Home > Projects >


The Early Medieval Period is of crucial importance to our understanding of the development of Wales as a nation and Christianity as practised throughout Great Britain and beyond. Furthermore, it is a period that has had a profound effect on the landscape, in both physical terms, through memorial stones, wells and chapels, and social terms, through the foundations of modern settlement patterns and administrative boundaries. However, it is a little-understood period within Welsh history, for a variety of reasons. These include the attitudes of researchers in relation to political/social trends, the paucity of detailed documentary evidence and reliable archaeological evidence, and the concomitant development of myths. Most importantly, it is a period that has only recently been subjected to the degree of rigorous study that it deserves. The result has been an imbalance in the archaeological record and on the Schedule of Ancient Monuments. The present project forms part of a pan-Wales project similar to that undertaken on the historic churches of Wales. It sets out to redress some of this imbalance.


April 2001-March 2002

The first year of the project comprised a desk-top assessment for the three counties of Carmarthenshire, Ceredigion and Pembrokeshire, assessing total of 1451 known ecclesiastical sites. The assessment resulted in a list of 323 sites with potential for early medieval origins - 154 in Pembrokeshire, 108 in Carmarthenshire and 61 in Ceredigion. The sites fall into two main categories: -

  • Church/chapel sites, ie. 'developed' cemetery sites

  • Open cemetery sites, ie. 'undeveloped', never having received a building. These include a number of Bronze Age round barrows and Iron Age enclosure re-use sites.

Few ecclesiastical sites in the three counties can be positively dated to the early medieval period. For example, the tradition of long-cist burial appears to have persisted well into the post-Conquest period, as in the excavated cemetery at Eglwyswrw (Pembrokeshire). However, a number of sites possess one or more characteristics that may be significant indicators of early medieval origins. The criteria include:

  • Documented pre-Conquest churches.

  • Clas or portionary church evidence.

  • Direct or indirect archaeological evidence

  • Early Christian Monuments, in situ.

  • Artefacts of definite early medieval date.

  • The presence of a saint's grave or capel y bedd.

  • Siting of churches within Roman forts etc.

Indicators such as churchyard morphology, British (ie. 'Celtic') dedications, place-name elements such as eglwys, llan and merthyr are treated with more caution and do not constitute criteria in themselves.

However, it must be stressed at the outset that only 30 sites have demonstrable pre-Conquest origins. Of these, only five have been absolutely dated by C14. The remaining 25 sites are known from documentary references and the precise locations of only some can be currently proven with any finality. Nevertheless, the assessment demonstrated that many traditional indicators do seem to apply, ie. circular churchyards, in situ ECMs and native dedications are often accompanied by other evidence for early medieval origins. Re-use of Iron Age enclosures, and large, outer churchyard enclosures were suggested at a large number of new sites. Some of these outer enclosures appear to variously incorporate Neolithic chambered tombs and Bronze Age standing stones, but this is apparent at only a small number of sites.


April 2002 – March 2003 Pembrokeshire

Year 2 of the Early Medieval Ecclesiastical Sites project comprised map research, aerial photograph analysis and selective field visits to the 154 sites, in Pembrokeshire, for which early medieval ecclesiastical potential emerged from the Year 1 desk-top study. A large number of these sites produced evidence for associated cropmark or earthwork features. However, the context and date of many of these features is at present unknown.

Some early medieval ecclesiastical sites in Pembrokeshire can be identified from the study of later medieval source documentation. The survival of clas churches as ‘portionary’ benefices, divided between the church and powerful laymen or ‘portionaries’, recognised in north Wales, appears to be paralleled in Pembrokeshire where many more churches were held in a variety of forms of multiple patronage into the modern period. A number of ecclesiastical holdings are described as ‘clas tir’ in 15th-16th century sources, when recollections of sanctuary or noddfa were also recorded. In most cases, the sources supplement existing evidence for early origins, or can be used in conjunction with other, more circumstantial evidence. In other instances, they have led to the identification of new sites. The terms of a number of early 12th century grants of churches, both Welsh and Anglo-Norman, make it clear that they refer to pre-existing ecclesiastical sites and, in some cases, allow the pattern of pre-Conquest ecclesiastical dependence to be determined; the terms of one source may be a reference to an otherwise unknown mother-house, possibly monastic, at Little Treffgarn in St Dogwells parish, which was later the temporary site of the Whitland Abbey community. A number of late 11th century churches were later recorded by Gerald of Wales, among others, and variously refer to surviving, abandoned or relocated sites, including a chapel at Upton - which may be represented by a circular enclosure at ‘Churchfield’. Other early 12th century sources suggest the displacement of ecclesiastical establishments, from sites which can occasionally be identified.

Eglwyswen churchyard, Pembrokeshire: aerial photograph of cropmark/outer enclosure

A number of putative early medieval church sites appear to have occupied a larger outer enclosure, usually - but not always - circular, and concentric around a smaller, inner enclosure. A large number of such double enclosures, many occupied by surviving churchyards, have been identified in Pembrokeshire. They are normally represented by cropmarks but occasionally by standing earthworks or field boundaries. In a number of cases, the inner enclosure appears to have survived to become a formal churchyard. In a very small number, the outer enclosure defined the extent of the later churchyard. The morphology of a number of these sites suggest that they may have their origins as iron age domestic enclosures of the ‘concentric antenna’ type, a class of monument becoming increasingly recognised in southwest Wales. In others, such re-use may have provided a model for de novo ecclesiastical enclosures. The outer enclosure may have had a variety of functions - as a cemetery, as space for an additional chapel - the sites of capeli-y-bedd (or founder’s grave chapels) have been identified within some - as land for agricultural plots, as an area of sanctuary/noddfa, or to define the limits of ecclesiastical ownership.

Ludlow N 2003 Early Medieval Ecclesiastical Sites Project: Pembrokeshire 2002-2003, Dyfed Archaeological Trust
Report No. 2003/39.


April 2003 – March 2004 - Carmarthenshire and Ceredigion

Year 3 of the Early Medieval Ecclesiastical Sites project comprised map research, aerial photograph analysis and selective field visits to the 169 sites (115 in Carmarthenshire and 54 in Ceredigion) for which early medieval ecclesiastical potential emerged from the Year 1 desk-top study.

Archaeological evidence for early medieval ecclesiastical activity is, in general, more sparse in Carmarthenshire and Ceredigion than in Pembrokeshire. For instance, burial in stone-lined graves or ‘cists’, so widespread in Pembrokeshire, is almost absent in these two counties. Nevertheless cist burial, although primarily associated with the early medieval period, continued to be practiced throughout the Middle Ages.
Other themes apparent in Pembrokeshire continued to be observed in Carmarthenshire and Ceredigion. More double enclosures, many occupied by surviving churchyards, were identified, represented mainly by cropmarks but occasionally by standing earthworks or field boundaries. As in Pembrokeshire, some sites produced evidence for associated - but undated - cropmark or earthwork features, but these were fewer in number. The discovery of Llandeilo’s Roman fort in 2003 allowed the origins of both Llandeilo and Llandyfeisant churches to be seen against a background of persistent prestige attached to the site, similar to the relationship between Carmarthen’s early medieval churches and the site of the Roman town.

The analysis of both contemporary, and later medieval source documentation was perhaps more effective in Carmarthenshire and Ceredigion and allowed a significant number of early medieval sites to be identified. Some of these, eg. Llandeilo Rwnws, Llanegwad, Carmarthenshire, may be associated with cropmark evidence. The pattern of pre-Conquest ecclesiastical dependence, and its development over time, could also be suggested in the parishes of Llandeilo Fawr and Cynwyl Gaeo, while the displacement of ecclesiastical establishments was noted in Newchurch parish, and perhaps within Kidwelly parish.

Aerial photograph of Llangan churchyard, Carmarthenshire, showing cropmark enclosures and possible structures


April 2004 – March 2005 – publication

This represented the culmination of the Early Medieval Ecclesiastical Sites Assessment Project. A paper, summarising the results of the project, was read by the project officers from each of the four trusts at the ‘Archaeology of the early medieval Celtic churches’ conference at Bangor on 9-12 September 2004.

This was jointly hosted by the Society for Medieval Archaeology and the Society for Church Archaeology. The proceedings, which are currently being edited, will be published in Medieval Archaeology. The paper by Dyfed Archaeological Trust was entitled ‘Identifying the Early Medieval Church in southwest Wales’ and examined the methodologies employed during the project, a summary of its findings and a review of the significant new discoveries.

The well-evidenced early medieval ecclesiastical sites of southwest Wales


Early Medieval Cemeteries in Pembrokeshire


Project contact: Ken Murphy



[click for navigation menu if not present]