Development of the Castle

Excavations at Nevern Castle have revealed that all the earthworks are medieval: there was no prehistoric settlement on this site. A single leaf-shaped flint arrowhead attests the presence of Neolithic hunters, possibly using the site as a lookout for animals in the valley below.
The Welsh of the Early Medieval period have left little archaeological trace at Nevern beyond the early medieval inscribed stone monuments in the churchyard. A series of thin, narrow ridges from ploughing were recovered from beneath the motte. This indicates an active Welsh farming community on this site immediately prior to the Anglo-Norman conquest. This is the only evidence we have uncovered for such a community. Elsewhere in Cemais evidence is also scarce; a round house dated AD980-1160 inside a defensive bank was found beneath the castle at Maenclochog.

The first defence constructed on this site was a small bank, around 1m high, which probably had a shallow ditch in front of it, enclosing an area of ground at the top of a steep sided natural promontory. Dated to the early 12th century through the associated pottery, this formed a ‘conquest’ castle, established circa 1108 by Robert FitzMartin to protect his forces as they subdued Cemais. A motte (defensive mound) with a surrounding deep ditch was then added to this bank and a four-post wooden structure was constructed on the summit of the motte which, given its small size, probably acted as a watch tower.

Subsequently there was a period of domestic occupation of this castle site. Fragments of pottery, scraps of bone, horseshoe nails and the charcoal from many campfires have been recovered from these earliest levels of the site.

A little later, probably circa 1116 when there was considerable Welsh military activity, a series of substantial banks and ditches were constructed on this site, burying the earlier bank and traces of occupation. On the west side of the site the banks and ditches formed a triangular castle with the motte at its apex. The bank and ditch on the east side of the castle (castle-town ditch) is today only visible as a shallow depression in the ground. The west bank is buried under the present large bank and ditch. The south side of the castle was defended by the natural slope, with a bank, ditch and inner bank constructed at the top of the slope. To the east there was a defended area. This could be an outer bailey to the castle, but is more probably a defended township with three large banks and two ditches to the north and a ditch and palisaded bank to the east.

There was evidence of two phases of wooden buildings in the town; one constructed with posts and other with sleeper beams.
The entrance to the castle was probably in the south west corner and the town entrance in the south east corner. Both were approached up steep slopes. On the top of the north bank there was evidence for two phases of wooden palisade. The good defences of this earth and timber castle and town show the efforts being made by Robert FitzMartin to colonise and develop Cemais. His establishment of the monastery of St Dogmaels and the foundation of a series of smaller castles throughout Cemais also occurred in this period.

In the next phase of activity, the castle and town were combined into the large castle we see today, and the castle began to be rebuilt in stone. The bank between the castle and town was slighted (taken down) and the castle-town ditch partially filled in. A large circular tower was constructed on top of the motte, which was deliberately reduced in height in order to support this two or three storey tower. This Round Tower was built of slate and occasional faced glacial stone boulders, mortared together with clay. It had a first-floor entrance accessed by external wooden stairs. Access to the cellar or basement room was only through a trap door in the floor. It is probable that later in the castle’s history the Lord Rhys was held prisoner in the basement of this tower, as it is one of the most secure locations on the site.

A second building, built with slate and some faced glacial boulders mortared with clay, was the earliest phase of the Great Hall. This building, 15m x 6.2m, was built near the south west corner of the site. The southern ditch was probably filled in and the banks levelled. Analysis has shown that the clay used as mortar in these buildings is the natural subsoil of the site. This is a traditional construction technique in north Pembrokeshire and may represent 12th century Welsh construction tradition before the widespread use of lime mortar.

A third building, probably a Chapel, was constructed in clay-mortared slate on the southern edge of the site, with a clear view of Nevern church. The presence of a new substantial wooden palisade on the top of the middle north bank is probably part of the refortification of the castle which occurred at this time.

Though we cannot be certain who started to rebuild the castle in stone, the process had commenced by the middle of the 12th century when Rhys ap Gruffudd (the Lord Rhys) was in control of this area. He was building castles from at least 1156 and is recorded in the Brut-y-Tywysogyon as building with stone and mortar in 1171 at Cardigan castle. The use of clay mortar and using buildings as part of the castle’s perimeter were features seen in later Welsh castles. Consequently, it appears likely that this initial stone castle was the work of Rhys ap Gruffudd.
Subsequently the castle’s accommodation and defences were greatly expanded. A stone entrance was constructed through the clay bank south of the motte. This western entrance comprised a wooden bridge across the west ditch and a walled passageway through the bank, with access probably controlled through wooden gates in the passageway. The Great Hall was extended to 22.2m x 6.2m, with a substantial hearth at the centre of this large space. An adjoining East Hall was added. This may well have been a large open hall as no evidence of internal divisions was recovered.
In the same phase of building work, a curtain wall was constructed running down from the corner of the Great Hall to protect the south west corner of the site. The castle was now extended with the creation of an Inner Castle on the eastern promontory, separated by the rock-cut ditch. A curtain wall was built around this rock outcrop together with the North Hall. This effectively created a much more secure castle with an inner and outer ward structure. Access to the Inner Castle was via a wooden bridge over the rock-cut ditch from the bailey. The curtain wall of the Inner Castle and that protecting the south west corner of the site were of similar widths (circa 1.2m) and forms of construction. All these buildings were constructed with large slates mortared with clay (no faced glacial boulders).
These impressive constructions along the southern edge of the site were designed to be seen from the valley below. At a time when only a few churches were built in stone, these buildings were some of the largest secular stone buildings in west Wales; demonstrating the wealth and power of the Lord of Cemais.
Creating at least five widely spaced high status stone buildings in a three-acre site was very unusual. Welsh castles such as Dolbadarn or Anglo-Norman castles like Manorbier are far smaller, more compact structures. It’s fair to suggest that this site was the llys (court) of the young Rhys ap Gruffudd. It draws together the defensive features of a castle with the traditional range of buildings of a Welsh llys.

At this point in the castle’s development there appears to have been a substantial siege which resulted in the south west curtain wall being demolished. Evidence from a series of postholes indicates that a wooden palisade was erected in its place to ensure continued defence of the castle. The western entrance also suffered attack, probably resulting in damage to or loss of the gates, as a wooden palisade was constructed across the entrance. This was later replaced with a clay-mortared slate blocking wall, but both the entrance and blocking wall were subsequently damaged and partially dismantled. It appears likely these are all events of a large, unrecorded siege, possibly when the Lord Rhys (Rhys ap Gruffudd) recaptured the castle in 1159 or 1165, after he had returned it to the crown in 1158.
The castle was subsequently redeveloped with a Rhomboidal Tower built on the south side to replace the slighted curtain wall defence. A new entrance was constructed in the south western corner beside the Rhomboidal Tower, whilst the damaged western entrance was buried under a clay bank. A curtain wall was probably built to replace the wooden palisade on the top of the west and middle north banks of the castle and a substantial square or rectangular tower was added at its southern end. This Southern Tower overlooked the road and protected a new south west entrance. A small tower may also have been constructed on the north east end of the curtain wall. Traces of a revetting wall present at the top of the southern slope of the hill may also have protected this side of the castle. All these structures were clay-mortared slate—clearly in the Welsh tradition.
There is also evidence for a number of smaller wooden houses with clay or slate floors on both the east and west sides of the bailey. It is likely that the castle contained many such domestic structures, but subsequent ploughing has removed almost all sign of them.

A final phase of construction enhanced the appearance of the site: the western bank was made higher and much wider with slate derived from making the western ditch wider and deeper. A new entrance was constructed with gritstone blocks. The pitched stone threshold of this southern entrance incorporated slates with apotropaic symbols scratched onto them to ward off evil.
This phase saw the frequent use of gritstone: finely dressed square and rectangular masonry blocks used for the corners of buildings and around doorways and windows. The creation and use of finely chiselled, squared blocks of stone is an Anglo-Norman building tradition, though they are still assembled using clay mortar. These techniques may suggest a fusing of Welsh and Anglo-Norman building traditions and almost certainly occurred after the arrival of William FitzMartin in the 1170s. This would also explain the use of a classic Romanesque chip-carved saltire pattern in the gritstone, part of a frieze in the Great Hall. New doorways in these gritstone blocks were added to the east end of the East Hall and the Chapel. The gritstone blocks are largely unweathered, as would be expected for stone added late in the castle’s history.
Subsequently a large Square Tower with rounded corners was constructed in the Inner Castle butting onto the earlier curtain wall, possibly reusing some of the slate from the earlier North Hall, which appears to have been partially dismantled at this time. The ditch between the bailey and the Inner Castle was quarried much deeper, though the work stopped when only half done, possibly as a result of the recapture of the castle by the Lord Rhys in 1191. The slate from the ditch may have been used in the Square Tower, which was barely finished before the castle was slighted.

All buildings in the castle showed extensive signs of burning and walls were deliberately collapsed. The cracks and partially displaced section of the Round Tower wall may be part of this deliberate demolition. This destruction can confidently be ascribed to the historically attested slighting of the castle in 1195 by Hywel Sais, as dating of pottery and other artefacts corresponds with a late 12th century period.

After the slighting of the castle, the site was looted for the large pieces of slate in the walls and the squared or decorated gritstone blocks. These were probably reused in the construction of farmhouses, churches and other buildings in the area around Nevern or Newport. Consequently, few gritstone blocks now survive on the site. Interestingly, none of these gritstone blocks was re-used in the rebuilding of St Brynach’s Church in Nevern during the 13th-16th centuries.
A series of pits was dug down into the rubble to try and recover the large slates and gritstones. These pits were often dug in places such as the entrance to the Great Hall, where such stone was originally located. This has removed the evidence of the grand entrances to these buildings.
In some cases, the ease of access to structures like the Southern Tower at the end of the west bank, which is immediately beside the road, meant that they were almost completely quarried away. On the north eastern edge of the site there is a quarry face cut into the native slate bedrock. By the 16th and 17th century Nevern was recorded as exporting slate for roofing, quarried at cliff face locations like this. This industry may well have developed out of the looting of slate from the ruins of Nevern Castle.

Very little pottery dated between the 12th and late 17th century date has come from the site, suggesting that little more than stone looting took place over this period. However, substantial amounts of pottery were deposited from the late 17th century onwards, and the extensive plough damage over much of the centre of the site shows that it was used intensively for growing crops from the late 17th century through to the Second World War. Traces of a cottage (Pwll-y-broga), which was still in use in the 19th century, are visible in the southern end of the west ditch.
A cottage and associated building, together with 18th century pottery and a nit comb, were uncovered at the south end of the rock-cut ditch. This cottage could have been associated with either the slate quarrying or farming on the site. There were also fragments of the walls of a further cottage or farmyard building found on top of the shattered remains of the Great Hall, partially buried beneath a later modern bank. These remains plus traces of a pitched slate path and the ruts from farm carts suggest a third post-medieval cottage or farm building here.

Traces of a slate-lined drain were recovered from both the castle-town ditch and the inner north ditch, indicating that efforts were made to drain water from the centre of the site to improve agricultural productivity during the 17th–19th century. A later modern bank which still cuts off the south west corner of the site is all that remains of a series of walls and banks, recorded on the tithe maps of the 19th century, which originally criss-crossed the site. These walls suggest that the grazing of animals on the banks and ditches was separated from crop growing in the bailey.
Throughout this period the castle continued to erode down the southern slope of the hill. The southern walls of the Chapel and Rhomboidal Tower have both collapsed down the slope.