Featured post

Gender in the Merovingian World

Saturday, 19 February 2011

Northern Britain and the Fall of the Roman Empire

[The text - of my Anderson Lecture, delivered at the University of Saint Andrews, on 18 February has now been published (you can read it here) so I have removed it from the blog, except for the piccies. 

The argument in the article goes essentially as follows: we ought to think about Roman-barbarian relations in the north of Britain more in the context of those on the other frontiers, African as well as Rhenish; we ought to think about Roman-barbarian relations much less exclusively in terms of conflict and confrontation - the two worlds were inter-twined; on the Rhine frontier it is possible to suggest a rough three-band conceptualisation of barbarian polities, with those in the middle, intermediate band most affected by the imperial crisis around 400 AD; the North Sea should be seen as a cultural zone of two-way interaction and not just as a frontier across which one-way 'migration' or invasion took place; the 'Pictish' confederacies discussed by late Roman sources started at Hadrian's Wall, not the Forth - the rough frontier of the seventh-century Pictish kingdom; a military reorganisation of Britain took place in the reign of Magnus Maximus which involved a movement of regular troops away from the line of the Wall and the (probably only temporary, at least as initially envisaged) handing over of authority in the highland  zone to local military leaders and 'irregulars'; this affected the southern Pictish areas between the walls and perhaps areas further north too; by the middle of the fifth century it produced crisis in that area and a fragmentation of an earlier extensive but weak confedearcy into smaller competing units; it might be that the British on the wall expanded north and became a dominant power; in the period around 600, crucial changes led to a shift in the balance of power towards the English in the south-east (the Scottish east coast should be seen as in the North Sea cultural zone) and the Scots and other powers on the west coast; the British in the intra-mural zone might have been squeezed militarily from both sides and an English political identity might well have become more popular in local competitions for authority; shifts in these years produced change and perhaps political crises in the Pictish areas north of the Forth; it might be to this period that we should trace the creation of the Scottish and indeed other kingdoms, such as Bernicia; internal Pictish strife might explain why the Picts do not seem to be a very active player in the early seventh-century politics that are visible to us.]

Major Barbarian groupings
in the fourth century
The structure of Gildas' historical

Late fourth-century Roman
 military metalwork in Britain
(from a 1986 article by H.-W.

The Traprain Law Treasure
Traprain Law


  1. A North Sea zone seems a very fruitful idea. Was there also a Northern Italy/Southern Germany zone?
    I'd like to be convinced by your Magnus Maximus as the superb tyrant theory as I've thought this likely for other reasons, but I'd like to reread Gildas including his introduction, and all his possible sources before being completely convinced.
    On curious but probably unimportant questions: why do you think Maxen Gwledig has the name of Maxentius rather than Maximus as he should, and why is Aetius three times consul called Agitius which is the name of Aegidius, never consul? Had one scribe read too many chronicles and had too much wine? Also I think you made a slight mistake in that appendix:Faustus was the son of Vortigern according to "Nennius" not son of Maximus (not that he really existed).

  2. Thanks David. Yes, you are quite right about my slip between Maximus and Vortigern re. Faustus. mea culpa. That Faustus is allegedly Faustus of Riez so the whole thing is nonsense! Currently I think that I can make a decent argument that by the ninth century people were confused between Maximus and 'Vortigern'. I may on the other hand be less convinced that Vortigern is necessarily legendary...

    Agitius doesn't actually need to be a corruption of Aegidius; it's a plausible phonetic rendition of Aëtius (probably pron. something like 'Ahitsiu', whence modern Italian Ezio).

    As to Macsen and Maximus or Maxentius, I'd have to ask someone who knows more about Celtic philology than I do!

    By the sixth century at least there is an 'Alpine zone' to some extent at least - for sure - and I think that that would count earlier, but seas make for better connections than mountains.

  3. Higham has a plausible scenario of how the stories of Vortigern were elaborated in the competition between the spin doctors of Gwynned and Powys, and further developed by writers in Kent and Northumbria.
    Vortigern himself is merely the eponym of a small and unimportant Welsh kingdom adjacent to Powys.If a real founder he was perhaps Irish, as the name is quite common in Ireland, and Irish adventurers were busy at this time in parts of Wales, for example Dyfed and elsewhere in Wales and Western England.
    The name, which could be interpreted as over-ruler, though tigern does not mean tyrant, was taken up as an invented ancestor amonst the contradictory genealogies of the princes of Powys The famous cross at Llangollen has too, like the kings of Dyfed, an invented ancestry in Maxen.
    It was the ideologists of Gwynned who then, to blacken the repute of the Powys princes, identified him with the proud tyrant who invented the English over, and woe to the Britons.
    Then in Kent this character was used to tell the story of how Anscehis, the basis of Hengist, Oesc. Ossa etc, first tricked his way to power.
    And from Kent the stories came to Bede and used as another part of his tale deigratlng the Britons (not that Bede's antithesising the British gens to the Gens Anglorum was universally accepted).

  4. Sorry, this is a pretty prosaic comment, but please, please - how might I cite this? Are you publishing the text elsewhere? Much to digest here.

  5. @ David. As far as I can see, there is cetainly something in this although I don't know how far one can take that precise theory which seems a bit elaborate.

    That there were traditions about 'good' and 'bad' Vortigerns (that is to say, the 'bad' Vortigern who let the Saxons in, and the 'good' Vortigern who one claims as a dynastic founder, associated with Maximus) by the ninth century is demonstrable. The Venedotian 'bad Vortigern' probably goes back to the eighth since Bede had that version. Whether Bede got his version from Gwynedd or Kent is, as far as I can see it today, impossible to decide.

    By the ninth century a good argument can be made that 'Vortigern', whoever and if ever he was, had become confused with Magnus Maximus, perhaps because of a confusion of (Aurelius) Ambrosius and Ambrosius (Aurelianus). That would imply either a separate fifth-century Vortigern or at least a separate set of Vortigern stories located in the fifth century.

    I'm thinking aloud here and may well have forgotten something crucial but I wonder if the 'Kentish' origin legend comes from Gwynedd via Bede. Much of it originates in Gildas, but Gildas doesn't mention anything more specific than 'the east of the island' and the earliest Anglo-Saxon archaeology is in East Anglia rather than Kent.

    It's clear too that his name seems to get attached to different pre-existing stories by that date.

    Beyond all that I'm not sure how far we can go.

  6. Really interesting stuff here- a few comments/thoughts
    - I'm having increasing problems with the increasing use of the way 'Late Antiquity' is being used by British scholars writing about Britain (and I include myself in this). It seems to be essentially replacing the old Rahtzian 'sub-Roman' ie anything 5th century that isn't clearly Germanic. if we have to use this term in Britain we need to see it as encompassing rather than excluding the areas effected by AS settlement. One of the key narratives of Late Antiquity is surely the process of accomodation of Barbarians communities; to write about British 'Late Antiquity' without any Anglo-Saxons seems somewhat wrong-headed

    - Magnus Maximus: just to throw into the mix, he was of course the last emperor to mint gold solidi in Britain

    -SFBs: again something that has been playing on my mind recently: we're getting increasing evidence for SFBS or at least, buildings with sunken working hollows, from the Late Roman north-east, such as at the villas at Faverdale (Darlington), Quarry Farm, Ingleby Barwick (Teesside). Not exactly the same was classic AS grubenhaus, but certainly similar.

    - military/official metalwork between the Humber and the Tyne: I've been going through the PAS data for this recently and there is one interesting pattern. In general, the belt-sets are being found to the east of York (Ryedale; Vale of Pickering; Derwent valley; Wolds), but very few are coming up north of the North York Moors/Tees. However, crossbow brooches are being found both south and north of the Tees. Are we seeing differences in between the location of limitatenses and comitatenses?
    I think that we need to start unpicking the extent of variation in settlement and other activity across the northern frontier south of the Wall- there is a huge difference between Rydedale and Weardale for example. I'm not sure this is always taken into account.

    - military continuity. There appears to be an increasing orthodoxy that continuity of occupation on Roman forts in the 5th century is evidence for emerging warbands evolving out of pre-existing military formations, and that these are acting as the cores around which political continuity is structured in the norhtern frontier (particularly south of the Wall). Again, I wonder whether we are over egging the pudding. There is certainly significant evidence for continuity within forts in the 5th century - we're busy digging it up each year at Binchester!! However, we need to remember the massive bias towards excavating forts in the north. Interestingly, when we dig other sites, such as the the villa at Quarry Farm, Ingleby Barwick and the rural settlement at Newton Bewley (near Hartlepool), there is also continuity into the 5th century. At other sites, such as the small villa at Faverdale, there is occupation well into the 4th century, but there is significant ploughdamage to the layers above, which means we can't assume that the absence of ephemeral 5th century stratigraphy is real (though to be fair, we can't automatically assume it was there either). We need to be aware that there are other nodes of settlement in the north that might form cores around which power might coalesce in the 5th century and not limit ourselves to stories about forts.

    idle speculatio to avoid marking...

  7. Thanks Dave. These are all really useful points. This late Roman metalwork really needs unpicking. The Bohme 1986 map that I was working from (and which was still quite recent when I first started thinking these thoughts...) seems way out of date and/or misleading. I need to look at the PASS more.

    That there are cruciform brooches to the north doesn't surprise me and in a way fits my model. They were the classic rank badges and on the Rhine frontier the Alamannic leaders appear to have made their own imitations. You could see them as the sorts of things that 'federate' leaders of the type I am proposing might be interested in and use.

    That is interesting about SFBs. More grist to the mill. I think they are just part of a standard architectural repertoire in situations where stone building is less (or un-) feasible. Subdividing them into Saxon and other forms doesn't seem very helpful to me. Indeed counting posts and measuring door-widths seems like an absurdity in thinking about things like ethnic identity.

    You're right too about 'late antique Britain' and the Saxons. As you imply, using one to exclude the other just perpetuates the binary ethnic divide that so distorts our way of thinking about this period.

    That there was an Anglo-Saxon migration is indisputable. People need to accept that and live with it. But it needs explaining and it needs much more subtle thought. And whether it is that helpful in thinking about material cultural change in northern Britain is something that needs rethinking.

  8. A couple comments from me, too (fully admitting I'm part of Dave's 'growing orthodoxy' of continuity at fort sites):

    Finds distributions: it is definitely worth looking at PAS data, though the biases in search patterns impact on this, as you hinted. Lindsay Allason-Jones and I tried to map as much as we could to accompany Jon Coulston's paper in Finds from the Frontier, and I looked at crossbow brooches in my paper. A rather significant point (at least in terms of data to date) is the higher occurrence of the earlier types of developed crossbows - the types 1 and 2 of the late 3rd to early 4th c - than the more frequently encountered type 3/4 on the continent. There is also more suggestive evidence for local production that does not rely on the fabricae, again in contrast to continental data. In sum I would suggest the finds and the structural archaeology indicate a significant social and geographic difference from the imperial court in the northern frontier, in contrast to southern Britannia, and even more strongly to the Rhine and Danube frontiers. I think this relates to imperial ideas about the stability (not to be equated with the level of peace) of any given frontier zone.

    Environmental/pollen evidence: I'd also disagree with the use/presentation of the evidence suggesting reforestation. Certainly there is some, and making sure that you use more recent calibration models for C14 dates impacts on this, but so does the basic method and date of sample collection. From what I've been able to discern, and using the general palynological method, pollen cores from the region between Humber and Forth do demonstrate reforestation, occurring first in upland areas in the '5th century', but not really spreading into the lowlands until the '6th' and '7th centuries'. We also cannot assume it is all anthropogenic, as there is a wet-shift that begins in the later 4th or early 5th century that must have impacted on agricultural regimes.


Note: only a member of this blog may post a comment.