h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

South Pembrokeshire is a part of Wales where Welsh is not generally spoken and many of the place names are English. It is known as "Little England Beyond Wales"

The unusual thing about south Pembrokeshire is that it is not physically connected in any way to England but lies in the south west of Wales, far from the border.

Apparently, this English colony began in the 12th century, after the Norman invasion of Wales, when many English and Flemish speaking settlers arrived in south Pembrokeshire.

“The Chronicle of the Princes (Brut y Tywysogyon) states a colony was established in 1105 when Henry I allowed a number of Flemings from modern-day Belgium to settle in the area around Haverfordwest. They were later joined by English settlers - the Flemish and English languages were similar at the time.

“This led to the extinction of Welsh in the area, and a legacy of aggression towards the language which has only softened in recent times.”

http://www.bbc.co.uk/wales/storyofwelsh/content/thenormans.shtml

In this map of part of south Pembrokeshire you can clearly see the English place names. Hopefully, this long string works when posted. If you scroll north and east on the map, you can see where the Welsh names begin to reappear.

http://www.multimap.com/map/browse.cgi?client=public&X=190000&Y=215000&width=500&height=300&gride=&gridn=&srec=0&coordsys=gb&db=&addr1=&addr2=&addr3=&pc=&advanced=&local=&localinfosel=&kw=&inmap=&table=&ovtype=&keepicon=&zm=0&scale=100000&down.x=186&down.y=4

You must agree the south Pembrokeshire place names are very English, and that this seems to point to an aggressive removal of the Welsh language, and existing place names, by the English (and Flemish) settlers.

What I have read so far does not make clear whether the Welsh were driven out of south Pembrokeshire, or whether the settlers forced them to use English and changed existing Welsh place names.

What I find interesting is that this cultural change, shown so clearly in south Pembrokeshire’s place names, happened in a Celtic region in the 12th century, long after the Saxons arrived, and in a period when we regard national and cultural identity as having become more settled in Britain.

South Pembrokeshire raises the question of whether Devon was fully Anglicized in the earlier Anglo-Saxon period, or whether the full force of the process actually occurred after the Norman Conquest. Accounts of post-Norman Celtic speech, known lost place names, and the apparent forced alteration of Celtic church dedications by Bishop Grandisson of Exeter in the 14th century suggest Devon could have undergone a similar process to this part of Wales during the medieval period.

Here, the author Terry Faull describes the medieval church’s opposition to Devon’s Celtic religious traditions.

Yes, I have posted this before! Yawn... sorry.

“A further example of the continuation of this central control may be see in the 12th and 13th centuries when a number of Devon church dedications seemed to have lost their link with their Celtic founders and been replaced by more main stream patrons (the folklorist Sabine Baring -Gould even claims that the 14th century Bishop Grandisson of Exeter was still positively anti-Celtic and took every opportunity of renaming churches away from original links with their Celtic founders).”

http://www.holywells.com/html/opposition_to_holy_wells.html

Terry Faull describes a process of change stretching through the 12th to 14th centuries. If Baring-Gould is correct and Bishop Grandisson was “positively anti-Celtic”—as late as the 1300s—then you must suppose there was still something Celtic for him to be against.

The 14th century is roughly 500 years after Devon is assumed to have become completely English.

In my opinion, south Pembrokeshire shows that Anglicization was still happening in Celtic Britain—with loss of place names and native language—hundreds of years after the first Germanic settlers arrived, and that the transformation of this part of Wales appears to have been as complete as in areas of England settled centuries earlier.

There is no doubt the same thing happened to Devon—and north-east Cornwall too—the question is when.

ExeValleyBoy,

I was aware of the accounts of English and Flemish settlers in South Pembrokeshire, and of the distinct change in the place-name type in this area, which was associated with the name ‘Little England beyond Wales’. I was therefore very interested to see the results of the Capelli et al Y-Chromosome survey, which included Haverfordwest as one of its sampling towns. If there had been significant numbers of English or Flemish descendents still occupying the town and surrounding areas, one would have expected this to be reflected in the Y-Chromosome signature for this area. However, Haverfordwest came out as being about 95% Indigenous and 5% Germanic, i.e. virtually the same as the Indigenous datum used by the UCL team (Castlerea and the Basque Country).

I then looked to see how sensitive the analysis was to other known cases of influxes of people from outside the sampling area. Another of the Welsh sampling sites was Llanidloes in Mid Wales, where the population of the area was reputedly massively boosted by migrant lead miners from Derbyshire at the beginning of the 19th century, when the rich Van Lead Mines were coming into full scale production. Even I was quite surprised to find that the Y-profile for Llanidloes, far from being typical for Wales, was actually almost indistinguishable from the Derbyshire/East Midlands Y-profile obtained from sampling sites at Ashbourne (Weale et al) and Southwell (Capelli et al), at 55% Indigenous and 45% Germanic.

Therefore I think that, concerning Haverfordwest, we can safely conclude that either:-

1. The English/Flemish settlement didn’t happen (unlikely in the light of the place-name evidence).
2. The English/Flemish settlement was very small in terms of numbers, but had an inordinately large influence on the place-names of the area.
3. The English and Flemish migrants left the area at some later date, leaving the Welsh in an area with predominantly English place-names.

Although I cannot recall where I saw it, I do remember reading that many of the English and Flemish migrants did in fact leave the area at a later date. Although the area now appears to be almost entirely inhabited by the descendents of Indigenous Britons (Welsh), this is still an almost exclusively English-speaking region. Yet only 20 miles or so northeast of here, in the Newcastle Emlyn region, Welsh is still the first language.

Whether any direct parallels can be drawn with the situation in Devon, it is difficult to say. Certainly the incidence of Brythonic place-names in South Pembrokeshire is somewhat lower than it is in Devon. Interestingly, at the northwest border of the English place-names in Pembrokeshire, where the Welsh names become the norm again, we find the name ‘Penycwm’, which seems to be an exact equivalent of ‘Pennycombe’, a farm southwest of Exeter. This is also the root of the Plymouth district name ‘Pennycomequick’.

I don’t think there is evidence for a mass change in Devon place-names following the Norman Conquest and later in the 14th century. Certainly a number of the town names can be traced back to pre-Norman times. However, most Devon place-names of this period were small farms and hamlets, many of which were ‘combe’ names. A fair percentage of these were (and still are) of the simplex form ‘Coombe’, with no specifier, although the majority have prefix specifiers, which are mostly English. Margaret Gelling in her book “The Landscape of Place-Names” states that many ‘combe’ names were simplex in form prior to the Norman Conquest, but given affixes soon afterwards, and appear in the later form in the Domesday Survey. Such a process could certainly account for the massive number of such names in Devon.

I have recently completed a survey of Devon ‘tre-‘ place-names, which appear on the OS 1:25,000 scale maps, together with known examples of this type that are now lost, i.e. those which appeared on early maps of Devon, but did not survive to the present time. One such ‘tre-‘ name of this type was ‘Trehoro’ in North West Devon. Another is ‘Awliscombe Tremenet’, which was apparently the old name of Awliscombe.

Even in recent centuries, names change in subtle ways. On the original OS maps of 1809, there appears the name ‘Buryet’ in North Devon (just north of High Bickington). On current OS maps the same name appears in the form ‘Burriott’ (in ‘Burriott Barton’) and as ‘Boreat’ (in ‘Boreat Moor’). The element ‘-yet’ in the 1809 version ‘Buryet’ could be from the Brythonic (Cornish) word ‘yet’ meaning a gateway, but who would recognise this possible element in either of the current forms?

If Baring-Gould was correct and Bishop Grandisson did come down heavily against things Celtic in the 14th century, could this have been influential in the remnants of the Brythonic language dying out at that time, as per Risdon’s account?

The situation in North East Cornwall is a bit of an enigma, as Cornish place-names here seem to be somewhat rarer than they are in neighbouring North West Devon, where Richard Coates (Celtic Voices English Places) indicates a fairly high concentration of attested Brythonic names (Trellick, Rosedown, Clovelly, Portledge and Cartland). My own ‘tre-‘ name survey also indicates a cluster of seven such names (including Trellick) in close proximity to the Cornish border in North West Devon. This would make North East Cornwall an enclave of English place-names surrounded by areas where Brythonic/Cornish names are somewhat more frequent.

Plymouth Exile,

The south Pembrokeshire genetic data do not seem to make a lot of sense, as something called the “Landsker line”, a group of castles and other fortifications, historically separated the English and Welsh speaking areas. This strongly suggests the Welsh were largely driven out of the area and kept out by the military. They may have returned later on, but this does not explain why the area held on to its ‘English’ identity and language so strongly and for so long.

“A line of over 50 castles and strongholds was built by the Normans and Flemish to protect south Pembrokeshire from the indigenous Welsh, who had been forced to move to the hilly country in the north of the county.”

http://www.bbc.co.uk/legacies/immig_emig/wales/w_sw/article_2.shtml

In the same article, geography professor Howard Carter is quoted as saying;

“If you look at the 'Brut y Tywysogyon' - the Chronicle of the Welsh Princes - it records 'a certain folk of strange origins and customs occupy the whole cantref of Rhôs the estuary of the river Cleddau, and drove away all the inhabitants of the land'. In a way you could almost call it a process of ethnic cleansing.”

There also appears to be strong Viking influence on the area too, from a somewhat earlier date to the Norman-Flemish incursion. A number of the ‘English’ place names apparently have Old Norse roots, including the Milford of Milford Haven.

This is taken from an article originally published in Pembrokeshire Life;

“Carew, Stackpole, Angle, Dale and Nash all bear traces of names of Scandinavian origin. Fishguard meant a 'fish yard', a place for catching fish, originally from the Scandinavian 'fiskr' and 'garthr'. It was recorded as 'Fissigart' as early as 1200. The Welsh name for the town, Abergwaun, has an equally long ancestry.”

http://www.benybont.co.uk/other/ess-art/wotname.htm

The Viking influence is dated from the 10th to the 11th century. The place name evidence suggests some degree of settlement must have taken place. Whether the decline, or the start of the decline in Welsh culture dates back as early as this Viking period is not made clear in either article.

This Wikipedia map (so I cannot vouch for its accuracy) shows a pocket of Old Norse speech established in south Pembrokeshire in the early 10th century.

http://en.wikipedia.org/wiki/Image:Old_norse%2C_ca_900.PNG

I also see, in addition to the Viking place names mentioned in the Pembrokeshire Life article there is a Lower Freystrop and Freystrop Cross just south of Haverfordwest. ‘Trop’ is a common element in Scandinavian place names, so these too seem to be Viking.

Given, all of this, I find it remarkable that there is so little trace of Germanic settlement in the DNA of Haverfordwest’s local population. However, I noticed that Haverfordwest is quite close (some 10 miles) to the resumption of the Welsh place names in the east. Testing more towards the core of the English speaking area, to the south and west, may have revealed more Germanic ancestry. I appreciate time and money limit the scope of such studies, but I think more would have been learned if the researchers had extensively studied just one region in depth, town by town, rather than trying to cover the whole nation using only a few sites

“As in England the DNA typically found in Wales either had an Ancient Briton (Celtic) signature or had the signature of the 'invading' populations (Angles, Saxons and Danish Vikings). Large parts of Wales, in particular in the western area of the country, were virtually entirely Ancient Briton, suggesting no Vikings settled in these regions.”

http://www.bbc.co.uk/history/programmes/bloodofthevikings/genetics_results_05.shtml

But the place names in south Pembrokeshire suggest that Vikings did settle. Otherwise why would the Welsh in this area have adopted and then continued to use so many Scandinavian place names that arose from a language that would have been completely alien to them?

The ‘large parts of Wales’ described were actually only Llanidloes, Haverfordwest and Anglesey. I don’t think how testing three places can possibly offer an accurate genetic picture of all Wales, they can only give us the history of a very specific location; as in the genetic traces of the 19th century English immigration to lead mining areas in Llanidloes that you mentioned.

Going back to place names...

I noticed, in addition to your observation of Penycwn at the north west edge of the English area, that about 2 miles to the north of Haverfordwest, there are three ‘comb’ names; Pelcomb Cross, Pelcomb Bridge and Pelcomb. Although these use the English spelling of ‘comb’, rather than the Welsh ‘cwm’ they could be Brythonic survivors.

And going back to Devon...

I think that Devon has some things in common with south Pembrokeshire. Originally a Celtic region, it has experienced Irish, Viking and English settlement over the centuries. All these cultures have left traces; Viking in the place names of Croyde and Lundy, Brythonic Celtic in many place names, local traditions and church dedications, Irish in the Ogham stones, and English in the huge number of Saxon and medieval English place names.

I don’t think that there was a massive programme to change Celtic place names in Devon after the Norman Conquest. Most of today’s Devon towns appear in the Domesday Book with their present day, mostly English names. But there is some evidence, as with Bishop Grandisson’s apparent dislike of Celtic church dedications, the loss of certain place names that you mention, and reported evidence of Celtic speech in medieval Devon, that there was a substratum of pre-Saxon Celtic culture in Devon existing into the post-Norman period (and maybe until much later) that was seen as problematic or even just as unfashionable, difficult or archaic by the people and the authorities of the time. As I have said before, although these accounts are not proof, they originate from a time when it was not fashionable to support Celtic culture, and it is hard to see the motives of the various individuals, like Risdon or Baring-Gould, to make such things up.

ExeValleyBoy,

The DNA result for Haverfordwest is no more than a snapshot of the local area at about the end of the 19th century (due to the paternal grandfather birth-place stipulation). If there had been large-scale movements of people prior to that time, these would not necessarily have shown up in the results, unless they had resulted in permanent changes in the genetic composition of the population at the end of the 19th century. As I stated in my last post, I have read somewhere that most of the Flemish migrants to the area had departed again before this time, and if (as seems likely) they had not interbred with the native Welsh, they would have taken their gene pool with them when they left.

Even the continued existence of English/Flemish place names and the continued use of the English language are not too difficult to explain. If the departing Flemish had been replaced by English speaking Welsh people from South and South East Wales, rather than by the Welsh speaking people of nearby Cardiganshire, the new residents would have had no incentive to rename the towns and villages in Welsh. However, the new residents would have been genetically indistinguishable from the Welsh-speaking people of Cardiganshire and further north. I agree with you that the ‘comb’ names of Pembrokeshire may well indicate pockets of Britons at the time that the Flemish were in occupation of the area.

I do not see the situation in Devon as being the same as that in Pembrokeshire. There is certainly no evidence of a new population arriving in Devon in the 13th or 14th centuries and departing again by the late 19th century. Yes there are a lot of English place-names in Devon (especially among the towns), but some of these are later additions, often resulting from the rapid expansion of the Woollen industries. The earlier ones were undoubtedly founded by Saxon migrants. The majority of the Brythonic names are to be found among the isolated farmsteads and small hamlets.

Incidentally, your presumed etymology for ‘Croyde’ is not now the accepted one. Andrew Breeze has shown compelling evidence for ‘Croyde’ being derived from the the Brythonic (primitive Cornish) word ‘crud’ meaning ‘cradle’, which precisely describes the ‘trough-like’ valley in which Croyde is situated.

Although the (limited) English settlement of Devon was undoubtedly real, the supposed Irish and Viking settlements were more probably in the nature of transient visits and probably confined mostly to the northern coastal regions, such as Lundy, whose name is definitely Norse in origin. Many of the small islands off the coast of Wales also have Norse names. The evidence for Irish influence is mainly that of the Ogham stones and the names of Irish (Celtic) saints, who were probably few in number. As the Irish DNA was found to be similar to the local Devon variety, we cannot tell for sure whether the Irish stayed in Devon.

I would describe the pre-Saxon Devon culture as rather more than a ‘substratum’. Although the Brythonic Language itself seems to have dwindled to nothing sometime around the 14th century, other aspects of Celtic culture seem to have survived for much longer (often into the present era). I see few (if any) specifically Saxon cultural remnants in the artistic or sporting fields.

>>The evidence for Irish influence is mainly that of the Ogham stones and the names of Irish (Celtic) saints, who were probably few in number. As the Irish DNA was found to be similar to the local Devon variety, we cannot tell for sure whether the Irish stayed in Devon.

Whilst there is no definitive evidence on DNA on any Irish component toDevon's modern day population (at least not yet), those untrustworthy scientists of the Victorian era used to look to physical characteristics.

According to a number of those (and I will try to dig out some references) they suggested that North Devon and Cornwall had a relatively high proportion of dark hair/grey eyes which (they claimed) reflected an Irish origin.

Whether this has an element of truth, or whether they were looking for something to fit their preconceptions I do not know.

“Incidentally, your presumed etymology for ‘Croyde’ is not now the accepted one. Andrew Breeze has shown compelling evidence for ‘Croyde’ being derived from the the Brythonic (primitive Cornish) word ‘crud’ meaning ‘cradle’, which precisely describes the ‘trough-like’ valley in which Croyde is situated.”

Pembrokeshire aside for a moment, I find this really interesting. From what I can see the Viking explanation is still deeply entrenched, the village even having a Viking festival in June to celebrate its apparent Nordic founder. This is believed to have been a personality called Crydda who landed there and founded a settlement.

On a web search about Croyde I found Crydda described mostly as a Viking settler, but others describing him as a Saxon.

The thing is this Crydda name is not made up. There is a Crydda valley—the valley in which Croyde is situated, and a Crydda stream flowing through the village. The Crydda Valley presumably being what Andrew Breeze identified in his Cornish interpretation of the name.

You say Andrew Breeze has ‘compelling evidence’ that Croyde-Crydda is actually Celtic. If you still have the text or web link at your disposal please post it as I would very much like to see what Breeze has to say about Croyde’s origin.

If Andrew Breeze is right and Crydda is actually Old Cornish, this actually has important implications for Devon place name research. It shows that Crydda the Viking was a legend created in later times to explain the Anglicised name of the village taken from its older form surviving in the clearly non-English Crydda valley and stream.

I also found that Croyde is generally believed to be a pre-Saxon era settlement, and that in the village there is a 12th century chapel dedicated to the Celtic Saint Endelentia who, it is said, brought Christianity to the village in the 6th century.

ExeValleyBoy,

I think that to a large extent, you have answered your own question in your last sentence. If Croyde was a pre-Saxon settlement, and St. Endelentia brought Christianity there in the 6th century, then it could not have been founded by a Viking. The Viking raids and settlements in Britain did not commence until the 8th century, and those raids (no evidence of settlements apart from Lundy) in the South West were even later. Thus the chronology is all wrong for a Viking founder of Croyde. The ‘Viking’ festival in June would appear to be an invented excuse for a celebration.

None of the respected place-name etymologists derive the name from that of a Viking, or indeed of a Saxon. Ekwall in 1960 suggested that the name of the village derived from the name of the headland, which he tentatively derives from OE ‘cryde’ from ‘crudan’, meaning ‘to press, to make ones way’. Another derivation by Campbell in 1972 suggests OE ‘cryde’ meaning ‘weeds’. Breeze finds neither of these derivations at all convincing, and neither would explain the name of the valley or of the stream. He therefore suggests that it is much more likely that the name of the village comes directly from the name of the valley and stream, and proposes a name, which is descriptive of the valley, i.e. Brythonic ‘crud’ meaning ‘cradle’. Not only would this tie in well with the villages presumed pre-Saxon origins and Celtic saint, but Breeze has also shown that, when later Anglicised, the pronunciation would have changed from a ‘u’ sound to a ‘y’ sound, hence ‘Crideholde’ in the Domesday Book and ‘Crideho’ in 1242. It is interesting to note that a later form of the name, ‘Crude’ (1276) and the present day ‘Croyde’, both seem to be closer to the original.

The above etymology is from the book “Celtic Voices English Places” by Richard Coates and Andrew Breeze (2000).