had some local position, Sant Sadurni de Tavèrnoles as a mountain monastery with farflung properties, Melias of Symposion as governor of Lykandos and Mūsā ibn Mūsā as castellan of Arnedo, but their big frontier breaks came from commissions from their respective states. Al-Ŷillīqī just chose his own spot, whether the centre liked it or not, and they got no effective say in the matter. So I wondered if there were more frontiersmen like this, whose strength was in their ability to use their location against their relevant centre rather than for it. And obviously I thought there were, but my third example surprised even me when I first thought of him.

Muş, Turkey

Muş, Turkey, once the chief town of Taron, image retrieved in 2018 from the now-dead site

So in ‘Our Man on the March: three frontier lords and their geopolitical positions’, my chosen lords obviously included al-Ŷillīqī; he was my type case. He wasn’t the first one I discussed, however; that was in fact another set of Caucasian lords from Constantine VII’s time, the lords of Taron, who had had a minor civil war between themselves in the time of Constantine’s dominant co-emperor Romanos I, which Romanos had been able to use to take an entire province off them in exchange for one he could less easily control. It was good state-craft on Romanos’s part, as Constantine grudgingly admits, but it was enabled entirely by the Taron family’s willingness to engage with him for status; he had no control in their lands which they didn’t give him, only control in Constantinople over stuff they came to care about.3 And then my third example was Saint Columba.

Now, OK, you may be looking at that and thinking, “you were stuck, weren’t you, Jonathan? That is the play of a man with no good cards in his hand.” But consider. Columba was Irish nobility in exile. He set up in an island which may have been in the ‘Scottish’ territory of Dá Riata, or may have been in Pictland; even to relative contemporaries this was apparently not clear. (And we already know how special islands can be in frontier terms, amirite.) From there he visited, advised and threatened—threatened quite a lot, actually—kings across most of the north of the two islands of Britain, and part of his strength indubitably came from being in no-one’s territory. I’m sure that if they had ever wanted that kind of bad luck the Dál Riata could have raided Iona, and given his apparent foresight I imagine Columba would not have been there, but in any case they didn’t. But he was also stationed on a frontier of a different kind, with the next or other world, and unlike most normal people he seemed to his contemporaries to be able to see over it and speak from the other side. Now, there’s no question theologically of Columba’s various royal and non-royal clients being able to exert power over Heaven through him; the best the Life of Columba lets you assert is that they were able to get some information on what Heaven had planned for them. In this power relationship, Columba had the upper hand, and though he might not himself be able to go back to Ireland, his monks rapidly became powers in the land there in several places, almost precisely because their patron (and of course his Patron) were out of Irish reach.4

St Ronan's Bay, Iona

You have to admit, it is a good location: St Ronan’s Bay, Iona, by Rich Tea, licensed for reuse under Creative Commons via

Now, I’m not, really, saying that Columba was to the Uí Neill and Dál Riata (or indeed Picts; actually, I maybe am for the Picts…) what al-Ŷillīqī was to Emir Muhammad I, but I am saying that their respective positions of power derived from being able to use their location as security against the centres between which they sat, rather than from using that location to those centres’ advantage as had my previous exemplary frontiersmen. It’s another part of the model, I think. Obviously this is going to be a complex model; but if it was simple, someone would hopefully have come up with it already! As it is, I may be collecting a few more parts yet, and you can probably expect to hear about those developments here.

1. For example, beautifully epitomising those nineteenth-century clichés even in he mid-twentieth, see C. Previté-Orton, The Shorter Cambridge Medieval History Volume I: the later Roman Empire to the twelfth century (Cambridge 1952), pp. 334-335.

2. Jonathan Jarrett, “Engaging Élites: Counts, Capital and Frontier Communities in the Ninth and Tenth Centuries, in Catalonia and Elsewhere” in Networks and Neighbours Vol. 2 (Binghamton NY 2018 for 2014), pp. 202–230, online here as of 12 April 2019.

3. My key sources here were Constantine Porphyrogenitus, De Administrando Imperii, ed. Gyula Moravcsik & transl. Romilly J. H. Jenkins, rev. edn. (London 1962), repr. as Dumbarton Oaks Texts 1 (Washington DC 1967, repr. 1993 and 2008), c. 43, and Stephen Runciman, ‘Cc. 43–46/165’, in Jenkins (ed.), Constantine Porphyrogenitus, De Administrando Imperii: a commentary (London 1962), repr. as Dumbarton Oaks Texts 4 (Washington DC 2012), pp. 156–80, but for something more up-to-date on the literally Byzantine politics of this region at the time, see now T. W. Greenwood, “Armenian Neighbours (600-1045)” in Jonathan Shepard (ed.), The Cambridge History of the Byzantine Empire c. 500–1492 (Cambridge 2009), pp. 333–364.

4. Here both source and up-to-date interpretation were from a Penguin, I admit, but no ordinary Penguin, a Penguin by the late great Richard Sharpe, to wit, Adomnán of Iona, Life of St Columba, transl. Richard Sharpe, Penguin Classics (Harmondsworth 1995). If, however, you wanted a background read not made entirely of endnotes, Tim Clarkson, Columba (Edinburgh 2012), is hard to beat. Specifically on the Iona family of monasteries, including the Irish houses of Kells and Durrow, see Máire Herbert, Iona, Kells, and Derry: The History and Hagiography of the Monastic Familia of Columba (Oxford 1988).

9 responses to “Different sorts of rulers on the edges

  1. I am not sure that a model based on center-peripheral powers can successfully fully explain what was going on then. I use to think about those topics more as center and periphery being both more like clusters of powers…?
    P.S. Good to know you had the chance to see the country again!

    • Yes, only briefly and passing through Barcelona only for transport, or I would have been in touch!

      In general, meanwhile, I agree: a frontier zone can very easily be its own centre, and I suppose here Iona might most easily be the example, a new centre created by the saint and his work. In the other cases, however, I do think that the point I was making was that their positions on the edge of a bigger polity was what gave these lords their particular power; they did not negotiate or exchange as equals and they recognised, sometimes, the legitimacy of another authority in their own areas. I don’t mean the deny your model in general, but I think there really was a centre-periphery in these cases, being used wittingly by both parties in them.

  2. Allan McKinley

    Fascinating piece, which elicits a fair bit of reaction from me. Although I must apologise that my first point is the classical postmodern observation that you may not really be talking about Columba but Adomnán who created the image. We actually have historical evidence for Adomnán using his frontier status as well: see Cain Adomnáin (sic?) and the release of the Irish captives in Northumbria, so there’s a good case for this. It’s probably somehow relevant that by Adomnán’s (I’m happy that I can spell the English genitive anyway) time his branch of Cenéll Conaill seem to have been been definitely excluded from the kingship.

    In terms of the model here, is there any contradiction between the symbiotic and itchy (and if you publish the final model please keep that distinction) modes of engagement? Both at root are based in the potential of a someone in the frontier zone to use their power, and look to be differing manifestations of the same underlying situation: that pretty much by definition a frontier was where the centre(s) had the most difficulty exercising power directly.

    As for applying the model, I think it potentially works for two of my favourite families/groups: the dukes of Alsace and the (sub-)kings of the Hwicce. Both are visible as both independent actors on the edge of a larger polity, and also as seemingly subservient agents of the same power. Both disappear as the central power strengthens/shifts it’s central zone towards them. In both cases the inter-relationship is complex and nuanced, and narratives of Mercian supremacy or Merovingian decedence fail to actually elucidate how these relationships actually worked.

  3. Hi Allan; this is as ever helpful. Firstly, yes, good and fair point about Adomnán; my sense is that the Vita is so distinctive in its miracles even by Irish standards that Columba as heavenly hinge-man had left him something genuine to work with, but you may well still be right that the purposes to which that’s put in the Vita is the hagiographer’s not the saint’s. Though Adomnán must have believed Columba would know!

    As to the latter point, I mean yes, these are different things a person in power on the frontier can do, and variation in the power of their centre of reference might decide whether symbiosis or itch is a better tactic. I think that dynamic is worth distinguishing though, or there isn’t really more complex this can get than ‘people far away but close also to other power centres are hard for a first power centre to control’, which is true but not very deep. In particular I think it’s important to stress that whether centre or edge dictates the terms is a variable, not a systemic factor, which some of the literature on modern frontiers wouldn’t necessarily have you believe (powerful violent state vs. subaltern craftiness all the way for them). But we agree at the end that a good account will need both sides’ perspectives considered, I think!

  4. Really interesting post that gave me a lot to think about. I can see a lot of the patterns you identify for the first type in action in the Carolingian and Ottonian marches generally. For that I might like to add the tendency for rulers to create more frontier lords to undermine the power of existing ones. For example, I highly suspect that Charles the Simple made Rollo count of Rouen to help undermine the power of Robert of Neustria. And with the Ottonians, we have some pretty clear examples of this – Otto the Great appointing Markgraf Gero to command the lower Saale district to undermine the power of the Billungs, Archbishop Bruno partitioning Lotharingia after the death of Conrad the Red or Otto II giving the Babenbergers control of what would become Austria to undermine the power of the Dukes of Bavaria. But generally, like you I think the old model that creating these frontier lords was bad for royal authority/ state power, that it inevitably led to “overmighty subjects” or powers that would inevitably go quasi-independent or break away, is inadequate.

    I also really like the model exemplified by Iona, an area I no virtually nothing about. I swear I’ve seen some other examples of that model, but I can’t pin them down right now ..

    • I can certainly see the argument with Charles the Simple and Rollo; he must have had some policy gain in mind there, after all, and that one makes sense. An earlier Carolingian example might be the Pannonian prince whom Louis the Pious seems to have tried and failed to manage, Ljudowit; I don’t see how the Carolingians can have understood as little about his doings as Thomas Lienhard thinks, but they obviously didn’t have him in hand; yet there was clearly no other viable option.* With the Ottonians I know less to check with but it sounds plausible. As to Iona, Mount Athos and Saint Catherine’s Sinai probably occupy places like Iona despite not being islands; but then as I’ve maybe shown, sometimes an island doesn’t have to be an island…

      * Thomas Lienhard, “Les combattants francs et slaves face à la paix : Crise et nouvelle définitions d’une élite dans l’espace oriental carolingien au début du IXe siècle” in François Bougard, Laurent Feller & Régine Le Jan (edd.), Les élites au haut moyen âge : crises et renouvellements, Haut Moyen Âge 1 (Turnhout 2006), pp. 253–266.

  5. Have you considered modern equivalents? I assume that Washington DC didn’t have much control over General Custer’s distant campaigns. Though I suppose they felt confident that he wouldn’t sell out to the Sioux.

    • Well, part of the idea of this project is to get the modernists comparing with us rather than the other way round as per, but if I were making that comparison, I would actually make it with the Islamic conquests in North Africa and the Iberian Peninsula, because the sources there stress exactly that problem, with the caliph not being sure he could control his general in Africa, Mūsā ibn Nuṣayr, and then Mūsā not being sure he could control his subordinate commander Tāriq ibn Ziyad, and both trying to pull the other in with sanctions they perhaps hadn’t deserved. But the concerns were real enough, as Mūsā’s son then married the widow of the last Visigothic king and some of his commanders killed him because they thought he was planning to go independent or native. And from that I would pull out several points: first, and probably most important, the sources that tell this story are at least a century later and writing under the next dynasty of caliphs, so we can’t be sure it’s real rather than a story to show how the Umayyads couldn’t control Islam properly and justly—but if so then our chroniclers were capable of seeing just the same capacity for difficulty that you’ve just highlighted, including a possible pact with the locals; secondly, that there might be a difference between the frontiers where the power gradient permits dominance, if you can marshal your resources there, which I was looking at in the post—when the Umayyads can get armies onto these frontiers the frontier lords lose, but often the Umayyads can’t do it for reasons closer to home—and those where when you marshal those resources there’s another army on the other side which may beat them (Little Big Horn or the Guadiana in 711); and thirdly, er… Not sure about thirdly, something to do with the importance of being able to set subordinates against their superiors in order to remain in ultimate command, perhaps. But it’s not a classically brilliant way to run an army, is it?

  6. Pingback: Flying Visit to Montserrat II: the place itself | 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.