The geography of the place

Cunliffe: Estimated territory of the Dobunni

B. Cunliffe (1991): Estimated territory of the Dobunni

So, to the north, the Dobunni, with Corinium (Cirencester) their civitas capital. It’s supposed that they spread down into Somerset as far as the Mendips. For Professor Cunliffe, judging by Iron Age pottery and coin finds, the River Axe was the estimated southern boundary (1).

In that case Bath would have been in their region (rather than the Belgae) at least during the Iron Age, and Professor Cunliffe suggested that there might have been a ‘southern splinter group’  at a later date (Ptolemy was writing c. 150AD), like the one around Ilchester for the Durotriges. Bath would then possibly have been the ‘sub-capital’ though there is no evidence for this.

So far, so … probable/possible.

To the south west were the Dumnonii in Cornwall, Devon and west Somerset, with their civitas capital Isca Dumnoniorum (Exeter). Their northern boundary seems to have extended as far as the River Parrett.

To the south east are the Durotriges, civitas capital given as Dunium by Ptolemy which scholars seem happy to identify with Hod Hill, a hillfort about 4 miles north of Blandford Forum. Why this is preferred (if they were going to choose a hillfort) over Maiden Castle just a stone’s throw south west of Dorchester, I don’t know – perhaps there are more traces of sub-Roman civilian habitation;  but Ilchester – Lendiniae – was certainly part of the Durotrigan territory in the 4th century and was the most north westerly centre. Ham Hill and South Cadbury are both described as ‘Durotrigan’ at some point. Hengistbury Head on the south coast, where the Stour and the Avon reach the sea at Christchurch, was an important trading centre with a mint and is another suggested site tribal centre.

This leaves the Belgae. Using circles to show the (very rough) possible settlements of the four tribal groups, as discussed:

Circles don't accurately show likely geographical boundaries

Circles don’t accurately show likely geographical boundaries: ellipses would have been a bit better

We know where the River Parrett was, we know where Ilchester/Lendeniae was, we know where the Mendips and River Axe were. This leaves a kind of No Man’s Land in the centre. Did the Durotriges – or the Belgae – get to the Somerset coast in the 1st-2nd century AD? Ptolemy thought the Belgae. At any rate, that No Man’s Land is worth looking at next.

(1).B.Cunliffe, Iron Age Communities in Britain: an account of England, Scotland and Wales from the seventh century BC to the Roman conquest 3rd edn (1991), pp.170-5. Referenced on a website about Saxon Bath.

So, where were the Belgae? (1)

I’ve started this blog four times already and got stuck each time, so here’s the fifth attempt (Part 1). Thanks to ‘Dusty’ for the Christian Marx article on Ptolemy: the maths looks a bit complicated but the maps resemble the ones I drew up. A bit.

Belgae: There’s general agreement that they weren’t a separate tribe but a loose ‘confederation’ of tribes. If you look at the maps showing the tribes of Roman Gaul, there are no ‘Belgae’, just various tribes who occupied Gallia Belgica; so the likelihood is (I presume) that the ‘Belgae’ of Roman Britannia designated a grouping made up of neighbouring tribes from north east Gaul distinguished as being the last Celtic migration, which arrived during the first century BC. They were as a group culturally distinctive enough for archaeologists to identify certain artifacts, including coins, as ‘Belgic’.

Wacher's hypothetical civitas Durotrigum

J. Wacher (1974): hypothetical civitas Durotrigum, the Belgae to the east

Some authorities, like Wacher, confine the tribe in an area roughly corresponding to southern Hampshire, which tallies with Winchester – Venta Belgarum – being the capital of their civitas, the administrative area set up by the Romans. The Atrebates to the north, with their capital at Silchester – Calleva Atebatum – may have been closely connected, and also the Reg(i)ni to the east. All ‘Belgae’ in the loose sense.

According to this theory, the neighbouring Durotriges, centred on Dorchester (Durnovaria), hypothetically spread up through Dorset to reach the Somerset coast somewhere between the mouth of the River Parrett and Brean Down. We know that  Ilchester (Lindinis or Lendiniae) in east Somerset was held by the Durotriges in the 4th century and was possibly  the πόλις of a separate northern division of the civitas Durotrigum because inscriptions have been found at – of all places – Hadrian’s Wall which refer to ‘C(ivitas) Dur(o)tr(i)g(um) [L]endin(i)e(n)sis’. This indicates that men from Lendiniae helped with repair work on the wall. The inscription is dated AD 369:

Oppenheimer (2006): Tribes of northern Gaul

S. Oppenheimer (2006): Tribes of northern Gaul

Meanwhile, other authorities (including Oppenheimer) have the Belgae spreading north west through Dorset and reaching the Somerset coast, at that same point – that is, between the mouth of the Parrett and Brean Down. If that is correct, the civitas Belgarum might just have included Aque Calide (Bath), as Ptolemy said, as well as the mysterious ‘Iscalis’, a little further to the west of Bath.

However, I haven’t found (as yet) any primary evidence that they reached this far, other than the charts drawn up from Ptolemy’s coordinates; and Ptolemy could have been mistaken, or the surviving medieval manuscript sources could have mistranscribed. Ptolemy appears to be the sole evidence for their presence in north Somerset (correct me if I’m wrong), and the modern sources derive directly from him (wisely or unwisely).

The evidence seems to be stronger that Bath was a centre of the Dobunni who spread down through Gloucestershire as far south as the Mendips, with their civitas capital at Corinium (Cirencester). Bath was obviously an important centre for Celts as well as the Romans, Aquae Sulis being named after the Celtic deity Sulis (Minerva); but the administrative status of Bath isn’t known.

Having examined these options, I shall next look more closely at the geography of this part of Somerset in light of these alternative theories. Quite soon before I forget what I wanted to say.

Ploughing on …

A week’s holiday gave me time to read (at last) Michael Costen, The Origins of Somerset; and a few points were relevant to my argument (re arx Cynuit).

I had wondered whether it was realistic to suppose that an OldWelsh/Celtic/British name dating back, presumably, to the pre-Roman era could possibly survive in a modern place name. Or, how unlikely is it that Condate/Condatis has survived in the name Countisbury? I found Costen had a parallel example (p 46), though there remains an element of speculation there as well.

Celtic deity Brigantia

Celtic deity Brigantia

Starting from Ekwall’s derivation of British briga, meaning ‘a high place’,  connected with the Celtic goddess Brigantia, the suggestion is that these lie at the origin of the names Brean (Down) and nearby Brent (Knoll), the Celtic deity Brigantia being the goddess of high places. And Costen notes that the churches both at Brean and at nearby Chelvey are dedicated to St Bridget (or St Brigid of Kildare), a very unusual dedication for early parish churches in England: it is argued that St Brigid was a ‘Christianisation’ of Brigantia, who in turn was equated with the Roman Minerva.

The remains of a 4th-c. Romano-Celtic temple on Brean Down hold no evidence to show which deity or deities were worshipped there.

To quote Costen: ‘Both names [i.e. Brean and Brent] have been derived from the British briga, which meant ‘a high place ‘ (Ekwall 1960), as well as referring to the goddess [Brigantia] (Ross 1967, p 360).’

So we have the development:

1. Briga ‘a high place’ 2. Brigantia the goddess of high places (and of water?)  3. modern place names Brean (Domesday Brien) and Brent, speculatively deriving from briga/Brigantia (Roman Minerva).

And the parallel:

1. Condate ‘a confluence/watersmeet’ 2. Condatis the god of confluences 3. modern name Countisbury (Domesday Contesberie), speculatively deriving from condate/Condatis (Roman Mars).

Brean Down and Brean village

Brean Down and Brean village

Just as modern Countisbury is not exactly at the confluence, nor is the present village of Brean on the height of the Brean promontory (where the prehistoric remains are). Quite the contrary: it is very low, level with the adjacent sandy beach – no coastal cliffs here. The parish church, for example, is about 2 miles from the landward end of the promontory, about 2.4 miles from its centre.

The archaeology of Brean Down is very rich: Bronze Age remains, an Iron Age fort and Celtic field systems – hillfort, occupation, burials and a 4th-century Romano-Celtic temple.

The parish church in Countisbury village is, as the crow flies (there’s no direct route), about two thirds of a mile from the Iron Age Myrtleberry North hill-slope enclosure, and from the watersmeet which is directly overlooked by the enclosure set at the end of the hill spur.

Myrtleberry, the watersmeet, Countisbury

Myrtleberry, the watersmeet, Countisbury

I’m still trying to find out more about Myrtleberry but there doesn’t seem to have been a lot of archaeological work done here, and nothing that seems very useful. It’s broadly contemporary with the Iron Age rampart by Wind Hill, and seems to have been a settlement rather than defensive. Its outer enclosure is of the type associated with the keeping of livestock (vide

Perhaps Wind Hill, where there are no signs of any settlement/occupation, was the defensive fortress  serving the (only slightly) more accessible Myrtleberry – the name, incidentally  seems very modern, named after Myrtleberry Cleave. No surprise in that, since there has been very little needing a name on the site for millennia: any original name will surely have been long lost, if it ever had one.

If so, then as at Brean, at a much later date a settlement would have been made in the lower, less protected site, firstly because the threat of enemy attack no longer existed; and secondly because in more peaceful times the priority became communication with other parts of the region and this was much easier from the more open position.

I think my examination of arx Cynuit, Countisbury and related matters  are coming to an end. Brean Down and the surrounding area, as well as being in Somerset, seem well worth revisiting.


What’s in a name (2)

Looking at Asser’s Latin text  Vita Ælfredi regis, there are interesting points about the place names that he uses. Sometimes he quotes a  name ‘saxonice’, sometimes ‘latine’, sometimes ‘britannice’/Welsh, or one, two or all three of them in various permutations. So for Selwood Forest, as we’ve seen, he first refers to it by its Saxon name Seluudu,  then Latine autem sylva magna and Britannice Coit Maur. The British/Welsh form and the Latin have the same meaning – ‘great wood’, but the Saxon name is not (apparently) equivalent, since Selwood is supposed to derive from the word for ‘sallow’ or ‘willow’. It  seems rather strange that there should be an extensive forest which is predominantly sallow  (salix) – which usually grows near rivers and streams, but … So Asser, being a Welshman, simply translated the Welsh name into Latin but does not comment on the Saxon.

The river Wylye (Guilou) near Wilton, lined with pollarded willows

The river Wylye (Guilou) near Wilton, lined with pollarded willows

Anyway, it seems that there are only two Welsh names in the text which have no gloss or alternative. The meaning of one is clear: the river Guilou is the Wylye in Wiltshire, because ‘Wiltun’ is said to lie on its southern bank of the river which gave its name to both Wilton and Wiltshire.

The other example is arx Cynuit, where there is no clue as to where it is except that it was in ‘Domnania’ and from the context it was probably on the north coast of Devon/Somerset. Latin arx appears a number of times in Asser’s text and is usually translated as ‘stronghold’ or ‘fortress’. The Penguin edition of Keynes and Lapidge translates it as in front of the ‘stronghold at Cynuit [Countisbury]’ which is partly best guess (Countisbury) and partly a puzzle. ‘AT Cynuit? Cynuit seems to be somebody’s proper name, so ‘Cynuit’s stronghold’ seems to be the meaning.

Cynuit needs to be scrutinised more carefully. There are two possibilities, which may be the same one, but there’s still a bit of a mystery. Next time …


What’s in a name? (1)

Before moving on to study the sites of the various Viking raids further, I’m moving back to have another look at the name arx Cynuit.

This appears only (I think) in Asser’s Life of Alfred which had the fullest description of the event: the Danish ships had sailed de Demetica regione (Dyfed) ad Domnaniam (Devon, sort of; though the boundaries of Dumnonia at that time are not clear). The Anglo-Saxon Chronicle says it was on Defenascire (Ms A).

Screen Shot 2015-11-18 at 23.18.46

The map shows the territory of the Dumnonii in the Roman period, but what Asser might have meant by the name, we can only guess. The West Saxons gradually pushed westwards, but at what point did they begin to call part of their conquered lands ‘Somerset’ which had had no Roman name? Perhaps my book on the Origins of Somerset will make this clearer.

The Latin word arx is generally translated as ‘stronghold, fortress’ – which isn’t much of a clue since the whole north coast of Devon and Somerset had Iron Age forts which were probably used by the Saxons as they were well-placed for watching for Danish ships coming.

Cynuit is clearly of British/Welsh origin. The later form was Cynwyd which I’ve seen identified with a St Cunetus, about whom I find nothing. The shadowy historical Cynwyd Cynwydion seems to have been both a king and a saint – but his kingdom of Cynwydion is traditionally placed in the north, by the borders of Scotland. Or in the Chilterns. Cynwyd Cynwydion was apparently born c. 491, died in c. 519 and was the son of Cynfelyn (Cunobelinus) ap Arthwys. But the Latin form of his name is given as Conuvitus, otherwise Conowit. Unfortunately, all the online sources get mixed up with family history, tracing ancestors and a lot of imagination. Screen Shot 2015-11-07 at 21.59.02

Cunetus or Conuvitus? They don’t seem to be originally the same name. But Cynwyd seems close enough to Cynuit and we do know of the church and village of Llangynwyd in Bridgend; consulting The parish of Llangynwyd by TC Evans (bardic name Cadrawd), Evans identifies the patron saint with Cynwyd Cynwydion, son of Cynfelyn, son of Garthwys. This, he said, came from the Genealogy of the Cambro British Saints, though what that is I have yet to discover.

Cynwyd Cynwydion, then, lived in the 6th century and Asser was writing in the 9th. From all we know, the arx or fortress in Devonshire had no connection with this Cynwyd but coincidences are coincidences, so I’m including a mystic map with all the esoteric Lost Knowledge of the lay lines of Glastonbury. Just for my satisafaction:

Screen Shot 2015-11-07 at 21.56.19

The mysteriousness of this is that Llangynwyd stands at the apex of an isosceles triangle, and at its base angles are Countisbury and Porlock Weir. I think this is meaningless, but nevertheless – it is so.

And in my next I shall look at Asser’s use of place names.

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.

Danish raids (7 & 8) – Watchet

The seventh and eighth raids, in the tenth century, were both on hapless Watchet. For 988 the Anglo Saxon chronicle entry is brief:

MS C: 988 Her wæs Wecedport geheregod, 7 Goda se Defenisca þegen ofslagen 7 mycel wæl mid him. Her gefor Dunstan arcebisceop, 7 Æþelgar, bisceop feng æfter him to arcestole, 7 he lytle hwile æfter þæm lyfode, butan .i. gear 7 .iii. monþas


‘In this year Watchet was laid waste, and Goda, thegn of Devon was slain and many men with him. In this year Archbishop Dunstan died [on my birthday, as it happens], and Bishop Æthelgar succeeded to the archiepiscopal see, and he lived but a short while after him, just one year and three months.’

However, Ms E reads, a bit annoyingly:

St Dunstan (Feast day 19 May)

St Dunstan (Feast Day, 19 May)

Ms E: 987 Her Wecedport wes gehergod.

988 Her wæs Goda se Dæ<fe>nisca þægn ofslagen 7 mycel wæl mid him. 7 her Dunstan se halga arcebiscop forlet þis lif 7 geferde þet heofonlice, and Æðelgar biscop feng æfter him to arcebiscopstol, 7 he litle hwile æfter þam leofode, butan an gear 7 .iii. monðas.

The meaning is just about the same, except that in Ms E the attack on Watchet was in 987 and the slaying of Goda in 988, giving the impression the two events were not linked, unlike in MS C. However, the MS C reading does seem to make more sense since MS E gives no explanation, no why, no where, no how, for the death of Goda.

MS D follows MS C but I don’t have enough knowledge of the relationship between the different manuscripts to know whether MS D was from a source independent of C.

The final raid was in 997 where MS E records:

MS E: Her on þissum geare ferde se here abutan Defnanscire into Sæfern muðon 7 þær gehergodon ægðer on Cornwealum ge on Norðwalum 7 on Defenan 7 eodon him þa up æt Wecedport 7 þær mycel yfel wrohtan on bærnette 7 on manslihtum …

‘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 …’

Cornwall (south Wales) and Wales (north Wales) are fairly easily located.  Severn MouthDevon too, except that its eastern border is not clear. But the mouth of the Severn?

This Wikipedia map is interesting: it might suggest that the Danes sailed as far as Steep Holm, which they seem to have used as temporary quarters on previous occasions: a convenient place to land and loiter offshore, looking for opportunities to raid the mainland.

And as far as I can see, these were the recorded raids on the north Somerset coast during the ninth and tenth centuries.