Seminar CLXIX: Spanish palaces in Winchester

The next seminar in my backlog of reports, horrifyingly lengthening not shortening despite the regular updates, was another visit of Anglo-Saxon England to the Earlier Middle Ages Seminar at the Institute of Historical Research, in the form of George Molyneaux from Oxford with a paper called “The Formation of the English Kingdom in the Tenth Century” on 13th March 2013. I went along, by way of showing the flag, but I’m not going to cover it here, for one thing because Magistra et Mater did already, and for another because it was quite acknowledgedly more or less the same paper I had earlier seen George give in Oxford and which I already blogged here. George did suggest I cover it anyway, with such phrases as “remarkably little development in the speaker’s thought” and so on but what can I say, I’m too far behind, and anyway if I miss that one out I now get to talk about me, always a hard temptation to resist.

The oldest part of the King Alfred Campus, University of Winchester, looking rather newer on the inside as I discovered

I have been quite looking forward to writing about this one, because it marked the first time I’d got up in public to talk about my work for quite a while. Reaching it thus marks some kind of exit from the slough in which I’d temporarily found myself at Oxford and my starting to gather myself for whatever was coming next, and apparently I did this by accepting the kind invitation of Dr Kate Weikert to host me at the Winchester Seminar on Comparative Medieval Cultures the very next day, on 14th March 2013. The word ‘comparative’ carries more strength here than usually it might, because the set-up with that seminar series was (is, perhaps) that there would be two papers in each evening, chosen to complement each other and provoke, well, comparison. So I stepped up first with “Brokedown palaces or Torres dels Moros? Finding the fisc in late-Carolingian Catalonia” and then local hero Dr Phil Marter followed on with “Archaeological Investigations at the Medieval Palacio de Ambel, Aragón”, and this actually worked really well, it was one of the more fun seminars I’ve been part of.

An aerial view of Sant Esteve and Santa Maria de Palautordera, Girona

I’ve been trying to work out what to do with my paper ever since, because it was something of an attack piece and I already have too much of a reputation as a negative scholar. All the same, you know, dear reader, that sometimes I feel scholarly outrage rather keenly and this paper was one of those. It was about places in Catalonia which bear a name in the form Palau- or Palou-, of which there are many and which are not fully understood. As long-memoried readers may remember, even some of the ones that are explained turn out not to be when one looks… The root is very clearly Latin palatium, which is what gives us English ‘palace’, but that can’t be what is meant here unless, as I said, conscious that there would be Anglo-Saxonists in the audience, you remember that sites like Cheddar have been called palaces, that is, big halls with some supporting sheds.¹ A team led by Professor Ramón Martí at the Università Autonòma de Barcelona have made these names their own and come up with ninety-odd across the counties of old Catalonia, and some of those I’ve come across are only a matter of miles from each other; they’re just too thickly spread to be major élite settlements. Up until quite recently, there was only really one going explanation of these sites, that favoured by Pierre Bonnassie, that they were fiscal estate centres that probably went back to the Romans, but Bonnassie’s sense of the word ‘fisc’ was so broad, applying to properties the counts bought and then sold again the next day, for example, that this doesn’t actually explain as much as it might.² And perhaps thus, it was in Bonnassie’s Festschrift that Professor Martí published the first version of his alternative theory, which he and a team of researchers have been filling out ever since, that they are in fact the relics of Muslim garrisons from the period of Islamic rule in Catalonia between 714 and 785.³

Distribution map of place-names in palatium and palatiolum in Catalonia, from Cristian Folch Iglesias and Jordi Gibert Rebull, ‘Arqueològia, documentació escrita i toponímia en l’estudi de l’Alta Edat Mitjana: els casos dels topònims pharus, monasteriolum i palatium’, Estrat Crític 5.2 (Barcelona 2011), 364-377 at p. 370

Although I did so in the seminar, I don’t want to reprise either Professor Martí’s arguments for this or mine against it here, partly because the open web is not where to start such arguments, but mainly because as I say, I don’t know what I’m doing with this paper yet. A basic problem that is worth expressing openly, however, is that as far as I can discover, the map above is as close to a list of their ninety-odd sites that Professor Martí’s team has published, which makes even agreeing with them quite difficult. One really wants to know what the evidence is. So, for the seminar, I made a list. I went through all my various charter notes and the volumes of the Catalunya Carolíngia and Atles Històric de Catalunya that I have and found forty-eight such sites, not too bad considering that I could barely touch the counties of Barcelona, Cerdanya or Urgell. The main thing I learnt from that was something that Professor Martí’s team would also acknowledge, I’m sure, that no one explanation will deal with all of these sites. The idea of them as fiscal complexes or ancient Roman centres runs into immediate trouble when one realises that the earliest recorded one in my list was a new-build, put up before 832 by Abbot Castellano of Arles in an area he’d just cleared from wasteland.⁴ And in fact, the only reason we can be sure that any of these sites were not new-builds, however unlikely it seems that they should be, is archæology, and here we mean actual digging since the eight sites that have been surveyed by Professor Martí’s team all produced ceramics that could be early enough for their argument, but equally could not.⁵

Actual digging underway at l’Aiguacuit de Terrassa, site of a Palacio fracta

Well, here again fuller publication of results would really help, but as far as I can discover, including from the Martí team’s own publications, only three of these sites have been dug: Palofret in Terrassa, Les Palats in Carcassonne and l’Hort de Pelat in Riudoms.⁶ The first of these was at least active in the right period, and the latter two of these produced unusual burials, one of which is in fact almost certainly Islamic, though I’m less sure about the crouched burials at Les Palats, which the original excavators thought Visigothic and which have now been lost. All three of the sites, however, and a number of others, were once Roman villas, and until we get something more like a list of sites with their evidence from Professor Martí’s team I do feel as if that might be a simpler explanation of these place-names, although I do note that many of these places do appear to have had fiscal connections and operated as estate centres.⁷ That is, however, firstly not incompatible with them having previously been villas, and secondly what we would expect from similar work in Castile and the Carolingian world where palatium is exactly what you call a rural complex at which renders are collected.⁸

Façade of the Palacio de los Hospitalarios, Ambel, Aragón, from Wikimedia Commons

As it turns out, this is also plausibly the case in Aragón, because the Palacio de Ambel, about which Dr Marter was talking, is or at least was indeed a rural estate centre where renders were collected. That rather minimises its very complex history, though, the earliest parts of which are pretty obscure. What you are looking at there is, functionally, the outside of a really posh nineteenth-century block of flats. The trappings of that have been stripped away, however, to reveal a Renaissance grange of the Hospital of Saint John, for whom its Preceptory died in the siege of Malta of 1515, which seems to be depicted in a few surviving wall-paintings inside the building. And it really is inside, too: the current roof, complete with arcade, is directly over three small towerlets invisible under the tiles, between which it’s actually possible to clamber on top of the vaulting that used to hold up the old roof, and now just holds up the ceiling on its underside. Before the Renaissance phase, this was a complex of buildings rather than the single quadrangle arrangement, and one of those buildings was a Gothic church, erected by the Knights Templar from whom the Hospital got the place when the Order was suppressed. Its rood screen is still there and behind it is Islamic decoration in geometric interlace, and internal decoration that includs fake bricks painted over the stone courses, but all of this is Templar-period, not Mudejar. The church, though, is probably the oldest bit of that complex, because it reuses a circular tower, which is reckoned a Torre dels Moros (as almost everything early in Spain seems to be) but which, being built of packed earth on a stone lower course, isn’t giving away much with its architecture. It is probably ninth-century, which is still Islamic in this area, but the dating evidence is basically guesswork, so other schemes could be considered.

Decorations now inside the church of Sant Miquel Arcángel, Ambel

Dr Marter and his team, or teams of which he has been part, anyway, have been working on this place for years, and what they are mainly doing is trying to stop it falling down and slowly restore it to its medieval configuration, which has involved such things as removing trees, finding a sixteenth-century letter hidden in the wall, and so on. But there was also time for some reflection on how the building had gone through its earliest sets of changes, and why the church wears such Islamic decoration. Was there an existing church in the area that this one replaced, and whose existing congregation, presumably Mudejar or Mozarab or whatever one wished to call them, culturally Arabicized, needed to be comforted that the new lords understood who they were and what was particular to them? Maybe, though if so sticking a Gothic rood screen in the way perhaps cancelled that message. Anyway, it seems clear that the place has lots to tell even after so many years’ work. And both of us got to think quite hard with each other’s examples about what one calls a Palacio on the Christian-Muslim frontier of Spain, and what work a palace really did anyway, and what it might once have been so as to wind up performing those functions. It was a good evening and I hope to see and indeed take part in more seminars so well configured in the future.

---

1. Obviously I have not yet got bored of citing John Blair, “Palaces or minsters? Northampton and Cheddar reconsidered” in Anglo-Saxon England Vol. 25 (Cambridge 1996), pp. 97-121, DOI: 10.1017/S0263675100001964.

2. Pierre Bonnassie, La Catalogne du milieu du X^e à la fin du XI^e siècle : croissance et mutations d’une société, Publications de l’Université Toulouse-le-Mirail 23 & 29 (Toulouse 1975-1976), 2 vols, I pp. 144-153.

3. Ramón Martí, “Palaus o almúnies fiscals a Catalunya i al-Andalus” in Hélène Débax (ed.), Les sociétés méridionales à l’âge féodal (l’Espagne, Italie et sud de France X^e-XIII^e s.) : hommage à Pierre Bonnassie, Méridiennes 8 (Toulouse 1999), pp. 63-69.

4. Pierre Ponsich (ed.), Catalunya Carolíngia VI: els comtats de Rosselló, Conflent, Vallespir i Fenollet, ed. Ramon Ordeig i Mata, Memòries de la secció històrico-arqueològica LXX (Barcelona 2006), doc. no. 17.

5. When I gave this paper, the most recent publication of the team’s theories seemed to be Cristian Folch Iglesias and Jordi Gibert Rebull, “Arqueològia, documentació escrita i toponímia en l’estudi de l’Alta Edat Mitjana: els casos dels topònims pharus, monasteriolum i palatium” in Estrat Crític Vol. 5 (Barcelona 2011), pp. 364-377; there is also Gibert, “La integració a al-Andalus dels territoris a ponent del Llobregat” in Butlletí de la Societat Catalana d’Estudis Històrics 16 (Barcelona 2005), pp. 39-72 at pp. 50-55; Ramón Martí, “Palacios y guardias emirales en Cataluña” in A. Riubal (ed.), II Congreso de Castellología Ibérica, Alcalá de la Selva, 2001 (Madrid 2005), pp. 293-309; and Ramón Martí, “Del fundus a la parrochia. Transformaciones del pobliamento rural en Cataluña durante la transición medieval” in Philippe Sénac (ed.), De la Tarraconnaise à la Marche Supérieure d’al-Andalus : les habitats ruraux (IV^e-XI^e siècle). Desde la Tarraconense a la Marca Superior de al-Andalus: los asentamientos rurales (siglos IV-XI), Méridiennes : Études Médiévales Ibériques 2 (Toulouse 2006), pp. 145-166, among other publications that more or less replicate these, though there might be newer ones I’ve missed.

6. The data for the latter two sites have to be strained from the publications in the previous note: there is no stand-alone publication of them that I’ve discovered, and these articles give you little more than a few lines on each. For Palofret, there is Joan Soler i Jiménez and Vicenç Ruiz i Gómez, “Els palaus de Terrassa: estudi de la presencia musulmana al terme de Terrassa a través de la toponímia” in Terme Vol. 15 (Terrassa 1999), pp. 37-51, online here. This article was written in the liught of Martí’s first publication of his theory, so that the interpretation of the site as Islamic is partly following him.

7. I get the Roman data also from the articles cited above, which is a bit master’s tools, but presumably the data is all equally valid.

8. See for example José Angel García de Cortázar & Ester Peña Bocos, “El palatium, símbolo y centro de poder en los reinos de Navarra y Castilla en los siglos X a XII” in Mayurqa Vol. 22 (Mallorca 1989), pp. 281-296; Josiane Barbier, “Les lieux du pouvoir en Gaule franque : l’exemple des palais” in Carl Ehlers (ed.), Deutsche Königspfalzen: Beiträge zu ihrer historischen und archäologischen Erforschung. 8: Places of power, Orte der Herrschaft, Lieux du pouvoir (Göttingen 2007), pp. 227-246; Darryl Campbell, “The Capitulare de Villis, the Brevium exempla, and the Carolingian court at Aachen” in Early Medieval Europe Vol. 18 (Oxford 2010), pp. 243-264.