Charter-hacking I: if every property is square…

Some while ago now the estimable Cullen Chandler asked me something in comments here about the possibilities of mapping the different plots in `our’ charters together to learn something about tenure density, this being one of the very few things he hadn’t then so far tried with them.1 I had, and found it unsatisfactory for a bunch of reasons, most of which I gave in those comments, but since we’ve subsequently had conversations here about how exactly people thought about boundaries between their lands in this place and period, and shown how in some cases those boundaries do allow properties at least to be located, I thought that one of the two cases in the latest set of charters I’ve processed where I tried this would serve to demonstrate why this is, and isn’t, a useful thing to do, using an example from Pla de Fals, as the transactors would call it if they were speaking Catalan now, in Rajadell in Manresa.

Not much prospect of locating that on the ground now... Also a really good example of why you shouldn't use Google Maps as a road map!

Almost all the transaction charters from tenth-century Catalonia deal with land, and almost all of those give boundaries for that land. Those boundaries are expressed in terms of what the property lies next to. Sometimes this is a geographical feature like a river or a rock, but more usually it’s someone else’s land. It’s almost always four boundaries given, and though schemes differ for how those are described, as discussed here before, usually they are compass directions, starting in the east. The charter boundaries I expect therefore run, “et afrontat ipsa terra de orientis in vinea de Enegone, de meridie in vinea de Ricolfo, de occidente in vinea de Berane, de circi in in vinea de Ricolfo”, to pick a non-random [Edit: hypothetical] example. So, OK, it does look as if you ought to be able to play jigsaw with this kind of information, doesn’t it? If you share the usual assumption of the sources that all properties have four sides with one thing each on them, it looks even more so.

Unfortunately, on the rare occasions where we have measurements of plots of land, they are usually long rectangles, and OK, that only changes the game so much, but the possibility of disjointed bounds obviously arises. To make matters worse, you can’t assume you have all the charters, and in any case people might move or die without making any, so the names of the persons there in one document could all change by the time of the next, whereafter the property would be unrecognisable to us.2 So it’s only when you have lots of plots in the same document that you get a snapshot large enough to say anything from. The document I just quoted is one of those, a donation to Santa Cecília de Montserrat of 10 quarteradas of land in five plots.3 There’s trouble already, of course, as while the term quarterada might imply a square allotment (but probably doesn’t), if they’re in pairs then we must be dealing with rectangles even so. But never mind that! Just for the sake of the thing, let’s assume that all these plots are squares of the same size. What does that get us? Well, here is the information, schematically:


Click through for full-size version

Now, feel free to print this off and cut them out and try and arrange them or whatever, but some things will be apparent. Firstly, plot 1 must be separate from the rest, those neighbours don’t recur. Secondly, plot 5 cannot adjoin any of the others, once you work it out: Isarn would have to crop up again and he doesn’t. (This could be fudged with rectangular plots, but who’s to say whether they should be lengthened north-south or east-west? Let’s not.) In fact, I can only get three of these five onto a continuous layout, as follows:


Now, this is not, perhaps, analytically useless. What it tells us apart from anything else, is that at the bare minimum this neighbour Riculf, who is in fact probably one of the sellers but here holding separately from them, has three separate plots of land and another of heath (because his edge of plot 5 doesn’t join up to any on the sketch), and so he’s clearly more important than the other three transactors, which fits with him being named first. Berà, too, must have at least two; if the plots are all equal in size (IF) there’s no way he can have fewer. Ennegó, on the other hand, despite showing up as a neighbour twice, could plausibly just have the one plot. But of course the plots probably aren’t equal in size, and even if they are, we have two more here that won’t join up and that means that there are others between them that we just aren’t seeing, any of which could belong to any of these people. And in fact, of course, there’s nothing that means any of these plots have to join up, or that they must do so in the most economical way. But even if they did, Riculf might have had only these four plots, but Ennegó might have twenty! Just, not right here. This method cannot disprove that. In fact, it can’t disprove much, and neither can it prove anything really. I still keep trying it, occasionally, because such records do have this tempting jigsaw-like solvability, but every time I do the results really ought to put me off doing it again…

1. Cullen gives an extremely good round-up of what can be done with these documents as “Land and Social Networks in the Carolingian Spanish March” in Studies in Medieval and Renaissance History, Third Series Vol. 6 (Brooklyn 2009), pp. 1-33.

2. And if you think about it, that tells you that these documents were not expected to be relevant for terribly long. But, of course, after thirty years you’d be able to appeal to a different defence anyway!

3. The document in question is Ramon Ordeig i Mata (ed.), Catalunya Carolíngia IV: els comtats d’Osona i de Manresa, Memòries de la Secció històrico-arqueològica LIII (Barcelona 1999), doc. no. 1593, and its full boundaries run like this:

“Et est hec omnia in comitatum Minorisa, in castro Agello, qui vokant Plano de de [sic in orig.] Falcos, et afrontant ipsa I petia de vinea: de oriente et de meridie in vinea de Eldomar et de hocciduo in vinea de Sancta Cecilia et de circi in termine Falchos. Et ipsa alia pecia afrontat: de oriente in vinea de Wiskafredo et de meridie et de hocciduo in vinea de Richulfo et de circi in vinea Ennego. Et alia pecia afrontat: de oriente in vinea Ennego et de meridie in vinea de Richulfo et de occiduo in vinea de Bera et de circi in vinea Sancta Maria. Et alia pecia afrontat: de oriente in vinea de Richulfo et de meridie in ermo de Richulfo et de hocciduo in vinea de Bera et de circi in vinea de Richulfo. Et alia pecia afrontat: de oriente et de meridie in vinea Isarno et de hocciduo similiter et de circi in vinea de Richulfo.”

7 responses to “Charter-hacking I: if every property is square…

  1. That looks worryingly like a defective version of the kind of logic problem you get in puzzle books (defective because those are carefully set up so there is 1 and only 1 correct solution). But if you ever do think about doing this kind of thing seriously, get the ChartEx people on the case, because they are developing fancy tools to do this exact thing of boundary matching (but working for late medieval York, which has much better boundary clauses). I’ve just put up a blog-post on their sessions at IMC.

  2. Thanks to you, I abandoned the idea of mapping anything. I still think it would be great to see how neighborhoods were laid out, but such is pretty much impossible to achieve. It would seem a waste to have acquired copies of the editions of charters and not use them further, but lately I haven’t had many new ideas on what to do with them.

    • I think there might be some hope at least to judge size of people’s properties where there are a decent number of geographical bounds as well as proprietorial ones, though that gives us the issues about whether rivers and torrents still run where they ran then. That only really gets us at one sort of property, that out in the wilds, of course, so it could be checked against those properties that are measured in dextri, and from that get some kind of a sense of the range of property sizes. (This sounds very like a thing Jordi Bolòs and Victor Hurtado might already have done.) That might then furnish parameters for this kind of exercise… but it would still be very hard to make it probative, I think, alas.

  3. I’ve also tried this kind of reconstruction, some time ago, in a different document (I am sorry, I don’t remember witch one), only to find that at least some of the owners had to have more than one plot, at with point the whole question becomes highly speculative.

    But I found your example also interesting in another sense. The plot is described as a terra, but it’s sorrounded by vineas, so there’s probably a community going on here. I guess that’s what in other charters appears as village’s culture ?

    • Your question confused me, as in the bounds I quote in the footnote the word ‘terra’ doesn’t appear, but it’s in what I quote in the main text, I see. I’m afraid that’s not in the actual document, and I think I must have wrongly expanded my notes on it when I wrote the post draft. Sorry, I will amend. I don’t think it would that unusual to have lands surrounded by vines, though, and I would usually blame that on geography, a small piece of flat land surrounded by woods or slopes. I suppose it could be deliberate, sometimes, but whether that is the same thing as a cultura, which I think I have only seen belonging to individuals, not to villae, I am really not sure…

  4. Pingback: Charter-hacking II, From the Sources VIII, Feudal Transformations XVI: scribes who take us through the mutation documentaire | 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.