The more that I muse on it, the more I think that the biggest problem with that Neurocomputing paper I just discussed is that the links between people that they attempt to graph were in many cases probably not real. As I said, for them a link counted if two people were in the same document, if they were in two documents involving the same lord within 15 years of each other, and if they were in two documents involving the same notary in that space of time.1 The first one is debatable but an obvious starting point: we may know that sometimes people listed together in charters weren’t all present together, but there is at least a contemporary association involved in that someone so recorded them.2 The second is much more dubious, as the two people linked might never have been in social contact at all; but as I said, at least they are in some sense in the same familia, though I really don’t think that’s what they were after; and I don’t think the third criterion has any value at all.
All the same, even the first one has problems if what you are after is a genuine social network. There’s argument about how witnesses are chosen for charters. In Rosenwein’s Cluny stuff family is clearly a big factor, and Duby did some of his most important work on the changes of society around the millennium by showing how that family presence changed in the charters there. In Brittany, Wendy Davies has shown that local reputation was primary; and not enough other people have considered it. In my stuff, I think that by and large people bring a couple of significant persons with them, whom we can sometimes pin down as neighbours, family, or frequent collaborators; and that is exactly the sort of thing I’d be hoping to achieve with social Network analysis of my material. But sometimes, I’m sure, the transaction is being held somewhere public, and a proportion, or sometimes all, of the witnesses are just people who were present for their own reasons. I’ve written about this with the court of Barcelona here, but it goes on at much tinier levels, gifts to abbeys on the same day and so on.3 So just loading them all up into a links table uncritically may be dodgy technique. On the other hand, the analysis might, ideally, sort the wheat from the chaff anyway, and maybe give us a way to test whether we can put people in the one box or the other…
Anyway, today I have been thinking less about this, and more about actual people and their relationships, and the main reason for this has been reading more of Julia Smith’s Europe After Rome. The great virtue of this book and the reason why it must have very difficult to write, as I’ve said elsewhere, is the way that Julia manages to step around the various difficult controversies that abound in medieval social history without avoiding the actual questions. So in her chapter about family and friends, rather than get bogged down in debates of agnatic versus cognatic, whether the kindreds in the lawcodes are real or imaginary, or whatever, Julia firstly (and throughout the book, without however losing the general thread of European unity) stresses variation, and then stresses how the categories blur. Some of your kindred are friends: kindred who aren’t friends aren’t much help and friends who aren’t kindred may be lots, but ultimately, “in Marc Bloch’s memorable phrase, ‘friends by blood’ were the kin who could be relied on.”4
The great contradiction of kindred, as a historical explanation of things, is of course that they don’t always behave properly. There are probably as many examples of family feuds as family solidarity in our sources, and we could guess as much from our own experience if there weren’t. Julia does cover this, but doesn’t make one point that I’ve long felt allowed us to get past that problem question, “do kin help or hinder?” It’s a problem question that people don’t often ask outright, but that lies in other questions, for example about law overtaking the threat of vengeance as a method of social control. My sidestep point is, that whatever else kindred are, they’re the people you, as a medieval man or woman, see most of all, and know best. Now it is true, as Julia says and others have said, that kindred counts for little a lot of the time until someone activates it by making demands on it, when people either help or don’t. Schrödinger’s support… But we can’t say whether family ties worked on one person or another differently: what we can say is that because they existed, they must have come into consideration.
Sometimes extraneous people in the kindred will have got dragged into things, of course. Hypothetical example: for some reason, you, though living in the Westfalian countryside on the family manse, have business in Frankfurt. It’s only now that anyone thinks of your mother-in-law’s Bavarian uncle with a share in a town-house there, who’s been out of touch for six years or so: he wouldn’t ordinarily be on the kingroup radar at all. But if you need a loan to do your business in town, you have the Bank of Dad right there at home, or if not, Bank of Bros. or whatever. Now the Bank may be closed, because you’re an irresponsible wastrel, or because of that thing you said about your stepmother at Pentecost; but it’s the first thing you think of because it’s at home, no? Kindred is a factor because of proximity as much as anything. It’s also a big part of identity that has to be defended at times, of course, but as a support network the reason the kindred come to the fore so much is at least partly, surely, because they are the people everyone knows best. Friends could come to be as close, too, but there isn’t a contradiction between these two groups if you see them like this. You appeal to people you know well: and the people you know best are likely your relatives. Easy, no? One more argument that perhaps we don’t have to have…
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, here cited from independently-paginated electronic preprint online at http://w3.grimm.univ-tlse2.fr/smash/jouve/papier/neurocomputing-final.pdf, last modified 11th February 2008 as of 21st May 2008, pp. 8-9.
2. See J. Jarrett, “Pathways of Power in late-Carolingian Catalonia”, unpublished Ph.D. thesis, University of London, 2005, pp. 31-34.
3. One pet example is F. 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), doc. no. 37, which is three donations to the nunnery of Sant Joan de les Abadesses in completely separate locations, done on the same day and in the same document; witnesses from all three or none at all could presumably be present, as the transactions were surely therefore done at the nunnery itself. On the other hand, contrast C. Devic, J. Vaissete, Histoire Générale de Languedoc avec les Notes et les Pièces Justificatives. Édition accompagnée de dissertations et actes nouvelles, contenant le recueil des inscriptions de la province antiques et du moyen âge, des planches, des cartes géographiques et des vues des monuments, rev. E. Mabille, E. Barry, E. Roschach & A. Molinier & ed. M. E. Dulaurier, Vol. V (Toulouse 1875, repr. Osnabrück 1973), Preuves: Chartes et Documents nos 193 & 194, where two cases in Elna, in the same court on the same day nevertheless brought forth different, but overlapping, sets of witnesses. For more references on witnessing, including to Rosenwein and Davies, see Jarrett, “Pathways”, pp. 54-59.
4. Julia M. H. Smith, Europe After Rome: a new cultural history 500-100 (Oxford 2005), pp. 83-114, quote at p. 98.