29 September, 2011

Burgh Castle Roman fort: Cnobheresburg?

The remains of the Roman shore fort are by far the most impressive visible features on the site of Burgh Castle Roman fort, with three of the four walls and their massive bastions still standing to near full height (see my previous post for pictures). However, although the site may well have started with the Romans in the third century or so, it doesn’t appear to have ended with them. Archaeological evidence indicates occupation in the centuries after the end of Imperial Roman rule in Britain, and may even connect with some snippets of history.


A hoard of high quality glassware, of Roman and Germanic styles, was found buried in the fort (English Heritage) – see picture on the fort information board in my earlier post. The glassware is dated to the early fifth century, so must have been buried at some date after that (possibly considerably after, if the vessels were prized heirlooms kept for a long time).

The English Heritage listing record says that the field east of the Roman fort was the site of both a Roman military cemetery and an early Anglo-Saxon cremation cemetery, with several cremation burials discovered in 1756 (English Heritage). The report says “Most of the urns illustrated in the records are identifiable as having been of pagan Saxon type”. The report doesn’t suggest a date, but cremation cemeteries are typically associated with the fifth and sixth centuries. Stylistic dating, on the basis of changing fashions in the design of jewellery or other grave goods or the decorations on the cremation urns, can sometimes narrow the date range, but if the urns were excavated in 1756 that dating evidence may not have been recorded.

Inside the fort, an inhumation cemetery has been excavated in the south-west corner with the remains of a large timber building on the south side of the cemetery. The cemetery was radiocarbon-dated to the sixth to tenth century. In the north-east corner, traces of irregular oval timber structures were identified, associated with pottery dated to the mid-seventh to ninth century (English Heritage).

The south-west quadrant of the fort was later occupied by a Norman motte, constructed in the late 11th or early 12th century (English Heritage).


Bede mentions a site called Cnobheresburg, which was granted to the Irish monk Fursey by King Sigebert of the East Angles as the site for a monastery in around 633:

[...]Fursey set himself with all speed to build a monastery on a site given him by King Sigbert [...] This monastery was pleasantly situated in some woods close to the sea, within the area of a fortification that the English call Cnobheresburg, meaning Cnobhere’s Town.
--Bede, Ecclesiastical History Book III ch.19

The location of Cnobheresburg is uncertain. It was presumably in East Anglia, since King Sigebert was able to grant it to a monastery, and from Bede’s description it was some sort of fortification close to the coast. Bede’s phrase ‘within the area of’ may imply that it was a large fortification and the monastery did not occupy all of it. All this is consistent with Burgh Castle as a possible location for Cnobheresburg. The mid-seventh-century pottery and the sixth-to-tenth-century inhumation cemetery inside the fort are also indicative of occupation at the right sort of date, especially if the timber building beside the cemetery was a church.

However, there may have been other fortifications along the coast of East Anglia that would also fit Bede’s description and that have since been lost to coastal erosion (there was a Roman shore fort at Felixstowe, for example). Unless or until further evidence is found, the identification is open to interpretation.

St Foillan

Fursey’s brother St Foillan is said to have taken over as abbot of the monastery of Cnobheresburg and to have fled to Nivelles in what is now Belgium with the books, relics and remaining monks in 651 after Penda of Mercia invaded East Anglia and sacked the monastery (see Wikipedia article on St Foillan). If Cnobheresburg monastery was completely destroyed and abandoned in 651, this could be inconsistent with the date of the inhumation cemetery and the pottery at Burgh Castle, both of which suggest some form of occupation extending into the ninth or tenth century, well after St Foillan’s departure. However, it is possible that not all the monks left with St Foillan, or that others rebuilt the monastery after his departure (or that the story is unreliable; the source for it is a record from the monastery at Nivelles, and I do not know the date of the document or its reliability).


Nothing is known of Cnobhere. The second element of the personal name is ‘here’, meaning ‘army’, so it would be a suitable sort of name for a warlord, and a warlord is the sort of person one might expect to be associated with a fort, but this is pure speculation.

Since the name Cnobheresburg was established by the time the site was granted to Fursey, Cnobhere (whoever he was) presumably pre-dated the 630s. It is perhaps likely that he was long gone by then, or he might have objected to having his fort handed over to a monk, although he may have been a party to the transaction for all we know.

Church of St Peter and St Paul

The church of St Peter and St Paul stands just north of Burgh Castle Roman fort, an attractive small church with a round tower. The listed building record identifies the tower as late 11th century, with the rest of the church being later (British Listed Buildings). See earlier post for more details.

If the building near the cemetery inside the fort walls was an early church, the church on the current site may have replaced it at some point, perhaps when the fort site became unsuitable or was taken into use for some other purpose. Since the inhumation cemetery inside the walls has a date range in the sixth to tenth century, somewhere in or after the tenth century may be a likely time for the church to have moved. There are many possible reasons why the church might have relocated. I can think of at least three, and no doubt there are others:
  • The Viking raids of the ninth and tenth centuries, which hit East Anglia hard and may have led to abandonment of the monastery, with the church later rebuilt on a different site;

  • Norman takeover of the fortified area for the motte and bailey castle, requiring any church/monastery/inhabitants within the walls to move to a different site;

  • Collapse of the west wall of the fort – it is not known when the collapse happened, and it may have been sufficiently alarming an event to prompt relocation of the church in case the rest of the fort followed suit.


The various archaeological findings on the site of Burgh Castle Roman fort suggest a surprisingly long history:
  • third and fourth century use as a Roman military base;

  • hoard of expensive Roman- and German-style glassware, dated to the early fifth century and therefore buried at some (unknown) date after that;

  • ’pagan Saxon’ cremation cemetery on the site of the Roman military cemetery in the field east of the fort, date uncertain but probably somewhere around the fifth / sixth century;

  • traces of timber structures in the fort interior with pottery dated to the mid-seventh to ninth century;

  • inhumation cemetery in the fort interior, radiocarbon dated to the sixth to tenth century.

Between them, these take us almost up to the existing church tower at the nearby church of St Peter and St Paul (eleventh century). Burgh Castle may not have been inhabited continuously, but it seems reasonable to infer that it was in use at least on and off over several hundred years.

The monastery mentioned by Bede that was founded at Cnobheresburg in the early to mid seventh century is an obvious candidate for association with the inhumation cemetery and the timber structures and pottery inside the walls. The dates are reasonably consistent, assuming that Cnobheresburg did not cease to exist when St Foillan left in 651, and an inhumation cemetery is the sort of thing one would expect to find on a monastic site, especially if the associated timber building was a church. However, the identification of Burgh Castle with Cnobheresburg is not proven.

If Burgh Castle is the site of Cnobheresburg, it’s an attractive speculation (but no more than that) to associate Cnobhere, who gave his name to the site at some date before the 630s, with the hoard of glassware and/or with the early pagan Saxon cremation cemetery identified in the field east of the fort. This field was also the site of a Roman military cemetery. This does not necessarily indicate continuity of occupation. Roman cemeteries often had tombstones and mausolea, some of which may have remained standing for a long time. It is perfectly possible that the people who used the cremation cemetery arrived on the site after the fort had been abandoned, recognised the Roman cemetery as a burial place – or simply as unsuitable for agriculture because of the standing remains – and used it as an appropriate place to inter their own dead. In this scenario, one could imagine Cnobhere as the leader of a group of Anglian raiders-turned-settlers (like the later Norse), who took over the fort and made it his base, either having found it abandoned or having evicted the previous inhabitants. However, the re-use of the Roman cemetery may also be consistent with continuity of occupation. The Roman garrison could have handed over to a replacement garrison of federate troops, who brought a different funeral rite with them. Or possibly the Roman garrison or their descendants simply changed the funerary rite they chose to use, possibly to reflect a change in their perceived identity. In such a scenario one could imagine the eponymous Cnobhere as the last Roman commander (the Late Roman Army had Germanic officers in high command) or his descendant, the sort of person who could have owned a hoard of expensive Late Roman glassware, deciding to go into business for himself as a local ruler when orders and supplies stopped arriving from HQ, and signalling his independent status by changes in social customs, including (but not necessarily limited to) the preferred funeral rites. Something similar may have happened at Birdoswald on Hadrian’s Wall, with a change in building structure to a warlord-style timber hall on the site of the fort granary, as discussed in earlier posts here and here. I need hardly add that this is speculative.

Bede, Ecclesiastical history of the English people. Translated by Leo Sherley-Price. Penguin Classics, 1968, ISBN 0-14-044565-X.
British Listed Buildings, available online
English Heritage listing, available online


Rick said...

This is more than a bit apropos of the last discussion. What did it mean to be a 'Roman soldier' after the legions left? Did any auxiliary or federate troops identify themselves as such?

Certainly there was nothing new about Roman soldiers obeying their commander instead of a remote central government - just ask Marius' mules.

But assuming some body of troops has local continuity, how long does it stay Roman? My impulse is to say probably a good long time, because the recruits stand taller when the sergeant reminds them what outfit they belong to.

Admittedly this is projecting the experience of modern armies, but what with the legionary eagles and all that, the Roman army seems to have been 'modern' in this sense. The well was deep, and could have taken a long time to finally run dry.

Carla said...

Indeed. The answer as usual is that we don't really know. Gildas seems clear that he thought of 'Romans' as an external people who periodically arrived in Britain, tried to educate the feckless natives, and went away again. But we have no literary tradition from any surviving Roman or ex-Roman auxiliaries (if there were any) to compare it with. My impulse would be the same as yours, that tradition might have hung on (however highly modified by circumstance!) for a long time. There might be a straw in the wind in that the East Anglian royal genalogy has 'Cesar' as a founder figure immediately after Woden. One way of reading this is to see it as the descendants of auxiliary and/or federate troops still identifying themselves as 'Roman' in some sense. What sense that might have been is up for grabs. It may have meant nothing more than the possession of power, with 'Cesar' standing in roughly for 'Boss', but it may indicate something along the lines that you describe.

Anonymous said...

Ech, short excavation reports from the late forties. I would want to have a look at those with new eyes, and to see what John Blair says about the place in his Church in Anglo-Saxon Society, before relying too much on dating from those. And indeed it would not be a bad thing for me to do those things so if I do I'll report back...

Carla said...

Jonathan - if you do, I'll be interested to hear the results. My feeling is that unless there's something definitive like an inscription, it would be difficult either to prove that Burgh Castle is Cnobheresburg or to prove that it definitely isn't - probably 'consistent with but not proven', as so often. If I remember rightly, the 2005 edition of John Blair's book treated Cnobheresburg as somewhere in East Anglia, possibly Burgh Castle, but location not definitively identified.

Rick said...

(Belatedly) very interesting about Gildas' perception of the Romans as an external people who periodically arrived in Britain, tried to educate the feckless natives, and went away again.

The inclusion of 'Cesar' in the East Anglian line is also interesting. But as you say the emphasis might be more 'Great Ruler' than specifically Roman. As a title, Caesar really got around, 'Kaisar-i-Hind' being (to my mind) the most remarkable.

Carla said...

Well, that's my take on Gildas's account (De Excidio II ch. 15-20 mainly) - there are no doubt other interpretations.

It did get around; I suppose it was the name for the most powerful ruler in Europe for a good few centuries, so it's not surprising that the title should stick. Even so, its longevity is remarkable.

Rick said...

My hat is off to anyone who has actually plowed through Gildas, even in translation!

Carla said...

I always try to go back to the primary sources if possible (where there are any)

Rick said...

Belatedly noting that in the period most discussed here, there often as not *aren't* any primary sources!

Carla said...

Indeed, you have to widen the definition of 'primary' out by at least a century. Still, even an account written two or three centuries after the event is over a thousand years nearer than we are...

Rick said...