A Research Framework for the Archaeology of Wales

Northwest Wales – Medieval






The chronological limits of the period reviewed in this paper start with the 11th century, which saw the first Norman incursions into Wales, their initial defeat by Gruffydd ap Cynan (d. 1137) and the start of a successful period of Welsh rule which was to end in 1282/3 with the death of Llywelyn ap Gruffydd and conquest by Edward I. The period ends in the late 15th century, by which time many of the hallmarks of the archaeology of the Middle Ages, in particular the castles and monasteries, had fallen, or had started to fall, into decline.


Archaeological studies of the period have tended to concentrate upon the more dramatic elements, and thus the literature is dominated, as was the landscape of the day, by castles and monasteries. These have left a legacy of dramatic ruins, which have excited the imagination and fed the scholarly mind for several generations. Only slightly less dramatic are the domestic structures built to house the leading families of the day with their highly ornate timber roofs and painted walls. Still less visually dramatic are the far more numerous remains of the homes of the lower status families, and the associated remains of agriculture and industry. Nonetheless these elements influenced the formation of the rural and urban landscapes in which we live, and a determined effort to understand and record the medieval landscape and the elements out of which it was formed is at the forefront of current archaeological research.


This study will examine the strengths, weaknesses, opportunities and threats of the archaeology of north-west Wales as requested by the steering committee of the Archaeological Research Framework. I am very grateful for the assistance I have received from Richard Avent, David Longley and Richard Suggett, whose ideas and contributions have influenced much of this paper. Research requires a suitable environment in which to flourish. It also relies heavily upon previous research, and the availability of readily accessible and reliable information. Gwynedd is relatively well served in this respect. Secondary archaeological sources, which provide a solid base for further research, include the Inventories of Anglesey and Caernarfonshire published by the RCAHMW and the more recent county history of Merioneth. Publication of prime historical documents is also of great help, and of particular note here is the publication of the 1352 extent of Anglesey by A D Carr. The area has also been well-served by historians and historical geographers – the likes of Jones-Pierce, Glanville Jones, Colin Gresham, E A Lewis, Harold Hughes, Arnold Taylor, Dafydd Jenkins, Anthony Carr, have done so much spadework that workers today are able to enter the fray with the experience of many cumulative years of study having been undertaken.




High status sites




The high status sites of the pre-conquest medieval period are those constructed for the principal rulers in both lay and ecclesiastical capacity. The former are the courts and settlements built for use by a peripatetic court that ruled by administering local justice and accepting dues in kind. The investigation of Llys and Maerdref sites conducted by GAT and funded by Cadw has considerably enhanced our knowledge and increased our understanding of location both within the landscape and within the jurisdictional boundaries of the day. The excavations at Rhosyr have confirmed the belief that these were complex sites, composed of all those buildings necessary for the running of an administration, including a hall, solar, chapel and ancillary buildings. Building techniques included both timber on stone foundations and stone buildings.

The motte was a Norman introduction, and yet appears to have quickly become utilised by Welsh rulers, and in three instances at least was built by them. There is widespread survival of mottes in the area, and though of mixed potential, some are well preserved. Several have attributable historical references to their construction and/or use. There is also evidence for early fortification by Welsh rulers at sites such as Carn Fadryn and Deganwy. Palaces of the Bishop of Bangor have been excavated at Gogarth and Bangor. Neither have revealed the full complexity of the settlements, though the large defensive hall at Gogarth, built entirely of stone, is of interest, as is the later development close by, which more closely resembles the llys at Rhosyr in layout and building composition.



The area is rich in masonry castles of the later medieval period, often with accompanying boroughs. These have long been studied, most notably by Arnold Taylor, though also by others. Major architectural studies have been published for the majority, and the Cadw guidebooks contain excellent summaries of the information available. The development of castle architecture has long fascinated scholars, and there is a wealth of literature that places the Edwardian castle in its historical and evolutionary context. Nearly all the stone castles and mottes are scheduled ancient monuments, and all the Edwardian castles are in Guardianship and managed by the state, thus ensuring preservation for future generations of scholars. There have been excellent studies of surviving medieval houses, but we are in particular indebted to Peter Smith for his descriptions and analysis of the development of higher status houses. Examples of note include Hafodty, Llanddona; Ty Gwyn, Barmouth; Plas Ucha, Llangar; and Penarth Fawr. From later in the period Cochwillan and Gloddaeth, with their hammer beam roofs, survive. The recent improvements in dendro-chronological dating have helped our understanding of the chronological development of domestic architecture.



Rhosyr is the only llys site to have been excavated, and thus our present knowledge is dangerously founded on evidence from this one site. We have only a hazy understanding of the chronological development of the llysoedd, and to what extent they may lie on centres of pre-12th century importance. The Welsh use of pre-Norman style fortification, as, perhaps, evidenced at Carn Fadryn, is poorly understood.


No excavation and little survey work has been undertaken on mottes. Therefore the date of construction of many of them is unknown, and thus arguments as to who built them and subsequently used them have little chronological basis. There are only two ringworks (Pen-ucha’r Llan and Tomen Fawr) and one certain moated site (Tregarnedd) in the area, which contrasts with other areas of Wales where these sites can be found in greater numbers, yet the reasons for this contrast are not fully understood. The 12th century earthwork at Trefadog may be classed as a ringwork also, and in which case so might Castell Crwn, but we have at present little idea of the status and background, or even nationality, of the people who built them, nor what influenced them to build in this way. Similar questions remain of some of the lesser Welsh stone castles such as Castell Carn Dochan and the enigmatic Castell Prysor.


Very few archaeological remains of high status houses of the 12th and 13th centuries are known. Where were the leading families living at this time? The monastic palaces at Bangor and Gogarth provide some information on high status halls, but few archaeological sites have been excavated from this period.


There is a danger that research tends to become more limited on subjects that have been apparently fully researched, and yet recent research into documentary sources for some south Wales’ castles, particularly Chepstow, Kidwelly and Raglan, has revealed new original material which has led to a major re-appraisal of the architectural development of the monuments.



Additional excavation on potential llys sites will provide a clearer understanding of their chronological development and the nature of the buildings. Pottery and finds from these sites will provide information to aid our understanding of trading relationships and the economic status of the sites. Excavation at selected mottes will result in a better understanding of their chronological development. A study of land tenure, landscape and political history combined with more informed knowledge from excavation will give a clearer understanding of the role of mottes as defensive and administrative sites.

Similarly, more research is required into the role of castles built by the princes of Gwynedd. Some lying on the border appear to be defensive (Ewloe, Dolforwyn, Deganwy and Castell y Bere), but the role of Dolbadarn, Dolwyddelan and Dinas Emrys is more puzzling, and they may be associated with protecting vaccaries. Landscape, documentary and archaeological analysis will help solve these problems.


Whilst considerable work has been undertaken on the origins and development of the Edwardian castles, there is opportunity for further understanding their internal layout, and the changes which occurred as the role of the castle changed from primarily defensive to administrative and domestic. Dendrochronological studies combined with careful structural recording and analysis will be able to offer firmer ideas on the chronological development of structures and timber constructional techniques.



Earthwork castles are inherently unstable monuments, and frequently suffer erosion and damage from burrowing animals and roots of trees.





There has been considerable work undertaken by historical-geographers and historians on economic, social and tenurial aspects of this period. The recently completed Deserted Rural Settlement project has identified large numbers of rural structures, many of them of potential medieval date. These have been added to the Sites and Monuments Record. Large numbers of surviving upstanding structures have been surveyed and mapped, largely by Investigators of the RCAHMW, and listed in the County Inventories and the new Merioneth County History. Work by Peter Smith and Richard Sugget has resulted in a clearer understanding of the development of the small hall house and the sub-medieval houses of the yeoman farmer. This work has been aided by dendro-chronological studies.


Though relatively late, the area is well-served by a number of excellent 18th century estate maps which clearly indicate remnants of medieval open field systems lying amongst more recently enclosed land. The use of these with earlier documentary references has resulted in a number of studies which have been able to chart the development of the landscape from medieval times to the present. The advent of environmentally sensitive agricultural schemes will help protect many of the minor archaeological elements of the countryside which have been so rapidly destroyed in recent times.



There has been a tendency to examine landscape developments from map and documentary evidence alone. Fieldwork has tended to lag behind, and thus the archaeology of medieval settlement and the wider landscape is less well understood and recognised on the ground. Whereas the DRS project identified many settlements in marginal and upland areas, we know of very few sites in the richer lowland areas of Anglesey, Lleyn, and the Arfon coastal belt. We need to know where these were sited. We need to know why they do not appear in archaeological assessments and watching briefs of pipelines and road schemes, in contrast to the late Prehistoric and Romano-British settlements, of which several have been found and excavated. If, as is often stated, they underlie existing farmsteads, then we need to adjust our methodologies to confirm this.

Very few excavations of medieval settlements have been undertaken. Cefn Graeanog has been the most successful to date, but we have no idea if this is a typical settlement, as there is so little with which to compare it. This lack of excavation has also led to a dearth of archaeo-botanical studies. Our knowledge of the operation of medieval open fields, and their origin and demise, is poorly understood. Similarly little work has been undertaken on agricultural techniques, and the development of the plough, use of ox and horse as traction animals and the development of plant and animal husbandry.



Archaeologists should build upon works by historians and historical-geographers (the two articles by F A Barnes on Cemais and Cleifiog are examples) and by using aerial photography, field walking and field evaluation techniques, attempt to locate the archaeological remains of the medieval landscape. It is very necessary to undertake multi-disciplinary projects linking environmental, historical and archaeological methods to provide a better understanding of the medieval landscape and its development. Archaeo-botanical analysis and experimental archaeology can be used to provide a clearer understanding of agricultural techniques, food sources and cooking. Computer aided graphics can be used to enhance maps and aerial photographs. This allows improved landscape modelling and interpretation. Project designs for contract archaeology need to incorporate research themes, so that return on expenditure is maximised, and money is not spent on projects which result in a mass of grey literature which is of benefit to no one.



Drainage has severely reduced the number of lowland locations suitable for providing environmental information.





In the pre-conquest period it was largely the maerdref settlements of the Welsh princes which formed any sort of urban and trading environment. Examples include Llanfaes, Aberffraw, Pwllheli and Nefyn, though major church sites at Bangor and Tywyn may also have provided early urban nuclei. The rapid decline of Llanfaes forced by Edward I and the subsequent conversion of the area to parkland by the Bulkeley family has resulted in much of the former town being preserved, though we still lack a clear understanding of its layout. Gwynedd contains a significant number of Edwardian boroughs with good historical records and the potential for providing archaeological evidence of high quality. Dendrochronolgy will be able to date with certainty some of the surviving timber buildings, as has been undertaken at Aberconwy house, Conwy, which has been dated to 1420. Historical references to ports are common, and early maps indicate the location of many of the major landing places. Some of these are now be-sanded or land-locked, and offer opportunities for research and the recovery of medieval wrecks. Examples include Penrhyn, where the earlier port lies upstream of the later slate harbour, Abermenai on Anglesey, the site of a medieval ferry, but also where Gruffydd ap Cynan is known to have landed, and the port which served the royal court at Aberffraw. A small number of medieval wrecks have been discovered within the area, two of which, both from an inland location at Llyn Peris, have been dated to circa 1200 and 1547-9.



Little concentrated archaeological work has been undertaken within urban environments. Major opportunities in Bangor and Caernarfon have been missed in the past. Much of the work is often undertaken as small watching briefs or evaluation trenches, and it is rare to be able to place the findings in a wider context. Small scale work has been undertaken at all the Edwardian boroughs and at many of the other urban centres, but the resources are not available within each of the projects to provide a full synthetic analysis of the results.


No work has been undertaken on port archaeology despite the remains of a number of known ports which were never developed in post-medieval times.



It would be of advantage to pull together the results of smaller projects on medieval towns and combining them with wider analyses of settlement morphology, street layout, location of seminalbuildings etc. This would provide a clearer understanding of the elements which made up a medieval town, and help identify surviving remains. Eastgate Street, Caernarfon, under which lies a medieval bridge, is an example where several projects have been undertaken over a period of considerable time, yet the results are disparate, and no project has brought together all the results, nor attempted to place the significance of the bridge within the layout of the town and the development of commercial areas immediately outside.

Though medieval ports are not easy to recognise from archaeological evidence alone, we have their location from historical evidence, and archaeological assessment of these sites will provide a clearer understanding of their location and form.




Utilisation of raw materials, including copper, lead, slate, limestone and coal, is known to have taken place, but very little work has been undertaken on the industries of medieval Gwynedd. Some knowledge exists of the coal mining industry, and some work has been undertaken on milling and textile mills.

Our knowledge of iron working in the medieval period in Gwynedd has been considerably enhanced by the work of Peter Crew at Coed y Brenin.



Our understanding of medieval industry, and the impact of the Edwardian conquest with its influx of craftsmen, is very poor, and few archaeological studies have been undertaken.



The very nature of resource exploitation means that evidence for early extraction and processing is nearly always destroyed by later workings, though excavation on potential metal extraction and processing sites should be undertaken. There is considerable opportunity for the analysis of building stone to identify their source. Water mills and wind mills were widely used throughout the Middle Ages, and though their primary use was for grinding grain, they were also used for fulling cloth and in the preparation of metal ores. The identification of medieval mill sites, and the location of those not supsequently used in later times, is one approach tried for Anglesey which has had limited success. Questions of particular importance which remain to be answered are the date of introduction of water power, and the date and frequency of use of horizontal water wheels. Anglesey possesses millstone quarries that were used throughout the medieval period, though no systematic survey has been undertaken.





A large number of churches survive with medieval architecture and fittings throughout Gwynedd. There is a bias towards earlier pre-conquest work surviving in Anglesey, particularly Romanesque sculpture. Later medieval work is distributed more evenly throughout the area, though the school of ornate woodwork which produced roofs and screens of high quality is located more in western Caernarfonshire and Merionethshire.


The Cadw funded study of medieval churches, combined with work by RCAHMW for the Anglesey and Caernarfonshire inventories, has provided a good base record for researcher’s to use. There are a number of good 19th century accounts surviving of medieval churches which were subsequently destroyed, and also paintings and prints which survive from before the 19th century restorations. The value of using these sources has been demonstrated by work undertaken at Beddgelert Church, where interpretation of the former monastic church has been enhanced. Monastic remains survive at a number of locations, though the wealthiest abbey in the area, Aberconwy, has not survived well. Nonetheless, there is good archaeological potential at a number of Cistercian and Augustinian sites, including the first site of Aberconwy in western Caernarfonshire, which, if correctly identified at Dinas y Prif, will preserve evidence of early Cistercian settlement.

There has been a considerable amount of work undertaken on the history of the Cistercians, particularly by David Williams, but also by Lawrence Butler. Work by the same authors and Colin Gresham has been undertaken on lands owned and managed by the monastic Orders.



Though there is a relatively strong history of the study of ecclesiastical architecture in the area, there is still considerable work to be undertaken in analysing trends, identifying schools or workshops, and recognising the primary lay and ecclesiastical influences behind architectural developments. Work is required on causal relationships between major historical episodes, such as the conquest of 1283, or the uprising of Glyndwr, and architectural developments. Gwynedd, and particularly Anglesey, formerly had a large number of field chapels, but the location of many of these is not known, and their status is poorly understood. Few detailed surveys of individual churches has been undertaken, though there are excellent articles by Harold Hughes and others which provide a good starting point.


No comprehensive excavation has been undertaken on any of the church sites in the area which matches that undertaken at, for example, Capel Maelog. Thus the origins of church sites, their development and their liturgical geography are unknown.

Our knowledge of the archaeology of pilgrimages is also very poorly understood. Though not easily evidenced, work is required on pilgrimage routes, the impact upon church architecture, the use of holy wells in the medieval period and the archaeology of relics.



More comprehensive surveys of individual churches are required. A study of sculpture, moulding profiles and window tracery will help provide a firmer chronology and identify more clearly the schools of craftsmen working in the area and the spread of techniques and designs. The impact of the Edwardian conquest on church architecture is rarely discussed, yet the importation of masons and carpenters from England certainly influenced the styles of the borough churches, with a subsequent impact upon native styles. Excavation undertaken both during church renovation and on deserted church sites is required to provide information on chronological development and liturgical layout.


Opportunities for research should be sought within major renovation programmes, particularly those funded by Heritage Lottery grants. These should include detailed survey, dendro-chronological dating and excavation. Work of this nature is being undertaken at Llaneilian, though an opportunity was missed at Holyhead.

The current programme of dendro-chronological sampling by RCAHMW needs to be continued and widened. Dates for only one Gwynedd church, that at Pistyll, have been published. We need to improve our understanding of pilgrimages, and how they impacted upon the development of church architecture.


Research on monastic lands has the potential to throw much new light upon the medieval landscape, agriculture, economy, trading patterns and building traditions. Studying a single grange allows a known part of the medieval landscape to be studied as a microcosm, and a real understanding can be obtained into its working. Work on two granges in Gwynedd is currently being undertaken as part of a Cadw funded pilot project.




Paper prepared by Andrew Davidson (GAT)




This document’s copyright is held by contributors and sponsors of the Research Framework for the Archaeology of Wales.