Castell Carreg Cennen

For many years Castell Carreg Cennen has been seen as the poor brother to the world heritage castle sites such as Caernarfon, Conway and Beaumaris built by Edward I in North Wales.  Now this image can be dispelled for new research strongly suggests that this castle is the princely pinnacle of Welsh castle building and was far more sumptuous than any Edwardian castle apart from the unfinished Caernarfon.

Standing on top of a 200' high limestone cliff Castell Carreg Cennen dominates the mountains and valleys of Deheubarth and controls the passage to the sea at Swansea.  The site was first used by prehistoric man and Roman coins have been found in the clearances of the castle.  The name is first mentioned before 1143 and there seems little doubt that a castle of some description stood here from that date onwards.  Unmentioned in the brutal wars that scarred Deheubarth in the twelfth century the castle first finds prominence in the Welsh Chronicles for 1246 where it was wrongly reported that the castle had been betrayed to the French.  In fact royal forces were holding the castle for the legitimate Anglo-Welsh heir and his widowed Anglo-Norman mother!  In 1248 that heir, Rhys Fychan of Dinefwr (d.1271), came of age and inherited the castle with the consent of King Henry III.  During Llywelyn ap Gruffydd's penultimate war in 1276b the castle was surrendered without a siege and was then burned and repeatedly rebuilt in the ensuing wars, changing hands twice in 1282 and once in 1287.  The Welsh castle, surrendered to King Edward I in 1277, was devastated by Dafydd ap Gruffydd in 1282 during his final war and repaired and refortified many times after this, being heavily rebuilt after a particularly heavy slighting by Owain Glyndwr in 1403 or 1404.  In 1462 the castle was finally demolished.

To this end the fortress can be seen as a mid thirteenth century Welsh princely answer to White Castle and Pembroke, which had been built on either side of the much disputed territories of the princes of Deheubarth. The princes of Gwynedd may have been the major power in Wales in the thirteenth century, but their rivals of the south undoubtedly built more ornate castles if we accept current wisdom. Certainly the grandeur and style of the Welsh-built Castell Carreg Cennen leaves Dolbadarn and Dolwyddelan almost in the Dark Ages, but other sites are yet to be properly examined in the light of current research.

As I have often stated it is well nigh impossible to accurately assess the date of the building of any piece of masonry.  Even ‘dateable' architectural features do not help as such things tend to have a shorter life span than the walls in which they stand.  Thus they are often and even regularly replaced.  All that can be given in regard to dating is broad outlines and inferences from occasional documentary references.  What really helps in suggesting dates is the history of a site and its owning families.  In this respect we are lucky with Castell Carreg Cennen.  The castle site has, since the coming of the Normans, been associated with one family, that of the kings of Deheubarth and their descendants right up until the wars of 1277 and 1282.  After this sufficient royal records survive to allow us a pretty good idea concerning the activities of the lords of this district after that date.  It is therefore well worth suggesting a building chronology for the castle remains deduced from this recent research.

It would seem that the first work built upon the possibly pre-historic site of Castell Carreg Cennen was a fine sandstone hall or llys.  This was possibly built under the auspices of Prince Anarawd, the grandson of King Rhys ap Tewdwr (d.1093), in the period 1138 to 1143.  Many would disagree upon such an early dating, because it has been taken as fact for hundreds of years that Welsh princes did not build stone castles, and if they did they had to be second or third rate structures.  Such ‘fact' is not based on any contemporary evidence, merely long established hearsay.

It should be remembered that the Welsh princes are said in contemporary sources to have taken to the Norman method of warfare during the reign of King Henry I (1100-35).  It is a small jump to go from Norman style cavalry to Norman style castles.  Indeed the earliest recorded Welsh castles built by the princes in this district are from the early twelfth century. A Welsh castle was standing at Aberystwyth in 1143 and others were in existence during this early period at Cilsant in St Clears, Cynfael, Llanrhystud and Ystrad Meurig.  Therefore it is historically possible that some form of building or fortification was constructed at Castell Carreg Cennen during the reign of King Stephen (1135-54).
Einion ap Anarawd (bef.1143-63) or even the Lord Rhys (1132-97) may have been responsible for adding the first eastern masonry defences which can still be seen at Castell Carreg Cennen.  This includes the early rectangular keep and the cistern tower both of which may have been built with adjoining curtain walls in the period up to 1162.  Then, quite possibly in the fighting that occurred in Deheubarth up to 1171, the castle was destroyed.  It may well have lain derelict for a long time judging by the height of the remains of the features just mentioned and built into the next phase of construction.  Some stone for the early castle seems to have been imported to the site from the area of the River Cennen about a mile to the east of the fortress.

The final stage of the castle's building has long been ascribed to the work of the Giffards of Clifford in the period 1283 to 1322.  This has mainly been alleged as the king was not responsible for its building and the Welsh were regarded as too ignorant to build a superbly ornamented castle.  However, as the history researched in this new book on the fortress has shown, the descendants of the Lord Rhys were not just ‘backwoodsmen', but were princes of the highest rank and well known in Norman society through power, wealth, marriage and lineage.  The history uncovered has also shown that the Giffard tenure of Is Cennen was far from secure and that the family showed little interest in the land.  It therefore seems unlikely that they were responsible for building such a showy and expensive castle as Carreg Cennen.  Indeed, it may only be John Giffard Senior (d.1299) who had sufficient reason to consider building such a castle, but his family history shows he had little to do with the district and apparently needed Dinefwr castle as a ‘refuge' in 1290.  After his death in 1299 Is Cennen would appear to have been held by his widow, the Dowager Lady Margaret, until 1308.  John Giffard Junior (d.1322) also seems to have had a full career elsewhere than Wales and again it must be doubtful as to how much time and money he would have had available to spend on refortifying Castell Carreg Cennen.  This is especially true after his capture and ransom at Bannockburn, even though it must be remembered that he was referred to as ‘le rych'.

The style of the castle also suggests a Welsh rather than an Anglo-Norman origin.  Wall passageways exist in great towers from the eleventh century onwards and occasionally appear in curtain walls, as at Caernarfon; however these are in a totally different league from those at Carreg Cennen.  In Norman castles such passageways invariably have the thin wall internally and the thick wall externally to face the enemy.  The engineer in charge of building the wall passages at Carreg Cennen thought the other way round would be preferable!  Such features are not common and perhaps show closest comparison to the late thirteenth century Caernarfon castle where the mural passages are much more advanced.

Finally, we come to the barbican and the similarities of this to other long barbicans at various Welsh castles.  At Carreg Cennen the structure, over 100' long, has superficial similarities to those at Castell y Bere, Dinefwr and DenbighDryslwyn too has a small barbican - or more accurately an extended gatehouse to the inner ward.  The middle and outer wards in some ways too, may be looked upon as large barbicans, both being impediments to reaching the inner ward.  After excavation the work at Dryslwyn is considered Welsh and the entrance to Dinefwr appears to be too.  The barbican at Castell y Bere covers a gateway next to a round tower.  It consists of steps rising up between two walls from a small irregular rectangular gatetower and covered by a rectangular guards' tower.  At the top of the steps was apparently a drawbridge which led to a slightly projecting hole-in-the-wall gatehouse next to the round tower.  It has been alleged that this was all built by Llywelyn ab Iorwerth between 1223 and 1240, but there is no evidence for this.  However this entrance does appears to be a much simpler version of the Carreg Cennen barbican, although the structure has often been dated without documentary or physical evidence to the 1286-87 building phase of the castle.  More likely this long barbican is from a long Welsh tradition as suggested above.

When all this evidence is considered there seems little doubt that Castell Carreg Cennen was a castle fit for a prince and it is possible that the current inner ward as it stands was built by Rhys Mechyll (d.1244) or his father, Rhys Gryg (d.1234).  Rhys Mechyll was an Anglophile with an Anglo-Norman wife.  She was to defend Dinefwr and Castell Carreg Cennen for the Crown and her children in the wars of Prince Dafydd (d.1246).  Further, the destruction of the castle of Llangadog in 1209 would have offered a good incentive for the local princes to build a new more powerful castle in the district.  It is therefore possible that Rhys Gryg was responsible for this act after he was confirmed in possession of Is Cennen and the bulk of Ystrad Tywi in 1216.

It should further be noted that no well has been found at Castell Carreg Cennen and that masonry castles built by Normans most often were equipped with such a feature, cf. Montgomery and Painscastle in the early thirteenth century and the castles of Edward I in the latter part of the century.  Certainly King Henry III (1216-72) was adamant that his castles needed secure water supplies provided by exceptionally deep wells.  Native Welsh castles like Criccieth, Dolbadarn, Dolwyddelan
and Harlech appear to lack wells as do the nearby castles of Dinefwr and Dryslwyn, though it should be noted that a well cum cistern has recently been excavated at Dolforwyn, but this was apparently made after the castle's fall to the English in 1277.  There is also a cistern that has been found in the gate passageway at Criccieth.  Another well cum cistern exists on top of the rock of Degannwy castle.  The lack of any proper flanking on the west front of Carreg Cennen also argues against an English design as too does the relatively poor ditching, especially when the amount of money Edward I spent on ditching his great castles at Harlech and Criccieth is considered.

The other additions of the ‘Welsh style' long barbican and curiously weak and unditched outer ward could have been the act of either Rhys Fychan (d.1271) or one of his sons, the last Welsh lords of the district.  Certainly the castle was strong enough to hold out during the wars of Prince Dafydd (1241-46) and during the disputes between Rhys Fychan and his uncle in the 1250s and 1260s when other castles in the district changed hands.  Similarly, when King Edward I acquired the castle in 1277 he did not think that any work was necessary at the site.  Nor did his viceroy Tibotot judging by the money expended on munitioning, but not repairing the site.  Over a hundred years ago a good historian, J.E. Morris, noted that during the period 1277 to 1282 ‘comparatively small sums were spent [on Dinefwr, Carreg Cennen and Llandovery], which points to the castles being already in a good state of repair'.  I can do no more than agree with this assessment.

The book Castell Carreg Cennen consists of 373 A4 pages with 336 diagrams, family trees and plans.  It can be purchased directly via the PayPal link below for £39.95.

Why not join me here and at other Lost Welsh Castles next Spring?  Please see the information on tours at Scholarly Sojourns.


Copyright©2011 Paul Martin Remfry