Al-Mansur’s failure to conquer the north of Iberia

This is the second of the reaction posts I promised following my much-backlogged report on the 2018 International Medieval Congress. One of the papers I’d been to see was by one Josep Suñé Arce, and was called "Was the Umayyad Caliphate of Córdoba as Strong as Arab Chroniclers Claimed?" I wanted to see this mainly because I couldn’t see how the question could be answered as framed, on the one hand because it’s two different grades of subjectivity – imagine an answer like, “Well, the chroniclers give it a clear 5 across the board but I score it B-!” – but on the other because the evidence from which the chroniclers would have to be falsified is, well, their chronicles. The only way I could intellectually conceive of the question being answered was as some version of ‘can we tell if the chroniclers were making stuff up?’, but if they were, then how would we know what was really the case?

Now, if he ever reads this I hope that Dr Suñé will forgive me that scepticism, because actually his paper was way more interesting than I’d unjustly feared. But, additionally, in early 2019 a fuller version of it came out in print.1 But because my first reactions were to the 20-minute conference paper, not the final article, I think it’s interesting to start with what I thought of the paper, and to see how the article differs. The argument of the paper fell into three parts, as follows:

  1. The sources are biased: they are based on official records which had no interest in a neutral viewpoint; and they are, especially in the case of the eleventh-century chronicler Ibn Hayyān, tinged by a nostalgia for the strong caliphate born of living through the subsequent taifa era in which Christian raids helped break up the Andalusī state and Ibn Hayyān’s own family were ruined and he had to flee to Morocco.
  2. The sources report victory much more than defeat, even when the Christian sources of the time tell the opposite story; but these victories didn’t lead to conquest or elimination of any of the targets.
  3. From other sources we can even see that the Catalans and Navarrese gained ground against Islam in the period; we also see that the Caliphate expended a lot of gold to try and keep the Christians from banding together, so they were actually making their enemies stronger.

So the answer to my methodological objection was, obviously, use other sources, and fair enough, that’s me told. But I now had further questions, as the saying goes. The question I actually raised was whether this was even what Ibn Hayyān cared about, because at this stage I had a different dissertation pupil with whom I was coming to the conclusion that what Ibn Hayyān’s overall historical argument was that Islam has struggled with internal divisions pretty much from the death of the Prophet onwards, but that wherever they were involved the Berbers made things worse. But now I’d want to ask, most obviously, surely the Christian sources have exactly the same biases; can we really use them as a check on later Arabic sources when they’re just as interested in presenting their own side as ever-victorious and righteous, especially since the most relevant chronicles were actually written for rulers prosecuting campaigns against Islam whereas at least Ibn Hayyān wrote after the fall of the Umayyads and so out of reach of the distorting effect of their patronage (though admittedly, his main sources did not).2

On the other hand, and the reason I flagged this paper as one I needed to think about more, Dr Suñé was not wrong that despite everything that al-Mansur inflicted upon the Christian kingdoms, their territory did expand in this time. In the case of Catalonia, I think I can explain it as the filling-up (by government, rather than by people) of a substantial unclaimed space between the two polities; it wasn’t that the Muslims were losing ground, it’s more that since about 827 no-one beyond it had really ruled the space between, say, Lleida and Barcelona, and now, by creeping settlement and governmentalising processes I’ve written much too much about here already, the Christians partly did. I don’t know enough about Navarra but there it seems to be more complex: a land which had been notionally under the pact, and thus inside the dar al-Islam, but was really only controlled by the intercession of our favourite frontier warlord clan, the Banū Qāsī, was lost when they rebelled and were crushed, and so the edge of direct control actually expanded under the caliphate, as their territory was taken over, but its notional extent shrank because Navarra was now lost.3 Navarra then expanded somewhat during the reign of al-Mansur, sure, but mainly with respect to its Christian neighbours and, as I’ve pointed out, was by 1030 or so pretty much in charge of the north under Sancho the Great.4 But was that anything to do with the Caliphate? Perhaps only because its raids had diverted and weakened Navarra’s more exposed competitors, I’d say. So here I would have had counter-arguments, and it was presumably some sense of what these were that made me flag the paper.

(Obverse of) gold dinar of Caliph Hishām II of Spain, 999-1000, Grierson Collection, Fitzwilliam Museum CM.PG.1192

(Obverse of) gold dinar of Caliph Hishām II of Spain, 999-1000, Grierson Collection, Fitzwilliam Museum CM.PG.1192

Reverse of the same coin

Reverse of the same coin

But the bit I can’t contradict, or even explain, is the flow of gold, because it is very evident in the charters I know so well. I was by no means the first person to spot that over the course of the 970s and 980s the main operating currency around the city of Barcelona became Muslim gold – that was Gaspar Feliu, whom all praise – but I could see what he’d seen very easily. Even in the years after the sack of Barcelona in 985, a decent part of the payments that were made to sort out land tenure and endow repaired foundations were in mancuses.5 If they were at intermittent but perpetual war with the Caliphate, where was all the gold coming from? Both Feliu and after him Pierre Bonnassie opted for trade that we can’t really see as an explanation; we see the foreign goods people could acquire with this money and we see the money but we don’t see anyone actually doing the trade, and no-one is able to explain what the people of Barcelona had to sell that was bringing in so much gold except for waving their hands at the idea of slavery, which is fine except that again there’s no positive evidence; Arabic sources don’t talk about buying slaves in Barcelona and Christian ones don’t talk about importing slaves to Barcelona or enslaving people, or even really feature slaves in any number.6 Of course the goods were, in Graham Swift’s immortal wording, ‘perishable’, and therefore so might be the record, but it’s still a big silence.7 Maybe, therefore, diplomatic pay-offs are part of the answer (though I have to say that, prior to the succession Ramon Borrell in 993 at least, the counts did not pay for things they bought in mancuses, something that I’ve only just really realised this moment, so how those payments got into the market I don’t know).8

So there matters could have rested, except that firstly Dr Suñé published the article, in a journal I was about to turn up in myself indeed, and secondly I had a dissertation student who was essentially asking the same question, so I grabbed it down. But did I read it? Well, got to admit, no, I just never got to it. But because I care about you, my readership, and also about not looking badly underinformed when I write this blog, I have read it now, and it is a serious piece of work that I’ve had to think hard with. The basic contention Dr Suñé wants to make is that the Umayyad Caliphate was never really strong enough reliably to dominate all its Christian neighbours, that even at its most militarised it was unable to prevent them overall gaining territory from it, and that we should not see its fall as internally caused, but as the result of it having had to feed gold steadily to its enemies for some decades when what turned out to be a fatal civil war broke out in 1009 and both sides did what the government had been doing for ages and enlisted Christian help. Of course the chronicles don’t say this like this, but they wouldn’t, would they, and you can see it in the whole source complex even so. Such, anyway, is the argument. It’s quite a complex thesis, and it rests on a knowledge of both Christian and Arabic materials I’m not sure anyone’s brought to it before, and in particular a deep knowledge of the works of Ibn Idhārī, a thirteenth-century African chronicler with a very detailed account of the events in question.9 I have found about at least two Catalan border skirmishes I’d no idea were recorded by reading this article. I also don’t find the basic argument implausible in this fuller version. That being said, there are still a couple of things I think aren’t fully proven, and one of them prompts me to wonder if there isn’t another, slightly different way to read what was going on in the Iberian Peninsula over about 970 to 1010.

So, the first question I have is over the amount of ground the Christian principalities supposedly gained over the Caliphate during its purportedly dominant phase. To start with, there is the argument I raised above, about incursions into no-man’s land rather than conquest from an enemy. It’s not that Suñé has no evidence for this happening, but a lot of it is either very early, as in, dating to before the Caliphate proper and therefore during the period of Andalusī civil war around or just after 900, when to be honest it’s not really clear who owns these places even when they’re lost; or it’s very late, as in during or after that civil war of 1009-1013 we already mentioned; or it’s a bit weird.10 What do I mean, weird? Well, one reference, which is supposedly to prove Catalan gains in Anoià and Penedès, in Manresa and Barcelona respectively, cites a charter covering Castell Cornil in more northerly and easterly Osona instead, and not the earliest one from there either; I don’t understand what the point of this is, and it certainly doesn’t establish that these had previously been Muslim territories, rather than unclaimed.11 Another cite is la Garde-Freinet, the Muslim coastal fortress site in Provence that I’ve written about, which was indeed lost to the Christians in 972/73, but which was not clearly an Umayyad possession and which in any case hardly reflects on the situation in the Iberian Peninsula.12 I’m left thinking that there may not be that much good evidence for what Suñé argues here, as opposed to the governmental creep I know we can see in the charters.

La Garde-Freinet, seen from the fort on Massif des Maures

La Garde-Freinet, seen from the fort ruins on Massif des Maures, unquestionably a Muslim possession lost to the Christians but, importantly, not in the Iberian Peninsula… Photo by Patrick RouzetOwn work, licensed under CC BY-SA 3.0 via Wikimedia Commons

But there is also the question of the gold, as mentioned above. This is a problem that Dr Suñé faces head-on, because he sees the flow of wealth northwards even during the belligerent phase of the caliphate’s existence as crucial to the way it undermined itself. To his credit, he does not say it was trade or slaving that brought this money northwards, but his explanation still isn’t very convincing. Firstly he notes that Berber troops recruited from Africa were usually paid; then he notes that al-Mansur’s armies (but not those of his caliphal predecessors) had Christian contingents; so he assumes they must have been paid too. To this he adds the infamous tribute payments, the parías, which Anna Balaguer has shown funnelled huge amounts of wealth into Catalonia and Aragón. But the thing is, she shows it for later, because the parías, again, didn’t start till the civil war of 1009-1013. Suñé’s citation of her work says that it demonstrates, “the inflow of Andalusian gold into the Catalan counties during the period 941–1180”, but that date range is really misleading, because actually Anna notes a single raid of 945 (not 941), which is itself much debated, and then there’s nothing in her tables till the civil war.13 And this makes sense, because you’d hardly expect the Andalusī state to be paying tributes to the same people it was cyclically raiding at the same time. So that only leaves payment for military service, which worryingly is not actually attested, again, until the civil war. The only good example Suñé has before that is of exactly the thing he thinks didn’t normally happen, conquest and then subject military service, as some Leonese frontier counts around Astorga were apparently so subdued by al-Mansur’s attacks in 997 that they agreed to pay the Muslim tax on Christians and serve in the army when summoned, and were thus part of the infamous sack of Santiago de Compostela later that year, on the Muslim side.14 To me, this just doesn’t look like the kind of flow of money northwards that could fund a military recovery.

And yet, as said above, the gold did go north, whether we can explain it or not. So perhas a better question is whether Suñé is right about what the Umayyads, and their ‘Āmirid quasi-deputies, were actually trying to do. He reproaches them for not achieving the conquest of the places they attacked despite three or four or more goes, and says that without definitive military victory it’s clear that the Christians would always return to being a threat.15 I’m not sure what kind of victory would be definitive enough for him here; was such a victory achieved at any point in the history of Islamic Iberia? Maybe las Navas de Tolosa, supposedly the last hope of a unified Muslim resistance to the Christian conquests, but even that’s a debate.16 But it seems that Dr Suñé thinks that the natural aim should have been conquest and direct political takeover. To that I say that it just doesn’t seem to be what the rulers of al-Andalus set out to do, at least not since around 720 or so. Rather, their aim appears to have been to pillage, burn and terrorise, with the hope of two quite separate results: one, of immediate importance, booty with which to satisfy the military when they went off duty, and two, less important, possible submission from the enemy, thus guaranteeing safety of Muslims near the frontiers and, if there were any of those, beyond. Sometimes, as we’ve seen, that submission might one way or another technically bring the Christians into direct political subjection, at least for a while, but if they recognised the caliphate’s authority long enough not to interfere with it, that was often enough, as the caliphate usually had its own problems to sort out anyway. At any rate, this is how I see it after a few years teaching it.

Soldiers of al-Mansur, depicted in the thirteenth-century Cantigas de Santa Maria

Al-Mansur’s army, as depicted in the Cantigas de Santa Maria of King Alfonso X of Castile, public domain via Wikimedia Commons

In the 970s, however, this shifted as al-Mansur realised that he could to a large extent solve those problems by recruiting soldiery from many more sources, so as to cut down on the ability of any one military power-base to defy him, and by paying for that by raiding the Christians much more often.17 Now, I think al-Mansur was smart enough to realise what many historians have said since, that the problem with a polity based on conquest is that it has to keep conquering, or else change its power-base. Any change to the power-base would probably have toppled al-Mansur from power, though; the continual, successful jihad he waged was critical to his importance. So conquest was not his aim; instead, the Christian principalities were, I think, a flock of metaphorical geese who, if killed, would stop laying the golden eggs of booty with which he paid his troops and celebrated his triumphs. Even removing their productive capacity would have made it harder for him to stay in control in Córdoba. So if gold got north, by trade, slavery, diplomatic payment, whatever, I think that was good for him, not least because he probably expected to take a decent tithe of it back southwards again every six or seven years. It might even have been a stable and reproducible system for a while. Now, I think Suñé may well be right that what went wrong with the system is that the Christians got too big to quell, and that al-Mansur’s less successful sons either couldn’t win so easily or picked their targets less well, lost the assurance of success and became vulnerable to internal opposition. But I think he may be missing the point of the ‘Āmirid strategy to say that they failed to conquer their opponents when they should have; to do so would probably have been the end of them. Instead, I think, like other rulers of many different sizes at the same sort of time, they found themselves facing the problem of what to do when enough people around you are getting rich that they no longer need to pay you as much attention as before to keep them important.

There isn’t really a good account of the history of the Iberian Peninsula either side of the year 1000 in English. Roger Collins’s is good, but is only twenty pages; Peter Scales’s old book gets at many of the important issues, but essentially does so by silently transcribing Ibn Hayyān; and I find myself usually recommending the first chapter of David Wasserstein’s book on the taifa kings even though it’s even older and principally about something else.18 This article by Suñé is a big step towards one, but it’s necessarily involved in debates which would completely swamp someone who was new to the era. There’s still no adequate account. Maybe I have to write one, some day. Or maybe Dr Suñé should, because I would definitely read it if he did!


1. Josep Suñé Arce, “Was the Umayyad Caliphate of Cordoba as Strong as Arab Chroniclers Claimed?” in al-Masāq: Islam and the Medieval Mediterranean Vol. 31 (Abingdon 2019), pp. 35–49, DOI: 10.1080/09503110.2018.1553376.

2. Ibn Hayyān’s principal sources were, supposedly, mostly-lost chronicles by two father-and-son historians, ‘Isā and Ahmad al-Razī, who wrote under the Caliphate. For more on Ibn Hayyān’s chronicle and its problems for us, see Manuela Marín, “El «Halcón Maltés» del arabismo español: el volume II/1 de al-Muqtabis de Ibn Ḥayyān” in al-Qanṭara Vol. 20 (Madrid 1999), pp. 543–549. There’s nothing in English, which is going to be a bit of a theme for this post.

3. I covered this in an IMC paper of my own long ago, but it’s no closer to being in print so instead I have to refer you to Jesús Lorenzo Jiménez, La dawla de los Banū Qasī: origen, auge y caída de una dinastía muladí en la frontera superior de al-Andalus, Estudios Árabes e Islámicos: Monografías 17 (Madrid 2010), on which it was heavily based, because there’s (yes) nothing in English.

4. Jonathan Jarrett, “Before the Reconquista: frontier relations in medieval Iberia 718 to 1031” in Javier Muñoz-Basols, Laura Lonsdale and Manuel Delgado (edd.), The Routledge Companion to Iberian Studies (London 2017), pp. 27–40, DOI: 10.4324/9781315709895.ch3, at pp. 29-30. There’s nothing (else) in English on Sancho the Great, except as patron of sculpture, and I really wish there was.

5. Gaspar Feliu y Montfort, “El condado de Barcelona en los siglos IX y X: organización territorial y económico-social” in Cuadernos de Historia Económica de Cataluña Vol. 7 (Barcelona 1972), pp. 9-31, translated as Gaspar Feliu i Montfort, “El comtat de Barcelona als segles IX i X: Organització territorial i econòmico-social” in Feliu, La llarga nit feudal: mil anys de pugna entre senyors i pagesos (València 2011), pp. 63–91. Such work as there is in English just refers to this, including my own, Jonathan Jarrett, “Currency Change in Pre-Millennial Catalonia: Coinage, Counts and Economics” in Numismatic Chronicle Vol. 169 (London 2009), pp. 217–243.

6. The best recent attempt to put together what evidence there is for this trade is Thomas Freudenhammer, “Rafica: Frühmittelalterlicher Karawanenhandel zwischen dem Westfrankenreich und Al-Andalus” in Vierteljahrschrift für Sozial- und Wirtschaftsgeschichte Vol. 105 (Stuttgart 2018), pp. 391–406. There is nothing in English…

7. Graham Swift, Last Orders (London 1986), p. 285.

8. This’d be a long footnote if I gave all the references, but the short version is that in all of Borrell II’s sales he pays in solidi, or at least the price is given in them. Since there were no coins of that denomination in Catalonia, what the prices were actually paid in is another question, but since Borrell did run his own coinage—see Jarrett, “Currency Change”—you’d expect him to use that, really.

9. Very little of Ibn Idhārī is available to me, at least until and unless I actually learn Arabic. The bit I know is Giorgio Levi della Vida, “Córdoba de la primera a la segunda conquista de la ciudad por los berberiscos (Nov. 1009–May. 1013) seg&uuacute;n al-Bayān al-Mugrib de Ibn ‘Idārī”, ed. Claudio Sánchez Albornoz & trans. I. Arias in Cuadernos de Historia de España Vol. 5 (Buenos Aires 1946), pp. 148–169, which is fascinating but not the whole text by any means! It does, however, include the bit that Dr Suñé uses as a worked example (“Umayyad Caliphate”, pp. 4-5) of how, when faced with multiple accounts, the Arabic chroniclers choose the highest numbers for Christian casualties they can find. Now, this is actually an odd bit to choose, as though the sources do say what Dr Suñé says they say, on this occasion the victorious Muslims were Berber troops actually opposing the Catalan mercenaries bought in by one of the Ummayad claimants in the civil war of 1009-1013. This can’t really be pro-Umayyad bias, therefore, and I think this is an agenda that Dr Suñé misses. Ibn Idhārī seems to me, indeed, to have been writing at least partly to argue against Ibn Hayyān, who seems to have blamed Berbers for almost everything that went wrong in al-Andalus. Throughout his account of the civil war, therefore, Ibn Idhārī argues that the Berbers were always the staunch and righteous defenders of al-Andalus, but the corrosive fear and mistrust that they met from their supposedly fellow Muslims undid all their good attempts. What this means for the argument here is that we need to consider that the chroniclers had specific as well as general biases…

10. Referring especially to Suñé, “Umayyad Caliphate”, pp. 11-12, with conquests of 907-918 (p. 11) or of lands which we simply don’t know were ever Muslim ruled, because the evidence cited is the work of the team led by Ramon Martí on place-names in palatio. You know my views on that theory, but in case you thought I was a lone voice in the wilderness see also Xavier Ballestín, “Consideraciones acerca del termino árabe balāṭ, su equivalencia con la voz latina palatium y su presencia en las fuentes andalusíes, magrebíes y orientales” in Ballestín and Ernesto Pastor (eds), Lo que vino de oriente: horizontes, praxis y dimensión material de los sistemas de dominacion fiscal en al-Andalus (ss. VII-IX), British Archaeological Reports (International Series) 2525 (Oxford 2013), pp. 28–42.

11. Ramon Ordeig i Mata (ed.), Catalunya carolíngia volum IV: Els comtats d’Osona i Manresa, Memòries de la Secció Històrico-Arqueològica 53 (Barcelona 1999), 3 vols, I, no. 654, cited by Suñé, “Umayyad Caliphate”, p. 12 n. 74. The other primary source in the note takes a 1022 report of Islamic presence at Montserrat four generations earlier as true; but not only do we have documentation from Montserrat from nearly a century before (e. g. Ordeig, Catalunya Carolíngia IV, I, no. 273, of 924) that makes no mention of this, there was an obvious value to the rhetoric of conquest (which I discuss in Jonathan Jarrett, “Power over Past and Future: Abbess Emma and the nunnery of Sant Joan de les Abadesses” in Early Medieval Europe Vol. 12 (Oxford 2003), pp. 229–258, DOI: 10.1111/j.0963-9462.2004.00128.x), and here again I can’t but feel that Dr Suñé has not felt it necessary to subject his Christian sources to the same critique as he does his Arabic ones.

12. My piece is Jonathan Jarrett, “Nests of Pirates? ‘Islandness’ in the Balearic Islands and la-Garde-Freinet” in al-Masāq Vol. 31 (Abingdon 2019), pp. 196–222, but Suñé cites Kees Versteegh, “The Arab Presence in France and Switzerland in the 10th Century” in Arabica Vol. 37 (Leiden 1990), pp. 359–388, which to be honest is a better starting point, as I had other points to make in mine.

13. Suñé, “Umayyad Caliphate”, pp. 13-14, citing p. 13 n. 84 Anna M. Balaguer, Del Mancús a la dobla: Or i paries d’Hispània, Col·lecció J. Botet i Siso 2 (Barcelona 1993), pp. 42 & 53, the table in question being on the latter.

14. Suñé, “Umayyad Caliphate”, p. 10, mentioned again p. 13.

15. Ibid. pp. 10-12.

16. For example, compare Martín Alvira Cabrer, “Las Navas de Tolosa: the beginning of the end of the ‘Reconquista’? The battle and its consequences according to the Christian sources of the thirteenth century” in Journal of Medieval Iberian Studies Vol. 14 (Abingdon 2012), pp. 45–51, and Bernard F. Reilly, “Las Navas de Tolosa and the changing balance of power”, ibid., pp. 83–87, part of a special issue of nine short articles about the significance of the battle.

17. For al-Mansur I tend to use Philippe Sénac, Almanzor: el azote del año mil, trans. Antoni Furió (Valencia 2011), simply because I own it, but there is also Xavier Ballestín, Al-Mansur y la dawla amiriya: una dinámica de poder y legitimidad en el occidente musulmán medieval, UB 78 (Barcelona 2004), which must also be worth a look, and at least an introduction in English can be found in Roger Collins, Caliphs and Kings: Spain, 796-1031, A History of Spain 5 (Chichester 2014), pp. 185-198.

18. Ibid., pp. 185-204; Peter C. Scales, The Fall of the Caliphate of Córdoba: Berbers and Andalusis in conflict, Medieval Iberian Peninsula 9 (Leiden 1994); David Wasserstein, The Rise and Fall of the Party-Kings: politics and society in Islamic Spain 1002-1086 (Princeton NJ 1985), pp. 55-82.

4 responses to “Al-Mansur’s failure to conquer the north of Iberia

  1. Thomas Freudenhammer

    Dear Jonathan,
    Thank you for quoting my article on “Rafica” under note 6. There is now an English translation available in “Academia.edu” under my name…
    Best regards,
    Thomas Freudenhammer

    • Thankyou very much, Dr Freudenhammer, that is very useful. I have wanted to set that article to students before now, but a Germanophone student is a sadly rare thing in the UK…

  2. JJ: from a seminar I once learned that in the late 1950s the undergraduates in Chemical Engineering (then called “Technical Chemistry”) at Edinburgh entered the university with Latin and French, and learned German at university, with one of their Finals papers being in German. And all this for chaps whose main skills and interests were in maths, physics, chemistry, and engineering.

    On the other hand the economist Keynes was too lazy to learn German even though much of the most important work of his era was written in German.

    Intellectual seriousness would seem once to have been a mark of the Faculty of Science. Dear dead days.

    • A foreign-language exam paper in a non-language subject is taking that commitment fairly seriously! I have occasionally raised the idea that we need a language component in our History degrees but wherever I’ve done it, the opinion has been the same: it’d hurt our student recruitment horribly. As with so many of these things, the problem begins further down the pipeline…

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 )

Google photo

You are commenting using your Google 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.