Picturing Abbess Emma’s associations

Really long-time readers of this blog will maybe remember a debate that got going on this blog in June 2008, apropos of a paper in the Journal of Neurocomputing that was using medieval charter information to showcase visualisation of social networks data.¹ I was initially sceptical but talking to two of the authors got me much more interested and I subsequently talked one of them into delivering a paper in the final Problems and Possibilities of Early Medieval Diplomatic session at Leeds, a paper that we did want to publish but which sadly in the end could not be included in the final publication.² That’s still a shame as there was good stuff to think with there, but of course what any historian dealing with dense social data is going to want to know about such software and techniques is, ‘how will it help me with my stuff?’ And since answering that question usually involves a lot of data entry, it has tended to rest there.

There is also the question of how much a “Representation of the medieval social network with force directed algorithm” like this from Boulet et al., “Batch kernel SOM and related Laplacian methods for social network analysis”, fig. 1, can tell you by itself, which is why of course in that article they then spend a lot of time breaking it down

For my area of interest, this was changed in mid-2012 as the indefatigable Joan Vilaseca of the Cathalaunia website began to investigate applying such techniques to the database he maintains there, which includes quite a lot of the documents from which I ply my trade. Magistra et Mater, who was getting interested in the possibilities of these things around then too, wrote some initial thoughts about what Joan and others were doing at hers in December 2012, and I had already made a stub note to talk about it in October of that year but, well, it’s been queued ever since. There is still plenty to say, though!

The thing that particularly caught my interest was that Joan put up a post on his blog in which he produced a list of the best-connected people in his database, the ones who appear with the most other people, and once the kings who appear in dating clauses and their notaries were filtered out, pretty much top of the list was Abbess Emma of Sant Joan de Ripoll. Since there is perhaps no-one in the world who cares more about Abbess Emma than me,³ this seemed like a really good type-case with which to answer the quesion: does this kind of analysis actually tell us very much that we didn’t already know? And, weirdly, I think that my conclusion is that for me it’s perhaps most valuable for emphasising what we don’t.

Relationships of Abbess Emma in the Cathalaunia database coloured according to grade of connectivity

To talk about this it’s necessary to get you the reader clear about exactly what Joan has actually done here, of course. As simply as I can put it, what we have above is a graph built in the following way. In Joan’s database Emma appears in 50 documents and in those 50 documents she occurs with an awful lot of people. Looking for only the most meaningful, Joan excluded from the count all persons with whom Emma turned up only once, which is a lot given that she orchestrated the Vall de Sant Joan hearing in which about 500 people swore testimony for her and then there’s still 48 more documents with her in. That still leaves 112 people with whom she is recorded associating more than once, in fact the total of associations still in the count is 1292. Many of these people also relate to each other and what you have above is a computer-aided display of all those links, with Emma at the centre and everyone else pulled out to where you can see the links. But you can already see from the way that some of the links are made with thicker bands of darker colour that some of these people dominate the sample much more than others. So, who are these people? Well, if you load up the SVG version of this graph on Joan’s blog you can just click straight through to his database records, which is marvellous, but in short the top five are two priests called Gentiles and Guisad, and then three laymen, namely Reinoard, Guimarà and Tudiscle.

Archivo de la Corona d’Aragón, Cancilleria, Pergamins Seniofredo 39 (reduced-quality version), with Gentiles’s signature lower left centre

This all sounds more or less sensible to me: though I think only Gentiles and Reinoard, maybe Tudiscle, would have been in a top five I’d guessed, I can see why they’re all here. Gentiles was something like the chief scribe of Sant Joan: he appears in Emma’s first appearance as an adult, he went on appearing some six to ten years after her death, and in that time an awful lot of the documents of the abbey carried his scribal signature, even though as Federico Udina pointed out when he edited these documents, they’re not all in the same hand. This presumably means that he had subordinates signing stuff off for him and that his name was important enough that it still had to be there.⁴ Guisad was another frequent scribe for the abbey, apparently older, and he also appeared on the panel of a couple of the hearings in which Emma pursused people for her rights over their land.⁵ Reinoard was headman of one of the settlements in the Vall de Sant Joan and worked as court enforcer for Emma once as well as appearing in court when she called them, he was a collaborator of hers whom I’ve discussed elsewhere.⁶ Tudiscle and Guimarà present a more interesting case: these are two of the landowners whom Emma took to court, but in both cases those episodes were part of a longer relationship with the abbey which had here broken down. I’ve written about these two as well precisely because I think that in Tudiscle’s case the hearing was part of mending that relationship, as his importance seems not to have suffered subsequently, whereas Guimarà seems subsequently not to have worked with the abbess and in fact seems to have managed to shift quite a lot of property once donated to the nunnery onto Emma’s little brother, Count-Marquis Sunyer of Barcelona, Girona and Osona, as part of Sunyer’s campaign to clip his sister’s independently-ruled abbey’s wings.⁷ But before that he had worked for Emma, and these people certainly make good people to study if you want to understand how Emma worked, which is of course why I did.

The same relationships now displayed according to their modularity, that is, by the size of the groups internal to the data

So, the first answer to the great question about whether this tells me anything looks like ‘no’; I had already found these people by older methods. But I’m arguably not the target here: the thing is that those methods were very like what Joan’s programming all but automates. I went through the documents, made note of the names who recurred most, assembled profiles of their appearances and decided who were the people I could tell the story with. Joan’s database and graphing together mean that I could, if I was starting again, do in about ten minutes the same exercise that took me weeks when I actually did it. I could do (and may do) the same thing now with Count-Marquis Guifré II Borrell, Sunyer’s predecessor and brother, for whom I haven’t done the same kind of background data-crunching, with far less trouble than I was anticipating. So in terms of research facilitation for others, this is a huge step forward even if it doesn’t help me. I can in fact only tell how much use it is precisely because I’d done it already by another means! (Whether Joan had to put fewer hours in to make it happen than I did for my research is another question of course…)

All the same, as I said above, what it now makes me think is how imperfect our data sometimes is for the kind of questions we might like to ask. If, on first principles, we asked ourselves who the principal contacts of an early medieval abbess was, we would probably presume that the main ones were her nuns. So indeed they may have been here, but as I’ve observed in a supposedly-forthcoming paper, while Emma was in charge of Sant Joan we know the names of only two other nuns, and those are only seen as they join the nunnery, we’ve no idea what Emma’s relations with them were like.⁸ If we then allowed ourselves to remember that this abbess was a count’s daughter, we might then think about her family as an important second string. But Emma hardly shows up with her family, and when she does it needs very careful reading: I think she only occurs alive and in person with brother Sunyer in the Vall de Sant Joan hearing where she was in theory taking him to court, for example.⁹ (She also turns up as a neighbour of land he was transferring twice, but of course she wasn’t actually there for that, though it gives us another reason to suppose they had other dealings.¹⁰) Also on the defending end on that occasion was their probably-elder brother Miró, Count of Cerdanya, who in his will named Emma one of his executors and had her called ‘my most dear sister’; I think she occurs with him once otherwise.¹¹ She got Radulf to consecrate a church with her once, I think that’s it though.¹² We can more or less see from this that this set of siblings were close collaborators even if not always very willing ones but the quantity of occurrences doesn’t really reflect what we can guess the importance of those relationships would have been.

Abbess Emma’s memorial in the medieval church of Sant Joan de les Abadesses

The information we get from this, therefore, is not wrong but it is partial. Emma probably did see Gentiles and talk with him most days of her adult life. It’s not clear whether Guisad was also a priest of the abbey but if so, he also would have been a regular feature of her days. She placed a lot of reliance on Reinoard, and that relationship was probably important to both of them in raising Reinoard above his fellows and showing those fellows how Emma could reward her collaborators. Tudiscle and Guimarà, at least at first, were more of that sort of person and even if the relationships probably didn’t mean as much to Emma as that with brother Miró did, for example, they’re historically very interesting and anyone working on Emma would be well served by being pointed towards them. But there is also quiet and missing data that must have made up a great deal more of her life, and that we can’t really reconstruct. It’s not by any means the fault of this technology that it can’t bring that to our notice: it obviously can’t give us back information we didn’t put in. But that also means that the technology is no more than one of the tools we have to use to understand that information in its context, some of which context is simply what isn’t there.¹³

---

1. Romain Boulet, Bertrand Jouse, Fabrice Rossi & Nathalie Villa, “Batch kernel SOM and related Laplacian methods for social network analysis” in Journal of Neurocomputing Vol. 71 (Amsterdam 2008), pp. 1579-1573.

2. What final publication, you ask? Why, Jonathan Jarrett & Allan Scott McKinley (edd.), Problems and Possibilities of Early Medieval Charters, International Medieval Research 19 (Turnhout 2013). You could buy it here if you wanted!

3. See J. 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; idem, Rulers and Ruled in Frontier Catalonia 880-1010: pathways of power (Woodbridge 2010), pp. 23-72.

4. Ibid. pp. 29-30; see Federico Udina Martorell, El Archivo Condal de Barcelona en los siglos IX-X: estudio crítico de sus fondos, Textos 18/Publicaciones de le Sección de Barcelona 15 (Madrid 1951), p. 205 for the argument.

5. Jarrett, Rulers and Ruled, p. 59 n. 162.

6. Ibid. pp. 39, 41-42.

7. Ibid. pp. 52-53 (Tudiscle), 53-57 (Guimarà) & 64-65 (Sunyer’s pressure on the nunnery).

8. J. Jarrett, “Nuns, Signatures and Literacy in late-Carolingian Catalonia”, to appear in a Festschrift for Rosamond McKitterick first planned in 2010.

9. The hearing is best printed in Ramon Ordeig i Mata (ed.), Catalunya Carolíngia IV: els comtats d’Osona i Manresa, Memòries de la Secció Històrico-arqueològica LIII (Barcelona 1999), 3 vols, doc. nos 119 & 120, though the palæographic notes of Udina, Archivo Condal, doc. no. 38, are still very useful.

10. Ordeig, Catalunya Carolíngia IV, doc. nos 103, 105 & 155. There are a few other cases where she and Sunyer both turn up as neighbours, but not of the same properties, so I don’t think that really counts here.

11. The will is only printed in Prosper de Bofarull y Mascaró, Los condes de Barcelona vindicados, y cronología y genealogía de los reyes de España considerados como soberanos independientes de su marca (Barcelona 1836, repr. 1990), 2 vols. I pp. 88-90. In Ordeig, Catalunya Carolíngia IV, doc. no. 57 Miró presided over a hearing where Emma was the plaintiff, but she was represented by a mandatory and not present herself.

12. Udina, Archivo Condal, doc. no. 73. The two also occur as common neighbours in Ordeig, Catalunya Carolíngia IV, doc. nos 155 & 419, but again that only tells us that they probably met at some other point.

13. Cf. Jarrett, “Poor Tools to Think With: the human space in digital diplomatics” in Georg Vogeler & Antonella Ambrosiani (edd.), Digital Diplomatics 2011, Beihefte der Archiv für Diplomatik (München forthcoming), pp. 291-302.