Seminars CXXXI & CXXXII: searching the margins of Anglo-Norman England

I’m sorry, did I say ‘the next week‘? Apparently I meant ‘the next month’. Wow, that’s never happened to the blog before, I do apologise. I have, for what it’s worth, been trying to secure the short-term future of my sanity and balance by actually seeing some bands, the medium-term future of history at my college by marking admissions tests and the long-term future of your humble blogger by offering myself as employee to people, and of course if anything comes of that you will hear in due course. But in the meantime, this is the only evening at home I shall have for a while even now so I should put some blog up, and that blog should be seminar reports. Given how immensely behind I am with these, I will skip one that I’ve no useful expertise with, Robert Hoyland speaking to the Earlier Middle Ages Seminar at London’s Institute of Historical Research on 7th March 2012 to the title, “Theophilus of Edessa and the Historiography of the 7th-8th-Century Near East”—sorry, Byzantinists and early Islamists—because although it had certain detective elements to it as Professor Hoyland was on the trail of a lost source, I knew almost none of the names involved and don’t read any of the languages and I have no means of evaluating how significant what he was saying was. Cool stemma diagram though! If you’re eager to know more I can revisit it, but otherwise I’ll move on to stuff I do have opinions about, those being my erstwhile colleague Emma Cavell, addressing the Late Medieval Seminar at the I. H. R. on the 9th March with the title, “Did Women Cause The Fall of Native Wales?” and Stephen Baxter, Chris Lewis and Duncan Probert addressing the Earlier Middle Ages Seminar there on the 14th March with the title, “Profile of a Doomed Élite: the structure of English landed society in 1066”.

Clun Castle: capital of intrigue circa 1281!

Emma’s area of expertise is the March of Wales in the time of the Norman kings of England, and the Marcher lords who made their fortunes there, and even more specifically, the women in the Marcher lords’ families.¹ What she had for us on this occasion was that, while becoming yet more expert on these people, she’d come across a number of letters to such women, Maud de Braose wife of Roger Mortimer (the first one) particularly, from local lords on the other side of the frontier, and what these letters were reporting was nothing less than military intelligence about the composition and motions of the army of Prince Llewellyn of Wales. This comes from a time in 1281 when Roger was out on campaign on that frontier because Llewellyn had just fortified it. Maud, meanwhile, was at Clun Castle and apparently running the command post, this information presumably going back out to Roger and the lords getting information back and so on. Unlike my period, we only have the letters in here, whereas I’m more used to having letters out, but nonetheless there she was at the centre of a fifth-column spy ring and she wasn’t the only one; Howys leStrange (good name madam!) is apparently reported commanding the defence of Welshpool when Llewellyn attacked, and the text that tells us this also tells us that while she was doing that she took care to hide all the documents in the castle. Yeah, I’ll bet! That is a relatively rare mention of such activity in the chronicles of the time, but the letters make it clear that there is more to tell. Emma has been working this up since, including details of a juicy family conspiracy between these groups, and I believe it’s now in some kind of print process, so you may be able to find out more soon!

Now, I thought this was pretty exciting myself, spies, spymistresses, treacherous compacts made on battlefields between mutually-cautious relatives and the last-but-one flash of Welsh independence briefly burning bright in the pan, but Emma got quite a grilling from Judith Bennett, no less, about the role her title had given the women and whether it was fair, and whether this evidence told us anything the Paston Letters don’t, and various others likewise sang up saying such behaviour wasn’t unusual in their area. I’ve had these questions (the ‘it’s not unusual’ sort) myself and I’m never sure what they’re supposed to achieve other than perhaps to imply that the questioner’s area of expertise is somehow more developed than the speaker’s.² Well, great, but the paper isn’t about that area, so, can we talk about what was actually said perhaps? Anyway, you will see from my description that I thought it was good stuff and maybe you also think it sounds like that too.

The manuscript of Greater Domesday

Then the next week I was back in the same building to hear about a different native population being subjugated by the Norman yoke (MAYBE), slightly earlier, as Stephen Baxter and his team told us about the first results from the Profile of a Doomed Elite project that he is running at King’s College London. What they are trying to do is to properly, scientifically, electronically and most of all accurately count, identify, locate and describe the landholders of England in 1066 and work out what had happened to them in 1086 via the magic window of Domesday Book. This has, of course, been attempted before, but never so thoroughly, and in work that Stephen described as “riddled with mistakes” and “methodologically flawed”.³ There is a lot to do here, and it’s not easy: starting estimates are 27,000 pieces of property assigned to 1200 different personal names, only a very few of whom have titles and very many of whom might therefore be people with the same names. I am very familiar with that problem, as of course are they from the PASE Domesday project that Stephen also ran, and the digital solutions they were working out here were consequently of a lot of interest to me.⁴ They involve combining maps and tables of data, frequencies of names, their predecessors on the estates, their wealth and using all this stuff to arrive, not at solid identifications, but at confidence measures for possible identifications. I like this a lot because it avoids the two common problems with prosopographical databases where identification is uncertain, of either the database format forcing the user to decide where someone belongs before they have the full picture of the database completed, thus not actually allowing that database to help with the identification, or else that format not giving a way of assessing or making links at all, so that the identification always has to be done real-time by eye, and therefore not necessarily with consistency.⁵ Better still, it does not resolve this problem by having the computer do black-box identifications whose basis isn’t flexible. When our data is as variable as the Domesday data, pretending that we won’t sometimes get garbage out when we put it in is just unrealistic. This solution lets one measure how garbagey each result is, and as Stephen explained it’s solid enough to start doing statistics with, because adequate statistical methods can factor in things like confidence and make them part of the measurements. This should allow them to ask questions like: how long is the tail of small free independent English landholders left after the big guys whom we know lose out? how much of English wealth is actually peasant-held? How does the Church compare, how do women do compared to men? (A preliminary take at that last from 1066 suggests, apparently, that ninety per cent of lay wealth then was held by men and half the rest by Queen Edith! Lucky her?)

After Stephen had talked us through that in taut and dynamic style, Duncan and Chris filled in some texture. Duncan talked about the greater accuracy of micro-studies in this method because of small landholders pretty certainly not holding anywhere else so we see all their stuff; but most of a nation’s worth of micro-studies and a big enough computer of course equals one very detailed macro-study, so it will all add up. Chris, on the other hand, focused on the big identifiable people, not least Archbishop Stigand of Canterbury, who now emerges as the third largest landholder in England tempus rex Eadwardi (I presume after Earl Harold and the king?), and actually least, weirdly, Harold’s sister Gunnhild, although she was a professed nun apparently living on her own estates; nonetheless, they were only 30 hides, which makes her the smallest landholder the team can place in a secure family connection. Chris also showed us Danes settled in Wessex (described as such), mixed-name families, northern king’s thegns taking service with Norman earls and many other possibilities. I’m sure some of these have been spotted before, probably largely by Ann Williams, but of course they’re going to catch all that are reasonably catchable through this project and there seems no question that that will give them new things to say about how Normans became Anglo-Normans, how English dealt with or were dealt with by Normans and how that varied from place to place. There were questions, all the same, including a marvellously Heisenbergian one by Susan Reynolds pointing out that since the king’s commissioners themselves didn’t know the answers they were soliciting from the jurors at the inquests that made up the Domesday data, the enquiry was itself presumably changing the data; but, there wasn’t anything that the team didn’t have some means of testing for and trapping via the statistical analyses. It can’t be rock-solid accurate, of course, it just can’t, because of factors like Susan’s but also because of the variable data quality and so on, and also of course because of the large chunks of England not included in Domesday Book, but it might be as close as we can get…

---

1. For example the widows, as studied in Emma Cavell, “Aristocratic widows and the medieval Welsh frontier: The Shropshire evidence” in Transactions of the Royal Historical Society 6th Series Vol. 17 (Cambridge 2007), pp. 57-82.

2. One would like, generously, to suppose that it was to offer scope for Tom Jones filks, but if so no-one grasped that nettle.

3. I guess that by this was implied Robin Fleming’s Kings and Lords in Conquest England (Cambridge 1991), not least because esteemed commentator Levi warned us some time ago that Stephen makes criticisms of this work in his The Earls of Mercia: Lordship and Power in Late Anglo-Saxon England (Oxford 2007) but I don’t know if Stephen would also have meant Ann Williams’s The English and the Norman Conquest (Woodbridge 1995).

4. Cf. Chris Lewis, “Joining the Dots: a methodology for identifying the English in Domesday Book” in Katherine Keats-Rohan (ed.), Family Trees and the Roots of Politics. The prosopography of Britain and France from the tenth to the twelfth century (Woodbridge 1997), pp. 69-87; Jonathan Jarrett, Rulers and Ruled in Frontier Catalonia 880-1010: pathways of power (Woodbridge 2010), p. 19.

5. I have actually spoken in public about this, at the Digital Diplomatics conference in Naples that I blogged some time ago, and my paper there, “Poor Tools to Think With: the human space in digital diplomatics” is, I believe, still under review for possible publication at this time, though it’s possible that it’s in press and no-one’s told me. Now I’ve said this, proofs will probably arrive in my INBOX just as I head out of town this week…