Llancarfan Historic Landscape

The Glamorgan-Gwent Archaeological Trust Ltd.


Historical Processes, Themes and Background

Natural Landscape

The historic landscape of Llancarfan occupies a relatively narrow river or stream corridor cut through the gentle undulating lowland plateau of the Vale of Glamorgan. The natural landscape of the historic landscape of Llancarfan, and the Vale in general, has been modified by man's influence since at least the Neolithic period if not before. The general topography of the area is essentially that of the generally shallow river valley of the Nant Llancarfan, which extends southwest from its watershed just south of Bonvilston, with its tributary streams of Nant Whitton, Moulton Brook, which itself is fed by Ford Brook, flowing in to join from the northeast, until it reaches its confluence with the River Waycock to the east, from where, known as the Kenson River, their combined force flows west past Llancadle to join the River Thaw. The various tributary streams and rivers cut through the agricultural plateau to the east of the main river valley of Nant Llancarfan to form steeper side valleys delineating spurs of land approximately 60m OD in height at their confluences. On the western side the topography of the limestone plateau is less affected by stream action. The settlement of Llancarfan itself is located at a pinch point in the valley just north of the confluence of the Nant Llancarfan with Moulton Brook; to the north and south of this point the valley broadens out.

Significant parcels of ancient woodland are found throughout much of the historic landscape, notably on the steeper slopes of the valley sides, which were less favourable for arable use, such as within HLCA008, for example Coed Garn-llwyd is a SSSI (Nant Whitton Woodlands), and the Ancient semi-natural woodland on the slopes south of and below Caemaen Farm. Further remnants of Ancient Woodland can be found within HLCA002 to the north and south of Castle Ditches Hillfort, including the woodland of Breach Wood, apparently named after the local craft of tailoring. Indicators of formerly more extensive woodland are traceable, not only from remnants of woodland, but also irregular small enclosures, hedges with specimen and frequent woodland character, and woodland place name elements, such as Ty'n-y-coed (HLCA012). Elsewhere the matrix of woodland and enclosure has been shown to be the result of a mixture of medieval assart, i.e. woodland clearance, and enclosure of open waste ground; small areas within HLCA008 and elsewhere within the historic landscape may have evolved from a similar process, however, further detailed survey work would need to be undertaken to establish the period and extent of these landscape forming processes. Given the concentration of prehistoric defended settlements in the area it is likely that the plateau areas to either side of the Nant Llancarfan and its tributaries had been cleared and agriculture established from an early date.

The evidence for woodland management of the area's ancient semi-natural woodland has yet to be investigated, although it is likely that the area's woodland, in addition to pannage, would have provided a range of resources for the local community, a source for wood for construction, agricultural and domestic purposes for fuel and craft industries, and possibly charcoal burning.

Back to top of page

Historic administrative divisions

The historic landscape is situated within the post-medieval County of Glamorganshire, which is divided into two physically distinct regions, that of Blaenau Morgannwg an area of upland on high Pennant Sandstone plateau, and that of Bro Morgannwg, or Vale of Glamorgan, an area of lowland on gently rolling limestone plateau dissected by shallow river valleys; Llancarfan is located within the latter area.

Prior to the Norman conquest, the area formed part of the kingdom of Morgannwg, and prior to that the early kingdom of Glwysing, named after an eponymous early King, Glywys. According to tradition Glwysing was divided into seven administrative regions or cantrefi, while 12th century sources assert that these were named after the sons of Glywys. Traditionally each cantref was divided into cwmwd, or commotes, each comprising estates or maenorau made up of a number of trefi, or townships. The lands of Glwysing lay traditionally between the Afon Towy and the Afon Llwyd, but borders appear to have been fairly fluid with the loss of territory to the west to Dyfed and the gaining of land to the east to encompass Gwent and the Ergyng. From the late 10th century the territories of Gwent, Gower and Glwysing were united under the name Morgannwg by Morgan Hen ab Owain. The territory which formed Morgannwg prior to the Norman Conquest included the following cantrefi: Gwrinydd (Gorfynydd), Penychan, Senghenydd, Gwynllwg, the commotes of Afan ande Nedd, as well as territory in Gwent. Llancarfan, Llanfeuthyn (Llanvithyn), Bonvilston, Penmark, and Llantriddyd, as well as other adjacent areas fell within the cantref of Penychan. Following the Norman conquest of the region in around 1091, this area formed part of the 'shire-fee', within the Hundred of Dinas Powys, of the Norman County of Glamorgan (Richards 1969).

It is notable that given the quality of the agricultural land that the area would have supported a relatively high population from the prehistoric period onwards. There is a possibility that some of defended enclosures and hillforts may have continued in occupation from the late prehistoric period into the early medieval period, as at Castle Ditches and Castell Moel, indeed the latter site appears to have been adapted following the Anglo-Norman occupation to serve as a manorial centre.

The Norman invasion in the latter part of the 11th century precipitated the collapse of this Welsh kingship and resulted in the creation of the Norman lordship of Glamorgan. Robert Fitzhamon became ruler of this new lordship by virtue of conquest; he had been the key figure in the Norman invasion of the Welsh kingship from a likely base in Gloucester. It is not wholly clear how the invasion was achieved, but it is clear that the area conquered did not include the entirety of the kingship, the conquered area being limited to an area approximately corresponding to that of Bro Morgannwg or the Vale of Glamorgan.

Subsequent to this initial conquest the administration of the area underwent many changes during the medieval period. Fitzhamon granted lands within his newly conquered lordship to the Abbeys of St Peter's at Gloucester and to Tewkesbury and divided the remainder of the area as demesne manors and manors respectively between himself as marcher lord of Glamorgan and his followers; the degree by which the land grants reflect or alter the boundaries of earlier land holdings of the pre-Norman landscape is not fully understood, it is possible that earlier land boundaries and holdings were preserved within these land grants.

Parish and manorial boundaries use physical features in the landscape, and are usually long-lasting and conservative characteristics within the landscape; for this reason where possible the character area boundaries have adopted former manorial or parish boundaries, where these can be seen to broadly reflect corresponding variations in character.

Back to top of page

Agricultural landscapes

There is little direct evidence for use of the present agricultural landscape in the prehistoric period; though the existence of hill forts and defended enclosures, such as Llanvithyn Camp. Castell Moel, and Castle Ditches, dating to the Iron Age, indicates settlement at least by the late prehistoric period and associated with it extensive clearance of the surrounding woodland to allow mixed, though probably pastoral dominated, agriculture. Environmental evidence indicates a marked increase in clearance activity during the Iron Age, which continue into the Romano-British period, corresponding to the occupation of hillforts and defended enclosures.

The economy was predominantly pastoral, and contrary to earlier views most Iron Age and Romano-British sites in Wales where examined in sufficient detail do have evidence for arable cultivation. Little environmental evidence is directly available for the area in relation to early agricultural practice, though an attempt was made to estimate the minimum number of domesticated animals present in the assemblage from Castle Ditches (Parkinson 1976): sheep or goat and Cattle were found to be present in equal numbers with approximately up to a quarter of sheep killed young, and two pig bones immature, the remainder from mature animals. Waterlogged plant remains from the well at Whitton, a defended farmstead of Iron Age and Romano-British date, located just east of the historic landscape relate to a local environment of plants perhaps indicative of an established agricultural landscape with a mix of weeds common to both arable and waste land identified (Wilson 1981; Caseldine 1990). However, no physical remains of associated enclosures or agricultural features of the period have yet been identified.

The fieldscape of the greater part of the Llancarfan can be traced back to the open field system of the medieval period. Whilst a consensus of opinion views this system as a Norman introduction (Emery 1971, 155), it is just possible that the land was organised in a similar way for agriculture prior to the Norman occupation (Kissock 1991). It has been considered that the medieval period is the earliest period to which significant areas of the modern landscape can be traced; whilst at present this may be true, future investigation/study may shed more light on the subject and allow some reconstruction of the landscape associated with the early medieval settlements of Llancarfan.

The openfield systems would have been divided into strips separated by turf banks, known as land shares or quillets, (a landholding in a communal field). The management of openfield systems, such as those at Pen-onn, and elsewhere indicates a characteristic mix of arable, and communal grazing, regulated by agreement among the strip-holders. These types of holdings appear to have been worked communally. The most extensive and best preserved of the relict fieldscapes of the medieval period in Llancarfan are undoubtedly to be found within HLCA008 Llanvithyn and Garnllwyd, within HLCA011 around Castell Moel (Liege Castle), within the area around Middlehill (HLCA007), where a number of fossilised strip fields survive, and within and east of HLCA002 in the vicinity of Pen-onn. Here, partly under the ownership of the Dean and Chapter of Gloucester Abbey, the medieval pattern of strip fields survived as interspersed holdings into the mid 19th century. The pattern of dispersed holdings under both secular and ecclesiastic ownership would suggest that land formerly belonging to the early medieval church granted to Norman Benedictine foundations continued to be farmed on a similar basis to other holdings within the openfield system, and may have been farmed by lay tenants.

Other relict agricultural features survive to the southwest of Llancadle within HLCA004; these are part of the wider system, which extended beyond the current limits of the historic landscape north towards Llanbethery and is recorded on a manorial map of 1622. It is notable that perhaps the best of these field systems actually lie well beyond the boundaries of the current historic landscape on the Register. These areas contain to varying degrees remains of strip fields or quillets characteristic of the openfield system fossilised as boundaries within the current fieldscape of varied amalgamated fields.

The manorial map of Llancadle of 1622 surveyed by Evans Mouse as part of a wider survey of Bolingbroke lands shows the openfield system at a time when they and the manorial system instrumental in its development was in decline. This map is nonetheless important in documenting the extent of the surviving medieval field pattern of the area by the early 17th century and evidence farming practice. The openfields comprised groups of share lands or strips, each strip representing one day's ploughing by the village ox team. To ensure that the best of the arable land was shared out equitably, no adjacent plots could be owned by a single tenant. Meadow land was shared along similar lines, and common meadow such as existed in Llancadle (within HLCA003), was highly valued during the medieval period; when services were commuted to rents meadows yielded more per acre than arable, usually twice as much. Enclosed crofts adjoining the farmstead, where livestock could be reared, are also a noted characteristic of the system (Davies 1957).

Examination of the tithe map of Llancarfan indicates that isolated landshares had become fossilised within the enclosed fieldscape as narrow linear fields, adjacent to other enclosures created from amalgamating holdings into larger enclosures. This process included enclosure of openfield arable, common meadow and to a lesser extent common or wasteland. The current landscape of consolidated farm holdings is largely in place by the mid-19th century (Llancarfan tithe map), with the notable exception of the area east of Pen-onn under the ownership of the Dean and Chapter of Gloucester, and the area south of Bonvilston, where interspersed holdings of the openfield survive partially fossilised. Further amalgamation of strip holdings, and rationalisation of enclosures has occurred by the survey of the 1st edition 25-inch OS map. Little change then occurs to the fieldscape of the area until the late 20th century when changing farming practices are again reflected in a rapid increase in field amalgamation on the agricultural plateau areas (HLCA006 in particular, and HLCA005 to a lesser extent), with the removal of hedgerows to create the large prairie-like fields required for large-scale agricultural production and increased mechanisation. Topographical restraints, and different ownership, have to an extent limited these changes to the more level agricultural plateaux and have prevented loss of smaller enclosures within the confines of the river valleys.

Small parcels of irregular enclosure have been identified on the margins of surviving ancient woodland and other broadleafed woodland, for example within HLCA008 southeast of Gowlog and south of Caemaen Farm, and in areas with indicative woodland related place names, such as Ty'n-y-coed, at the southern edge of HLCA 009; this is perhaps indicative of later piecemeal enclosure and clearance of woodland, or assart activities. It is not clear whether this relates to medieval or early post-medieval agricultural expansion.

Back to top of page

Settlement Landscapes

Settlement within the area dates from the prehistoric period and is characterised by several prehistoric defended enclosures, chief among which is the hillfort within HLCA002 to the south east of the present settlement of Llancarfan, known as Castle Ditches (SAM GM071; PRN 00383s). This site is a univallate hillfort of c 4.2 ha occupying the west end of a ridge which rises gently to the east and falls away steeply on other sides, with its east approach defended by a strong rampart and flat-bottomed ditch, its other sides defended by a steep external bank, apparently originally revetted in stone, with ditch and counterscarp bank, and an entrance located to the south. The entrance has a slight inturn on east side is protected by two banks and ditches extending from main enclosure. The evidence is suggestive of continuity of settlement, with occupation extending from the late prehistoric period through the Roman and possibly into the early medieval period, as suggested by documentary evidence, though it is not known whether this was continuous or interrupted. The later hillfort seams to have been preceded by smaller enclosure of different plan defended by stone wall, whilst the interior has revealed the foundations of round hut associated with finds of plain Iron Age B sherds, as well as sherds with decoration similar to those found on the Somerset levels, and Roman pottery of 2nd-4th century date. Further defensive characteristics of similar date are to be seen at Llanvithyn Iron Age Camp (HLCA008), SAM GM293, and Castell Moel (SAM GM298) within HLCA011, where the remains of the univallate hillfort is graphically shown to have been re-occupied during the medieval period by the site of a moated 'manor' known as Liege Castle. Cropmark sites identified from aerial photographs within HLCAs 005 and 006 may also have been defensive settlements of prehistoric or possibly early medieval date.

To date little detailed analysis of the area's settlement has been undertaken which directly relates to Llancarfan and its surrounding area, though the early medieval settlement layout of the area has been touched upon by Jeremy Knight, who has argued that documentary evidence (chapters 48-54 of the Vita Cadoci) preserves indications of the way in which the monastic community at Llancarfan exploited the agricultural landscape of the area prior to the Norman conquest; a community of 36 canons, each with a prebend of 80 acres is thought to have existed at Llancarfan. The land was worked by a lay community, described as hortolani or bondmen, for the benefit of itself and the canons. Knight has also identified many of the places identified within the text relating to monastic site of Llancarfan, and these lie within the parish of not only Llancarfan, but also Penmark and Porthkerry (Knight 1984, 395-8).

Davies and others have stated that the nucleated settlement of the area, in conjunction with the formerly characteristic openfield system, is a result of Norman colonisation; however, Kissock in his article on the origins of medieval rural settlements in Glamorgan (Kissock 1991, 31-49) suggests that settlement nucleation at Llancarfan and elsewhere in Glamorgan could be a much earlier phenomenon. Kissock identifies Llancarfan as possessing the fundamental attributes of a multiple estate, of late Roman or Early-medieval date: 'a coherent landscape unit including both fertile lowland and less attractive clay lands of river valleys'. In the wider landscape (i.e. beyond the bounds of the current historic landscape, it is known that Roman farms existed in the fourth century at Aberthaw and Llanbethery, to the south and west respectively (and also at Whitton to the east). Kissock notes a settlement pattern of a single main village with eight associated hamlets with an area of strip field around each settlement, and the presence of a clas community as being indicative of the existence of a multiple estate.

A characteristic feature of the historic landscape and its surrounding area are the numerous areas of former openfield or strip field, such as that around Llancadle and east of Pen-onn; the latter appears to be the land mentioned in a late 12th century grant made to St Peter's Abbey at Gloucester by Robert Fitzhammon. This grant not only included the church of St Cadoc at Llancarfan and Pen-onn together with 15 hides of land, but also, significantly, the former property of the clas. The tithe map of 1840 shows land in the same area, then under the ownership of the Abbey's successor 'the Dean and Chapter of Gloucester', as enclosed strip holdings dispersed throughout the area of the former openfield.

This openfield may be an example of a type of land known as tir corddlan or nucleal land; this has been identified by GRJ Jones as having an association with monastic clas communities and is often noted through its late survival as fragmented Glebe or ecclesiastic land holdings intermingled with land owned by others (for example Jones 1972, and 1976). Kissock has presented a model which demonstrates that the emergence of early settlement nucleation could have resulted from the donation of multiple estates or parts of multiple estates to the church, with the accompanying change from secular to ecclesiastic ownership, and a change from tir cyfrif or 'reckoned land', where the dispersed single farmsteads or dwellings were found, to tir corddlan, represented by large areas of openfield (Kissock 1991). Whilst this model might explain some of the general processes behind the formation of the early settlement nucleations within the historic landscape, further study and investigation is required to fully investigate and confirm the nature and date of actual settlements, both deserted and surviving in the area.

There are no Roman or later landscapes within Llancarfan that can be said to be truly defensive, though several include defensive characteristics or elements. Of particular note is the medieval ringwork at Pancross (NPRN 307703; PRN 00904s), a defensive settlement feature, though more a likely constructed as a symbol of power than as a truly defendable site. It has been suggested that there may have been an earthwork castle at this site (RCAHMW 1991), potentially of 12th century origin is likely that this castle would have been associated with Norman settlement of the area, in particular with the Umfraville family of Penmark. The chronology of the development of defensive enclosures into ringworks is largely unproven: no excavation of the ringwork at Pancross has been carried out and, as a result of recent damage to the monument, it is unlikely to yield much evidence for period its of occupation/use. Continuity of use/or reuse of defensive sites, specifically enclosures and ringworks, between the Early-medieval and the post-conquest period has been observed, although the lack of recent excavation means the exact chronology and nature of this reuse is limited, or speculative at best; it should therefore be borne in mind that the ringwork as a site type is not necessarily exclusively Anglo-Norman in origin.

Another medieval defended feature is found in the remains of the small 'moated site' of Liege Castle, built within the ramparts of the late prehistoric hillfort known as Castell Moel (PRN 359s). The principal remains consist of a strong bank and ditch, about 15m wide by 3m high with traces of a counterscarp bank, forming three sides of a rectangle about 25m across. From the northeast corner, a slightly larger rampart with its ditch almost silted up extends southeast for about 30m in a slight curve. Slighter banks extend all these features for about 18m further south. The site is considered to have been built by the Norris family of Penllyn Castle, perhaps as early as the 13th century. Liege Castle was a sub-manor of Bonvilston, which was itself a sub-manor of Wenvoe. As a subordinate residence, its main function was presumably as the centre for the administration of the sub-manor, possibly housing a steward. A small hamlet appears to have developed in the vicinity of the present Liege Castle Farm.

A number of deserted or shrunken medieval settlements are to be found within the historic landscape, some of which have had some level of survey carried out by the RCAHM(W); for example the relict and buried archaeological features, such as lynchets, field banks, a possible pillow mound and house platforms, associated with Bradington Deserted Medieval Village at Llanvithyn (RCAHMW 1982, 229), a settlement considered to have been partly removed by expansion of the adjacent grange belonging to Margam during the latter part of the medieval period. An agricultural settlement (PRN01910s) considered to be a possible location of a medieval grange granted to St Peter's Abbey, Gloucester, has been identified at Pen-onn. This site is described as a 'large complex with a main terrace running along an old field boundary, with a main platform at west end, 15m x 8m, with wall bank, 2 smaller square platforms (8m x 8m), a lower platform (6m x 6m), and a central circular depression'. The latter visited as part of the Cadw-funded deserted rural settlement project (Locock 2001). Other shrunken settlements with associated agricultural remains include that on the fringes of Llancadle (PRN 00691s and 02428s) which comprises a hollow way, a series of level platforms and terraces and possible lynchets, and that to the south of Liege Castle, focussed on the area between Leach (Liege) Farm and the remains of Liege Castle Chapel.

The exact layout of the deserted settlements of the area remains obscure, though they are thought likely to have originally had some level of nucleation. There is a noticeable similarity with most of the locations of the older settlements in the area: they are frequently sited at the valley edge and the edge of the agricultural plateau, at upper break of slope, a location which allows convenient access to both the openfield arable on the agricultural plateau and the common meadows in the river valley. Llancadle is the most obvious as a surviving example, whilst the deserted rural settlement of Pen-onn, with its clustered layout at the junction of the main lane to Moulton and tracks giving access to the former openfield hinterland, is another. Pen-onn, formerly a cluster of farms, is set close to the upper break of slope of the Nant Llancarfan valley between its arable openfield to the east and its common meadow in the valley below to the west. Similarly, the deserted settlement at Llanvithyn appears to be located again at the upper edge of the valley side. The settlement of Llancarfan is at odds to the location being located within the valley itself at a convergence of routes and river crossings, though sited significantly close to the major prehistoric hillfort of Castle Ditches. It is likely that Llancarfan's location is governed more by Early-medieval ecclesiastic sensibilities and administrative requirements: the need to be centrally located close to a major pre-existing settlement and power base, and easily accessible within a wider local landscape, access to the nearby agricultural hinterland being less of a day-to-day requirement.

Two nucleated settlements survive in the historic landscape itself: Llancarfan and Llancadle; these settlements both have origins of medieval date, the former with confirmed finds dating back to the early medieval period. Llanacarfan is nucleated on its medieval church, set within its irregular enclosure, which is likely to be the Early-medieval monastic enclosure, though another potential site has been identified nearby to the southeast, with remains of 13/14th century date. A separate sub-focus has also been identified to the north at the former mill and road junction/river crossing, which has been linked to the main settlement by linear ribbon and more recent infilling of modern dwellings. The settlement or hamlet of Llancadle appears to have originally focussed during the 17th century upon the site of its former chapel of ease, now demolished, with the main farm located at the western side of the settlement. Its form is as a settlement cluster, which would have been ideally placed to serve the surrounding openfield and the meadow lands beyond in both the Kenson Valley and the Thaw Valley to the east beyond the historic landscape boundary. During the 20th century the focus of settlement has perhaps slightly moved to the route between Aberthaw and Llancarfan at the east end of the village, as represented by a discrete area of ribbon development, and the replacement of much of the original cluster of farmhouses and cottages by detached commuter and/or retirement dwellings. The status of Llancarfan is as the main settlement within the area, emphasised by its ecclesiastic focus, the parish church, and also by more recent elements shown on the 1st edition map such as the village school, nonconformist chapels and burial grounds, a public house, poorhouse, post-office, and smithy, most of which survive to this day. Historically Llancarfan would have not only been a religious, educational and administrative (a function which dated back to the period of its Early-medieval origins) centre for the locality, but appears to have continued to develop as a wider focal point for the area during the medieval and post-medieval period, at first through its connections with St Peter's Abbey at Gloucester, then as the parochial centre with functions of supply and distribution, education, a place supplying rural specialist crafts for the surrounding agricultural area, not to mention hospitality in its inn, the Fox and Hounds.

Llancadle on the other hand appears to have remained a subsidiary settlement based from the medieval period largely on agriculture, being a cluster of farms, farm buildings and agricultural workers' cottages: this relationship is emphasised by its close proximity to its openfields and common meadows. The agricultural emphasis appears to change little until the end of the 20th century; the Green Dragon public house, depicted as a public house on the 1st edition OS, for example was formerly an outlying farmstead in the 17th century. It is the nature of change in the agricultural regime, through enclosure of openfield underway by the early 17th century and largely complete by the mid-19th century, which had the most noticeable effect on the layout of the settlement of Llancadle. This process was accompanied by a reduction in the size of the settlement during the post-medieval period, which culminated with the renewal and expansion of the agricultural ranges belonging to the main farm at Llancadle by the last quarter of the 19th century. The main farm in the settlement appears to have expanded at the expense of smaller holdings, so that effectively a single farm with outbuildings remained within the settlement by the survey of the 1st edition, with the other farms becoming smallholdings or cottages.

On map evidence it is possible that several of the farmsteads, and associated agricultural buildings within Llancadle might date back to the early 17th century or possibly earlier; for example the Green Dragon, Lower Llancadle and Llancadle Farm are all shown on the manorial map of 1622. There is certainly cartographic and archaeological evidence to suggest that Llancadle was once far more extensive, with occupied crofts (now relict and buried archaeological features) noted to the south and west. Both settlements are now characterised largely by a mixture of post-medieval housing and farmsteads of 18th or 19th century date, with original features surviving and more recent 20th century development infilling. Within Llancarfan several of the older cottages are characteristically built into, rather than across the slope with gables facing on to the road, in line with earlier building traditions.

Both Llancarfan and Llancadle are protected through designation as conservation areas, though apart from the church of St Cadoc's within Llancarfan and the nearby telephone box, none of the buildings within the two main settlements of the historic landscape is currently protected through listing. It would also appear that conservation area status has not successfully controlled the use of inappropriate materials and styles in relation to existing building renewal and modern house design; replacement windows, chimney stack removal, and inappropriate treatment of rooflines and extensions, which conservation status should guard against, are all visible in the area.

Apart from the two main settlements mention above the layout of the majority of the surviving settlement is characteristically one of dispersed scatters of farmsteads, occasionally the post-enclosure remnants of more extensive settlement clusters or nucleations. A loose cluster is also noted at the road junction of Pancross, where post-medieval farms have been augmented by a terrace of council housing (Cattwg's Cottages), and other cottages during the 20th century.

A number of interesting late medieval/early post-medieval gentry or manor houses survive within the landscape area; these include Garnllwyd (Listed Grade II*; HLCA008), a first-floor hall house with early post-medieval additions, and Llanvithyn House and its dated gatehouse (both Listed Grade II; HLCA008), an early 16th century gentry house displaying affinities to the unit system group, whereby completely self-contained dwelling units are closely grouped on a single site. The hall house (PRN 01428s) at Crosstown, within HLCA005, is also of medieval origin. This belongs to the chimney backing onto cross-passage category, and has an inserted central chimney and ogee-moulded beamed ceiling (now removed) of 16th century date. Significant surviving medieval features at the site include a two-centre headed doorway with plain-chamfered jambs to the cross-passage, and a trefoil-headed single light.

Surviving farmsteads of the early post-medieval period are relatively rare, even within the historic landscape of note is Cliff Farmhouse (Listed Grade II), near Llancadle, a good example of a 17th century thatched roofed farmhouse of 3-room plan with a central doorway, and winding stair by gable fireplace, and retains original features such as ogee-stopped beams (RCAHMW 1988). The characteristic building materials would have been local limestone, usually lime-washed, under a thatched roof; some buildings were also roofed in stone slabs (as indicated by the name 'Ty-to-maen', though the previously common thatch has invariably been largely replaced by imported and locally produced pantiles, such as that noted on the steeply pitched roofs at Whitewell Farm, and slate, as noted at Crosstown.

The small size of the current historic landscape, and the small number of surviving older buildings does not allow any useful conclusions regarding building style or architectural development in the wider context of the Vale. There is little surviving evidence for hierarchy of building styles of early post-medieval date; most of the older building stock belongs to the gentry or upper yeoman class of farmer. Within the settlement of Llancarfan a broader range of buildings survives giving some indication of 18th/19th century social hierarchy, represented by subtle degrees in the size between the various dwellings from smaller terraced cottages, now Bridge Cottage, and attached cottages (for example Corner House) and detached cottages with small associated plots (such as Fern Cottage), to subtly larger buildings notably Great House, which along with the now demolished Ty-to-maen may have been the main farms in the settlement (the former may have had a strong association with the church). Agri-industrial buildings in the form of mills and the site of the former smithy also indicate the service and supply functions of the settlement. The differentiation between non-conformist chapel and established church is also visible in the built fabric of the settlement with separate burial grounds indicating changes in the social make up and aspirations within the settlement, and the wider rural landscape.

Back to top of page

Ritual and ecclesiastical landscapes

The earliest known ritual feature in the area is the recumbent stone north of Ty'n-y-coed, within HLCA012, a possible portal dolmen funerary monument of the Neolithic period, though the validity of its identification has recently been called into question (PRN 00934s). A Bronze Age standing stone, the Druidstone within HLCA008, is also considered to have had some form of ritual function.

Other ritual features are the areas numerous wishing wells or rag wells such as Llancarfan well (PRN 01849s) HLCA001, said to cure the King's Evil, Ffynnon-y-Clwyf (PRN 00391s), and Ffynnon Flamaiddan (PRN 00513s) both within HLCA002 also noted healing wells, the latter known for the cure of erysipelas. The name 'rag well' originates from the tradition of hanging pieces of material on bushes after drinking at the well (Jones 1954, 185-6). Ffynnon Dyfrig within HLCA008, which formerly had a small stone wellhead structure, is thought, though not confirmed, to have been a medieval holywell.

The ecclesiastic characteristic or influence largely dates from the Early-medieval period with the establishment of the monastic settlement by St Cadoc at Llancarfan between the 5th and 6th century AD; the irregular-shaped churchyard enclosure 'Llan Garbhan' (PRN 03736s), is thought to be the most likely site of this early medieval foundation (PRN 384s; RCAHMW 1976, 17 no 827), see HLCA001 for further information. The presence of a late 9th or 10th century pillar cross (PRN 780s; RCAHMW 1976, 62 no 940), and later medieval church (Listed Building Grade I; PRN 00385s), are strong ecclesiastic characteristics and had a lasting influence on the earlier and continued development and layout of the settlement at Llancarfan, right up to the post-medieval period, though the nature of the ecclesiastic focus would have subtly changed from early monastic clas to appropriated church, and later parish church.

A number of other ecclesiastic sites have been considered to have been Early-medieval foundations, such as Llanvithyn (see below), and Llancadle, tentatively associated with the early medieval site of Llan Hoitlan, where the remains of a chapel of 14th century date once stood. The physical evidence for these is now ephemeral, and the Early-medieval association recently open to question (Evans 2004,65).

Like many important early medieval monastic foundations, such as Llandough, and Llanulltud Fawr, the clas at Llancarfan passed into Benedictine control following the Norman Conquest, being granted to the Abbey of St Peter's at Gloucester during the late 11th or early 12th century. If the model for the establishment of monastic lands mentioned above is correct, it would appear from the cartographic evidence that the nature of the agricultural landscape associated with the early monastic foundation such as the holding at Pen-onn (partly within HLCA002), and the way in which it was worked appears to have continued much as before, though worked either by tenants or by lay brethren on behalf of the new ecclesiastic landlord, the Abbey of St Peter' and later Tewksbury Abbey. This may not have always been the case, as it appears elsewhere the loss of land from secular to ecclesiastic hands may have sometimes lead to reorganisation of landholdings and drastic consequences for the secular tenants and their settlements, for example at Llanvithyn (HLCA008), where an extensive area of earthworks defines the deserted village of Bradington, depopulated when the grange of Llanvithyn belonging to the Cistercian Abbey of Margam was enlarged (RCAHMW 1982, 229). Remains associated with the grange at Llanvithyn are as yet largely unestablished; documentary evidence suggests a grange chapel (dedicated to St Meuthin) was built on a plot of land especially donated for the purpose (RCAHMW 1982, 291), and burials have been identified in the area, though it is uncertain whether these belong to the grange or an earlier undeveloped cemetery site of the Early-medieval period (Evans 2004, 65). A curvilinear bank to the south of Llanvithyn, roughly concentric with the cemetery site, is considered by the Royal Commission to have been the grange boundary (RCAHMW 1982, 292), though this has also been tentatively identified also as being of Early-medieval origin, as it appears to underlie medieval settlement features. The area requires further investigation to clarify its landscape development.

There were numerous other monastic granges throughout the area such Greendown (HLCA009) granted to Margam Abbey by Hugh de Raelega c. 1161 (Williams 1990), and land associated with Margam, 10 acres of pastureland at a place named as Moys or Moyl on 13th and 14th century lists (PRN 03803s; Williams 2001, 306 no. 91, RCAHMW 1982, 297) considered to have been centred on Liege Castle in HLCA011. The exact effect on the landscape that the granting of land to the various monastic orders had in the area, remains to be explored in detail. It is likely to have depended on when the grants occurred and to have varied dependant on the monastic order to which the land was granted.

Other relict ecclesiastic remains include Liege Castle Chapel (PRN 00362s; HLCA011) and cemetery, part of a once larger rural settlement. It has been tentatively suggested, though unconfirmed that Liege Castle Chapel once formed part of Greendown grange.

Nonconformist chapels with associated burial grounds are a later ecclesiastic characteristic, mostly located within the village of Llancarfan itself (HLCA001), and include White Chapel (PRN 01419s) an early 19th century Wesleyan Chapel (now converted) with whitewashed rendered walls, lancet windows, slate roof and Bethlehem Chapel (PRN 1421s) a 19th century chapel with an interesting front gable elevation with round headed doorway and windows in a restrained simplified classical style, whilst within HLCA011, near Liege Castle is a the simple 19th century non-conformist chapel, Carmel Chapel (Independent), also converted to a dwelling.

Back to top of page

Industrial features

Industrial archaeology is represented mainly by milling, a characteristic of the landscape with early mills associated with the monastic grange of Llanvithyn recorded in medieval documents; in 1336 the 2 mills at Llanvithyn were worth 2 pounds. The same mills (both within HLCA008) were later leased to Sir John Raglan in 1519; mills or their converted buildings remain on the same site today, with notable post-medieval mill buildings mills (PRNs 01993s & 01848s; NPRN 24,942; LB 13,611 grade II), built of local stone with sandstone dressings comprises a 17th century mill house with original features and adjoining largely 19th century three-storey mill. Llanvithyn Woollen Mill (now the Old Bakehouse) retains its associated race and sluice. Other surviving mills are located within HLCA001, namely the converted Corn mill and New Mill at Llancarfan, whilst cartographic evidence, the manorial map of 1622, indicates a tuck mill formerly stood on western banks of the river Kenson, within the south extent of HLCA002.

The agri-industrial characteristic is further represented throughout much of the area by small-scale agricultural quarrying and associated lime production, typically quarries and limekilns, are noted on 1st edition OS maps throughout the area. It is likely that some of these features may date back to at least the early post-medieval period, though a noted increase in extraction during the 19th century represented in the documentary evidence is likely to have been the direct result of agricultural improvement associated with enclosure and the amalgamation of holdings. These features largely relate to agricultural lime production, though might also have produced building stone. The majority of these features are recorded in documentary and cartographic material and their current condition is largely unknown.

Back to top of page

Historical associations

An ecclesiastic charter of the 7th century witnessed by Iacob abbas altaris Sancti Catoci (LL144 of c650), and another of the 10th century including the phase simul cum dignitae pontificalis cathedrae abbati totius dignitatis ecclesiae Sancti Catoci Lann Caruaniae (LL243 of c980) (Davies 1978, 135; 1979, 97, 125), indicate the association of the area and Llancarfan with St Cadoc/Cattwg, son of Gwynllyw, a contemporary of Dubricius, about AD 500; the latter is also given (as Landcaruan/Nant Caruguan) by Lifris's Vita Sancti Cadoci of c 1100 (Wade-Evans 1944, xi, 52-5). It has been postulated that the name Llancarfan or 'Lann Gharban' as mentioned in the 9th-10th century Irish Lives of St Finnian relates to the 5th century St Germanus, an association with St Dubricius of the late 6th century, has also been raised. A association with St Meuthin, Cadoc's teacher, is also claimed for the chapel and possible early medieval site at Llanvithyn (Llanfeuthin).

The area is associated with 'Caradog the Historian', refered to by Geoffrey of Monmouth in his Historia, and author of the Life of Cadog and the Life of Gildas, and possibly the Vita Cungari, is considered also to have had a possible association with the Book of Llandaff. Also strongly associated with Llancarfan is the bard and discredited historian Iolo Morgannwg, born at Pen-onn in 1746 and noted particularly for his imaginative interpretation of history including the fabrication of several chronicles, such as the Brut Aberpergwm and the Brut Ieuan Brechfa (Lewis 1971, 449-554).

During the medieval period the area is associated with the Abbeys of St Peters, Gloucester and Tewksbury, and also the Cistercian Abbey of Margam, through grants of land by Fitzhammon and others. For example, Liege Castle (HLCA002) is known to have been associated with both Margam Abbey, and the Norris family of Penllyn Castell from at least the 14th century. During the 12th century Hugh de Raelega is known to have been responsible for the grant of land at Green Down (HLCA009) to Margam Abbey, whilst there are known associations with the Umfravilles of nearby Penmark, who are thought to have controlled the ringwork at Pancross (NPRN 307703, PRN 00904s; HLCA006), during the same period.

The major post-medieval landowners included the Dean and Chapter of Gloucester Cathedral, and a number of gentry families such as the Giles Family (HLCA005 and 006), the Carne family, the Raglans (of Carnllwyd) and Lewis family of Y Fan (HLCA008, HLCA011 etc), the St John family of Fonmon, who were lords of the manor of Llancadle and elsewhere. Another notable personage was Sir John Wildgoose, owner of Liege Castle (HLCA011) during the reign of Elizabeth I. The main landowners of the area by the mid 19th century were Sir Thomas Digby Aubrey and CKK Tynte Esq.

The area is also associated with William Griffith of Llanvithyn, a noted Glamorgan Catholic recusant, who with his father Hugh purchased the grange of Llanvithyn (HLCA008) in 1565 from Thomas Carne (Pugh 1986) and with Henry Williams of Llancarfan, a famous 18th century clock and watchmaker (Cloutman and Linnard 2003a). Without doubt further associations also exist to be revealed.

Back to top of page