What does Gaimar say?

Tracking down an Anglo-Norman text of Geffrei Gaimar’s metrical Estorie des Engles that a) doesn’t cost a 3-figure sum and b) is out of copyright took a while. Prof. Short’s now definitive 2009 edition fell down on both these counts. However, his new background information and the critical apparatus were easier to access. I shan’t consider all the interesting questions such as Gaimar’s identity or which sources he used,  as only the dozen or so lines dealing with the ‘Battle of Cynuit’ event are of immediate importance.

Thomas Wright, antiquarian (1810-1877)

Thomas Wright, antiquarian (1810-1877)

The text I’ve used is Thomas Wright’s edition for the Caxton Society, published in 1850, which he based on Ms. R – British Library Royal 13 A xxi – the same manuscript used by Prof. Short. Wright was a highly respected scholar, archaeologist and antiquary in his day: I have left his text unemended, even where obvious corrections could have been suggested. Though old, it is has a latter day relevance to the current question …

Gaimar completed his ‘Estorie’ just before 1140, more than 150 years after Æthelweard wrote his Chronicon. So this was not merely 250 years after the event but also after a seismic régime change in England: no more marauding Danes but Norman French rulers. With the changed society came the poet Geffrei Gaimar, in the service of the Anglo-Norman nobility.

We gathered from the earlier sources that a Danish force landed somewhere in Devon, with 23 or 30 ships; it was led by a brother of Hynguar, possibly his brother Healfdene or a third, unnamed, brother. They may have besieged a band of Saxon thegns (who may have been led by Odda, the ealdorman of Devon) in a stronghold called arx Cynuit, naturally defended on three sides but vulnerable on the eastern side. The two sides engaged in battle and 800-odd Danes, including their leader, were killed. The leader’s Raven banner may have been captured. Finally, unlikely though it may seem, the Danes might actually have won, though the general opinion was that they lost quite badly …

As far as a connection with the west Devonshire coast is concerned, that’s all we had. So what does Gaimar’s version add to the story? I’m going to take the lines 3147 – 3158.

Context: The Danes have attacked Chippenham and caused King Alfred to flee into the wastelands of Somerset ‘to escape their bloody hands’; he had difficulty in gathering any of his forces together to fight with him, while the Danes wrought destruction and caused the people to flee. Nevertheless, even then Alfred managed to kill some of them: ‘Sovent oscist de lur asquanz.’ (l. 3146)

This is the point where the narrative introduces such details as allow us (just) to recognise the battle at arx Cynuit, immediately after Gaimar relates that Alfred has killed some of the Danes:

Un frere Iware e Haldene               A brother of Iware and Haldene  l. 3147

En fu oscis el bois de Pene;             Was killed in Pene Wood

Ubbe out à nun, un mal fesant;     His name was Ubba, a miscreant;

Does that imply that Ubba was one of those killed by Alfred? Is En an adverb of time or place? Or is it a pronoun (‘of them’)? Or just a phatic syllable to complete the octosyllabic line? Where was the bois de Pene? Were those critics right in associating Penselwood in Somerset – Pen Selwood – with King Alfred’s Tower and the mysterious Egbert’s Stone, Ecgbryhtesstan? In Gaimar’s narrative, Alfred  builds up his stronghold in Athelney and four weeks later, after Easter he rides back to ‘Ecbrichstane, in the east of Selwood’: Après Paske quatre semaine/Chevacha à Ecbrichstane/ Co est del hest de Selewode. (ll. 3165-3167).


Penselwood in Somerset appeared in the Domesday Book as ‘Penne’

Ubba has been killed in battle, but so far nothing has linked his whereabouts to west Devon. Gaimar goes on (the  translation doesn’t quite correspond line by line because of the word order):

Sur li firent hoge mult grant                  Over him the Danes built …      l. 3150

Li Daneis, quant l’ourent trové:            A great mound, when they found him:

Ubbelawe l’unt apelé.                            They called it Ubba’s Mound.

La hoge est en Deveneschire.               The burial mound is in Devonshire

So Ubba is named as a brother of Iware and Haldene, killed in battle and buried by the Danes (no mention of Odda here). The mention of ‘Ubbelawe’ is part of the west Devon legend, yet Gaimar’s story doesn’t connect it with the coast. The Danes don’t even arrive in ships: one would assume they were part of the Great Heathen Army already in the Midlands. Apart from the single mention of Ubba’s burial place, all other indications are that this battle took place further east, somewhere in Somerset; it isn’t absolutely clear that the Danes were fighting against Alfred with such forces as he had managed to muster: they still might have landed on the north Devon or Somerset coast but the text doesn’t suggest it.

De gent i out bien grant martyre,           There was great slaughter of men,

Huit cenz quarante en i morurent,        Eight hundred and forty died there,

Quinchald, feluns, perjures furent;         So what? They were criminals and oath-breakers;

Conquis i fu le gumfanun                       There was Ubba’s standard won

Ubbe, ke Raven out nun.                         That was called Raven.

These lines show the clearest evidence that Gaimar had  Anglo-Saxon source material, directly or indirectly. The 840 victims ties in with ASC, but also the name Ubbelawe has the Old English element hlāw, meaning mound or tumulus; and Raven or hræfn doesn’t seem to show any similarity with Gaimar’s Anglo-Norman vocabulary. Deveneschire is also Old English.

The greatest puzzle is geographical: there simply is no textual evidence in Gaimar that the event was connected anywhere near the mouth of the River Taw, or on the River Torridge. And yet, the tradition grew up there. Why? An interesting antiquarian point is Thomas Wright’s footnote to ll. 3148ff. on p.108:

“Gaimar has here added to the information given in the Saxon Chronicle, but it is impossible to say from what source he derived his additional matter. The Chronicle merely says that Ubba was killed in Devonshire [NB No extant version of the Chronicle names Ubba – OS]; the life of Alfred which goes under the name of Asser calls the place Cynuit; it is now called Kinnith or Kenny Castle, near Appledore, in Barnstaple Bay. I am informed that there was formerly a mound on the “Barrows” or sand beach at Appledore, which was called Hubbaston, Ubbaston, and Whibblestan; but it has been long swept away by the tides.”

Detail: Benjamin Donne's 1765 map of Devon showing 'Henny Castle olim K[e]nnith'

Detail: Benjamin Donn’s 1765 map of Devon showing ‘Henny Castle olim Kenwith’ (nearby is Godborough Castle)

Which simply tells us what we already knew: by 1850 the legend had been firmly established for centuries and had now become accepted wisdom. In The Historical Gazetteer of England’s Place-Names, Hubbastone is the only name noted that seemed to have any relevance, a place in the parish of Northam, in Shebbear Hundred. The earliest reference given is the spelling Hubberstone, noted in 1765, although the sources consulted go back to Saxon times.


On to the past (arx Cynuit) (1)

I’ve probably gone as far as I can go with Iscalis. My final thought was that it was somewhere on the coastal plain around Burnham-on-Sea, and south to the River Parrett. The heavily Romanised area, with many Iron Age sites,  seems a more likely location (to me) than Charterhouse (on top of the Mendips). This area would have been a centre of the Durotriges, rather than the Belgae as Ptolemy said (so it’s a bit like believing you have the correct exam answer – provided the examiner has made a slight mistake with the question.)

There are no facts, only interpretations.

“There are no facts, only interpretations.”

As Nietzsche said: “We hear only those questions for which we are in a position to find answers.” So I’m  now hearing a different question.

I advanced a little further with Iscalis  than in my study of  arx Cynuit where I was fairly confident it wasn’t at Countisbury, but couldn’t come up with a likely alternative. And speaking of arx Cynuit …  my dismissal of the suggestion that it was Castle Hill, near Beaford, was challenged, so I thought my next investigation could be around the west Devon coast, yes, back to arx Cynuit again.

This is not so much a ‘Well, we shall never know for sure’ question (even if we shall never know for sure) but: How did the local tradition develop and why did it persist for so long? What lay behind it? Where did it start? It MAY have started when the Danes landed near Appledore in 878, if they did. But did they (we shall never know for sure)? It’s the tradition I’ll be focusing on here, not the historical fact.

Method: If this were family history, you would start with modern times and work your way back, because you start with what you know and it gradually uncovers the unknown. But here we already know the ancient sources and we need to uncover the successive stages which led to the present. So I shall begin with Asser and the ASC (which neither support nor contradict the tradition).

[To be continued]

If not Countisbury – where? (1)

The map of the north Devon/Somerset coast (click to enlarge) marks the supposed sites of the eight raids made by the Danes during the ninth and tenth centuries (the circle on the right is only larger because the area is more vaguely located; the one on the left is the supposed site of arx Cynuit at Countisbury).

The Danish raids in the ninth and tenth centuries

The Danish raids in the ninth and tenth centuries

Points to note:

1. Only Countisbury is located above towering cliffs which offer no place for ships to land. Porlock is set on a wide bay and had a harbour; Carhampton and Watchet are located in the low-lying land between Exmoor and the Quantocks, and Watchet had a harbour; the mouth of the Parrett extends over flat land towards the Somerset levels; there was a harbour at Combwich and a few miles further up-river the port of Crandon Bridge.

There seems no realistic way the Danes could have rowed up the East Lyn river to Countisbury either: it was probably too narrow, was rocky and ‘uphill’ with waterfalls taking the river from higher up on Exmoor down to the coast.

2. Arx Cynuit is the only one not precisely located with an identifiable name. All the raids were recorded in the Anglo-Saxon Chronicle; only the raid on arx Cynuit took place during King Alfred’s reign, and is therefore the only one of the eight mentioned in Asser’s Life of King Alfred. He gives added details not included in the chronicle, but the name is a British one, not a Saxon one (Asser, of course, was a Welshman). The chronicle says only that it was in Devonshire.

3. Geographical features must change over a millennium, but although the Iron Age promontory fortress at Countisbury corresponds with Asser’s description (he said he had seen it), he also said that there was no water on the site but ordnance survey maps certainly show a convenient spring. Perhaps it sprang up more recently.

Wind Hill fort, showing the spring

Wind Hill fort, showing the spring

4. The chronicle and Asser agree that the raid in 878 was on some part of Devon, but where was the border between Devon and Somerset? Could it have been as far east as Watchet, for example? As far as the earliest raid at the mouth of the Parrett is concerned the chronicle says:

‘845/848 In this year ealdorman Eanwulf with the men of Somerset and Bishop Ealchstan and ealdorman Osric with the men of Dorset fought at the mouth of the Parrett with the Danish host; there was much slaughter and they were victorious.’

It’s fairly safe to assume that the area was in Somerset at that point, since there is no mention of the men of Devon participating. But then, also in the chronicle:

‘997 Ms E: In this year the [Danish] host sailed round Devonshire to the mouth of the Severn and there they laid waste both in Cornwall and Wales, and in Devon and they  landed at Watchet and there wrought much destruction, by burning and killing …

They landed at Watchet, yet there is no mention of Somerset; and:

‘988 Ms E: In this year Watchet was laid waste, and Goda, thegn of Devon was slain…’

This is the slightly dubious entry, but at Watchet it is, apparently, the thegn of Devon who was killed and still no mention of Somerset. We know the river Parrett marked the border at one time, though earlier than this; however,  perhaps this area south of the Parrett was still considered to be Devon – so arx Cynuit could be around here.

More follows, if the Person from Porlock doesn’t arrive on business.

Follow every lead …

… even if just for a bit of innocent amusement:

When Alfred left Athelney in May 878, he gerade with his company to Selwood Forest in Wiltshire to rally his army which assembled at Ecgbryhtesstane. They camped one night there and then moved on to Aecglea/Iglea where they camped a further night. The next day they marched to Eþandune where the Vikings were defeated in a decisive battle.

It is not certain exactly where Ecgbryhtesstane was (perhaps not Penselwood, perhaps not Kingston Deverill); nor Aecglea/Iglea (perhaps Iley Oak, west of Sutton Veny); nor Eþandune (presumed to be Edington on the ridge above the chalkland of Salisbury Plain). Eþandune was NOT, alas, Edington in Somerset, in the Polden Hills not far from Athelney.

But, supposing it was. Here’s another argument:

The derivation of Aecg-lea suggests Ockley/Oakley. There is no Ockley in Wiltshire, though ‘Iley Oak’ was the meeting place of the Hundreds of Heytesbury and Warminster, south of Edington in Wiltshire …

Achelai Alwy tenuit T[empore] R[egis] E[dwardi] - Domesday

Achelai Alwy tenuit T[empore] R[egis] E[dwardi] – Domesday

But returning to Edington in Somerset, there is indeed a small area which bears the name ‘Oakley’ (Oakley Brook, Oakley Farm) which is recorded in the Domesday Book as Achelai: not a large place since the tax paid was ‘0’ and the number of households  ‘0’. It was to the south of Ilchester, near the A37, as the crow flies 14 miles from the Somerset Edington.

The remaining vestiges of Achelai: not much in Domesday, not much now

The remaining vestiges of Achelai: not much in Domesday, not much now

So, here’s the hypothetical reconstruction: Alfred rode to Selwood Forest to rally the Hampshire and Wiltshire fyrds at Ecgbryhtesstane but then, then, he backtracked to Somerset once more, pitched camp at Oakley (not Iley Oak), and next morning marched through Ilchester and Somerton to Edington in the Poldens where Guthrum was decisively defeated and subsequently baptised a Christian, along with 30 of the best men of his army.

Oakley, Aller, Athelney and Edington in Somerset

Oakley, Aller, Athelney and Edington in Somerset

And where were they baptised? Well, at Aller, just six miles south of Edington and a mere 3-4miles east of Athelney. And after eight days the baptismal robes were removed at Wedmore (where by some traditions Guthrum signed a  treaty)  – and where is Wedmore? Just six miles to the north of Edington.

Oakley, Edington, Aller, Wedmore.  The locations of Aecglea and Eþandune are still doubtful. But Aller and Wedmore are as certain as may be – Asser §56:

Nam post hebdomadas Godrum, paganorum rex, cum triginta electissimis de exercitu suo viris, ad Aelfred regem prope Aethelingaeg in loco, qui dicitur Alre [Aller], prevenit. Quem Aelfred rex in filium adoptionis sibi suscipiens, de fonte sacro baptismatis elevavit. Cuius chrismatis solutio octavo die in villa regia, quae dicitur Waedmor [Wedmore], fuit. Qui, postquam baptizatus fuit, duodecim noctibus cum rege mansit. Cui rex cum suis omnibus multa et optima beneficia largiter dedit.

And why would Alfred ride all the way over to Selwood Forest and then come back again to fight a battle in the Poldens? Well, one might equally ask why he would fight a battle in Wiltshire and then bring the defeated Guthrum back to Aller, 36 miles away, to be baptised? Why not at Chippenham, twelve miles away, a royal vill? Or Trowbridge? Warminster? Even Winchester which was almost the same distance away? Why Aller, unless the two protagonists were conveniently nearby?

These are deep waters, Watson …

[To be continued]

The rains came …

But returning to our island, below is a Met Office satellite picture (much enlarged, placenames added) taken on 23 January 2014. The southern Somerset Levels resemble the Great Lakes. Channel 4’s Time Team made excavations (about which, more later) on both sections of the island site. On the eastern section they found subterranean traces of the medieval Athelney Abbey (post-Alfred); and on the western section the remains of metal-working, considered to be Saxon.

The initial ‘A’ of Athelney is just below the abbey site, to the left of it is the fortress site and to the left again is the causeway leading to East Lyng. More than 1,000 years after Alfred sought refuge in the levels they become as watery as ever – now from the flooding of the rivers over the low-lying land, but before that by the receding of the sea.

The Jan 2014 floods

The Jan 2014 floods