A Research Framework for the Archaeology of Wales

Northwest Wales – Later Prehistoric






The conventional and convenient divisions of the past, based on a broad technological model and the principal raw material in use for tools and weapons, do not always correspond to social changes. At a time when bronze metal technology reached new heights of proficiency and output, other factors were at work, which provided a catalyst for far-reaching transformations. These were to characterise the next thousand years of prehistory.   A serious downturn in the climate may have provoked a retreat from marginal areas, which henceforth ceased to be viable for arable cultivation. Pressure on communities competing for the same space is likely to have increased and such pressure may have contributed to the appearance of early fortifications in the landscape. Elsewhere, similar climatic problems, accompanied by crop failures, are thought to have initiated disturbances and population movements, which sent shock waves across Europe. The last ripples of these washed against British shores. It is particularly significant that among the indicators of this phase are the first true swords and other items of serious weaponry. The adoption of iron working appears to have spread through Britain relatively rapidly, bringing with it social, industrial and economic repercussions.


The evidence for, and emphasis on, formal ritual and elaborate burial, so characteristic of the Earlier Bronze Age, falls away during the Later Bronze Age and becomes almost invisible during the Iron Age. Climate may, again, have contributed to the decline in popularity of the upland locations, which had previously been favoured for burial and ritual. In contrast, the evidence for settlement increases during this period. From the Later Bronze Age, hillforts begin to dominate the landscape. Undefended settlement is also well represented, as are field systems and other components of potentially contemporary landscapes. The quality of the field evidence in certain areas of Gwynedd is very high. The identifiable hierarchy of settlement reflects a perceived increase in the stratification of society dominated by a warrior aristocracy. It is now academically unfashionable to describe society as Celtic, even in a generic sense. Nevertheless, what we know of Late Prehistoric society in this area conforms to Professor Binchy’s archetype – ‘tribal, rural, hierarchical and familial.’ On the eve of the Roman conquest, most of Britain can be seen to be speaking a language ancestral to modern





Hillforts and related fortifications are the dominant feature, occupying naturally defensible positions from isolated and craggy hilltops to the promontories of the seacoast. Their functions are likely to have been diverse and to have included economic exploitation and interchange, defence and political control. Larger sites may have been tribal capitals with multiple functions; long-lived sites may have changed the emphasis of their use; small sites may have been little more than the defended homesteads and the power base of local lords.


Around 130 hillforts are known in the area west of the Conwy and north of the Dyfi. The areas enclosed range from less than 0.1 ha at small but defensible sites such as Bryn Y Castell and Pen Y Gaer, Aberglaslyn to around 11.0 ha at Garn Boduan and Carn Fadrun on the Llyn peninsula. The majority of forts (60%) are sited below the 200 m contour although a significant number (15%) can be found at altitudes between 300 and 400 m. The summits of Braich y Dinas (RCHAMW 1956, 85-6), now destroyed, and Tre’r Ceiri (RCAHMW 1960, 101- 3) were fortified at altitudes above 470 m OD. Enclosed, but less strongly defended, farms and other rural settlements occur and numerically must always have represented the majority. Such sites include isolated round huts, concentric enclosures, palisades, weak earthworks and the precursors of the stone walled hut groups, which constitute a major component of the Romano-British rural landscape. Some 1000 roundhouse settlements have been recorded within the region. A large number of these are single roundhouses or scattered groups of unenclosed huts, most of which (69%) occur above the 200 m contour. The largest single class of roundhouse settlement, however, and the most coherent in its classification, is the enclosed/nucleated group. With 373 examples recorded, this settlement type accounts for 41% of all roundhouse settlements in north-west Wales. The class is particularly strongly represented on Anglesey and on the adjacent mainland of Arfon. Most (60%) of these occur at altitudes lower than 200 m. This apparent differentiation on the basis of altitude between enclosed/nucleated settlements and single and scattered roundhouses suggests an economic, social or functional distinction rather than one of chronology. On the other hand there is a close correlation between the altitude distribution of enclosed settlements and hillforts in the region.


The available dating evidence, for the enclosed/nucleated group, principally diagnostic artefacts such as Roman pottery and coins as casual finds and from early excavations, has weighted the chronological balance towards the Romano-British centuries. However, more recent excavation and re-examination of earlier discoveries has shown that some at least of these settlements had their origin in late prehistory (Lynch 1991, 370,376; Kelly 1990). In north-west Wales, therefore, the landscape of the later Iron Age might be seen to comprise a hierarchy of habitation which included fortified sites within the upper stratum and undefended roundhouse settlements which one might be permitted to infer to be the houses of client farmers in a dependent relationship to the nobility (Cunliffe 1991, 260; Cunliffe 1993, 110-111; Champion 1995, 92; Kelly 1998, 29). It might be further suggested that enclosed/nucleated groups or their precursors constituted the basic farming unit and single and dispersed groups of roundhouses, generally at higher altitudes, represented a functionally complementary component of the economic regime, perhaps on the high summer pastures. The average enclosed/nucleated roundhouse settlement occupies an area of around 0.1 ha. The surviving walls of this group, generally representing the latest phase of use, are predominantly of stone. The enclosures are mostly sub-circular but the largest and most visually impressive are rectilinear or polygonal and likely to date to the Roman centuries. On Anglesey an additional class of settlement can be identified in a small group of morphologically similar earthworks for which Bryn Eryr is the exemplar. All are rectilinear earthworks enclosing areas in excess of the average enclosed settlement, of broadly comparable ground plan, occupying generally low-lying locations. It might be argued that Bryn Eryr represents a class of settlement occupying the middle ground in a hierarchy of size and status. Such sites are a product of the agricultural potential of the region and are not found in less rich areas. Recent excavations at, for example, Bryn Eryr and Bush Farm have shown chronological and structural development on the same site, continuing occupation from the Iron Age into the Romano-British centuries. Equally, the field systems that survive in association with farms have the potential to illustrate economic and land management changes through time. Excavations at Moel y Gerddi, Sarn Meyllteyrn, Bryn Eryr and Bush Farm have demonstrated the constructional changes and size variations in the development of house building during late prehistory. This is an area where chronology and detail could be refined.


Ritual and Burial


Evidence for formal burial during later prehistory is generally rare over much of Britain during this period. In north-west Wales it is very rare. At the beginning of the first millennium BC the common burial rite throughout Britain was cremation, either within an urn or in a pit. If a mound were present, then the new mounds of the later Bronze Age were consistently smaller than those of the Earlier Bronze Age, although earlier barrows could become the focus for Later Bronze Age cemeteries. Inhumation is rare and the more distinctive examples would seem to display continental affinities. Consideration should also be given, in this period, to the possibility that the remains of the dead, cremated or otherwise, were committed to the waters of rivers, lakes and marshes accompanied by the large quantity of high-status metalwork that has been dredged from, for example, the major river systems of southern Britain.


On mainland Britain inhumation burial makes an appearance in a number of areas during the 5th and 4th centuries BC. Among the most significant, numerous and regionally distinctive, are the ‘Arras culture’ barrow burials of Yorkshire. Elsewhere there are a few indications of inhumations with cart burials, as widely spaced as Suffolk, Dorset and southern Scotland. A particularly distinctive feature of the ‘Arras’ barrows is the circular, or more commonly, square ditch that encloses them. Relatively small square-ditch mortuary enclosures occur in continental later Iron Age contexts and recur in the suburban inhumation cemeteries of late Roman Britain and the rural cemeteries of western and northern Britain during the Early Middle Ages. They also occur infrequently in the burial grounds of Anglo-Saxon England. Inhumations in stone cists are known from Devon and Cornwall, Scotland and, perhaps, from north Wales, at Cerrig y Drudion. Some of these could be described as cemeteries with graves arranged in rows. Nevertheless, despite the clear presence of formal cemetery burial in certain regions – in Yorkshire and Kent during the Middle Iron Age, in Dorset during the Later Iron Age, for example - apart from individual high-status graves, it would appear that most of the population were committed to the afterlife either with little ceremony or accompanied by rites which evade the archaeological record. Articulated bodies have been recovered from disused storage pits on settlement sites and, frequently, part-bodies have been found. One possible explanation is that the initial rite may have involved exposure and excarnation.


In north Britain a few instances of formal interment are recorded, most of which are late in the Iron Age. These include crouched inhumation in short cists. Extended inhumation is seen in the north, too, in the early centuries AD. At Broxmouth, East Lothian, where crouched inhumations represent the Iron Age rite, an extended inhumation in a dug grave is dated by radiocarbon determination to the 2nd to 4th century AD. Ceremonial and ritual monuments in our area are represented only by the complex of natural lakes and pools at Llyn Cerrig Bach and the anecdotal evidence of Tacitus, describing the Roman campaign on Anglesey in the summer of AD60. At Llyn Cerrig Bach, quantities of metalwork, with many exotic pieces, are thought to have been ritually consigned to the waters at the end of the Iron Age. The analogy is with documented continental southern Gaulish practice and with the recovery of rich artefacts from pools and rivers in other areas of the ‘Celtic’ world. Natural features of the landscape such as groves, springs and rivers may have provided the focus for religious expression. Nevertheless, this should not preclude the possible existence of associated structural evidence or their adjuncts. It is in this context that the decorated stones from Trefollwyn and, perhaps, the stone head from Hendy should be seen and just, possibly, the small rectangular structure within its low circular bank at Capel Eithin.


Industry Economy and Production



External contacts had, throughout the Bronze Age, played an important role in influencing the course of British metallurgy. In the earliest phase of the Bronze Age, Irish contacts had been important in the West. Towards the end of the second millennium, cross-channel connexions had effectively created an industrial zone, which extended from north-western France across south-eastern England. These developments made their presence felt in Wales. A new and significant development, however, concerns the appearance of novel items of weaponry and horse gear of central European ‘Urnfield’’, and ultimately Mediterranean derivation, and of technically proficient, thinly cast and beaten bronze artefacts. Particularly important is the introduction of the first true swords to Britain. During the Great Orme phase of the Earlier Bronze Age, north-Welsh ores were a component of most of the metal in circulation in Wales. Production in the north-west was also high, to judge by the surviving evidence. Towards the end of the Earlier Bronze Age, however, the focus of ore extraction may have shifted south to Aberystwyth and mid-Wales. While it is possible that the Great Orme mines continued in production during the Later Bronze Age, their significance appears to have waned considerably.


The products of the earlier part of the Later Bronze Age, the Penard phase, and its successor, the Wilburton phase, are very poorly represented in the north-west. The Penard phase has been described as a crisis period. There is significantly more bronze in circulation during the 9th - and 8th -centuries or, at least, a greater survival of the products, and a reliance on recycled scrap for raw material. In some areas, notably South Wales, a high proportion of lead was added to the alloy. A certain amount of lead facilitates the casting of the thin walled and complex artefacts in use during this period but very high concentrations suggest a shortage of other components.


During this period north-west-Wales still maintained contact with Ireland as evidenced by the considerably greater number of gold artefacts of the Irish Dowris phase than in any other region in Wales (for example, Pigeon's Cave, Great Orme). The earliest iron artefacts appear in Wales (at Llyn Fawr, Glamorgan; and in Denbighshire) at a time when the latest full bronze industries are still in production, during the 8th -century BC. The objects, a socketed axe, sickle and a spearhead are essentially bronze forms recreated in the new metal. We must assume that iron became the metal for all major tools and weapons quite rapidly. Nevertheless, with the exception of a single iron ring-headed pin from the hillfort of Din Silwy, Llanfihangel, Anglesey, there is a complete dearth of evidence for both bronze and iron artefacts in north-west-Wales during the earlier part of the conventional Iron Age. It is unclear in the present state of knowledge how real this apparent situation is.


Considerably more material is known from the Later Iron Age, in its Bryn Eryr - Bryn y Castell phase, between the 3 rd-century BC and the 1st-century AD. Even then the greater part of our evidence derives from a handful of key sites, in particular, the votive deposit at Llyn Cerrig Bach (48 bronze and 63 iron artefacts) and the major iron smelting and smithing workshops at Crawcwellt (an estimated 23 smelting furnaces producing about 0.5 tonne of fully refined bar iron), and Bryn y Castell. Iron is generally reserved for tools, weapons and robust items. Bronze is the metal of choice for fasteners and fittings where intricacy of form, detail and decoration were more important than strength.



Field systems survive as above ground features in areas that are now outside the zone of intensive agriculture. Particularly good examples occur on marginal land in Arfon and Ardudwy. Some occur in clear association with settlement and provide evidence for droveways, tracks, paddocks and, possibly, where there is lynchet formation, for arable cultivation. Reuse, overlay and replacement of boundaries can also be seen, however, and the relative chronology of the association of multiphase boundaries to multiphase settlements is generally unclear, as is indeed, the absolute chronology of these features. Saddle querns and mortars are indicative of the processing of grain and other foodstuffs and cereals are a component of the environmental evidence at the very limited number of sites where such evidence has been recorded. The evidence of animal bone is even less well represented. At Bryn Eryr, both cattle and sheep were present. At Segontium, during the Roman period, cattle were an important component of the food resource, drawn presumably from pastures within the region. Cattle, particularly grazing the upland pastures of the major river valleys, continued to represent a very important economic resource during the earlier Middle Ages. This aspect of the agricultural regime is perhaps under-represented or not fully recognised in the evidence for later prehistory.


External contacts or lack of them


Axes are one of the diagnostic indicators of regional metalworking traditions during the Later Bronze Age. Later Bronze Age metallurgy in north-west-Wales would appear to have been conservative, retaining a preference for the traditional palstave, albeit in a distinctive late form, rather than developing a local socketed type. Nevertheless, imported or intrusive socketed forms characteristic of other regions, occur, and it is particularly significant that north-west-Wales has more Irish gold than any other region of Wales, indicating a contact across the Irish Sea which is a recurring theme through later prehistory and early history. The different permutations of metalwork assemblages across Wales have been seen by some to identify a regional distinctiveness, which presages the tribal divisions of the Late Iron Age. There is, on present evidence, no iron of the very earliest Iron Age in north-west-Wales to compare with the admittedly limited evidence elsewhere. There is, however, an iron axe from the Berwyns (Denbs.). There is also an important group of bronze buckles, rings, pendants and attachments from Parc Y Meirch, St. George, close to the hillfort at Dinorben, Denbighshire. These are of continental north-European derivation and are indicative of the more general appearance in Britain, circa 10th-9th century BC of bits, bridles, horse brasses and the fixtures and fittings of wheeled vehicles. Concentrically ridged and looped buttons, and also rings, from Llangwyllog and Ty Mawr, Anglesey may be from functionally related equipment. There is also a beaten bronze ‘bucket’ from Arthog, Merioneth, which displays a north- European continental origin. The bronze knobbed and hinged collar from Clynnog is a continental piece of, perhaps, 4th-3rd century BC date. Certain settlements, for example, Bryn Eryr, were in receipt of salt from the northern Welsh Marches during the later Iron Age, evidenced by the distinctive ceramic containers (VCP) that brought the commodity. Tacitus (Annals XIV, 29) states that the island of Anglesey had provided a refuge for fugitives, presumably from the Roman advance. There may have been similar instances, at other times, resulting from undocumented conflicts. Some material in the Llyn Cerrig Bach (Anglesey) hoard, on typological and metallurgical grounds might be suggested to derive from an original context in the late Iron Age of southern Britain and there is a smattering of other items. The inhumation burial from Gelliniog is another example of ‘external contact’, in this case in the individual’s person, rather than by proxy. Two groups of bronze bowls from the coast near Barmouth, found in the mid-19th century and recorded separately, may be from the same or similar hoards. The collection includes ornamented handled cooking pans and thin bowls. They appear to be pre-Roman Italian imports. Two Roman Republican coins from Anglesey (Tal Gwynedd and Croes Allgo) may have entered circulation in the region at this time as old issues included in the large injection of cash, which financed the military push northward in the Flavian period. A Flavian hoard of 39 coins from Llanfaethlu, Anglesey, contained 20 Republican issues.


Research Strengths and weaknesses


The strength of the current situation is that there is a framework in place for the compilation of a comprehensive data set and for this data to become freely available through the Regional Sites and Monuments Record. The weakness of the current situation is that the resources now being applied are not adequate to the task of maintaining and enhancing this record. It is a strength that a pan-Wales programme of thematic assessment (of, for example, hut circle settlement) grant-aided by Cadw: Welsh Historic Monuments, allows, in addition to threat related and management considerations, a revision of aspects of the data set. It is a potential weakness that research considerations are incidental to, rather than determining factors in, the selection of such themes.


Research priorities



It is unlikely that all types of defended settlement are functionally related or that they occupied the same contemporary landscape. An absolute chronology across the range of defended settlement is urgently needed, as is an assessment of the diversity of function. Some progress has been made in establishing a chronological sequence for certain individual non-defended settlement. More work is required here, as is an assessment of the relationship between defended and contemporary non-defended settlement.


Burial and ritual

The evidence is largely absent. Where do we look for it? In later prehistoric Ireland we might expect cremation within the circuit of a ring ditch perhaps under mounds or in the ditch itself. Significantly, certain ritual monuments of an earlier past are known to have provided a focus for such burials. Elsewhere in Wales both cremation and inhumation are known, but not in great numbers. The appropriation of, or focus on, earlier ritual monuments is also a feature of Welsh later prehistoric burial at, for example, Stackpole, Pembrokeshire and Plas Gogerddan, Ceredigion. This association of burial with earlier numinous locations is an enduring theme, which continues into the Early Historic period as a determining factor in the location of a significant number of 6th-9th-century AD extended inhumation cemeteries. The reconstruction of the ritual and funerary landscape of Later Prehistoric Wales as a counterpoint to the extensive evidence for settlement is as much a research priority as is an understanding of the settlement base in Earlier Prehistory where most of our knowledge derives from burial and ritual.


Economy and production

Establishment of the chronological and landscape association of field systems to defended and undefended settlement is a research priority. The gathering of plant macrofossil (from waterlogged contexts), pollen and animal bone (from limestone areas?) evidence for the agricultural economic base and, in particular, to identify topographic or chronological distinctions in the agricultural regime is a very desirable objective.


Peter Crew’s groundbreaking work on the process and logistics of iron production is an example of the persistence required to achieve meaningful results. Extension of technological analysis of this kind in the elucidation of the transition for conventional ‘Late Bronze Age to Iron Age’ is very much a priority. Peter Northover’s work on the analysis of copper alloys, identifying the chronology of the change in emphasis in the sources of raw material now requires the identification of production sites on the ground, through fieldwork.


External contacts

Are we shy of acknowledging the contribution of direct external influences in shaping the cultural landscape of Later Prehistoric Wales? It has recently been suggested, as ‘a reasonable assumption’, that the population of Iron Age Wales was ‘largely indigenous, subject to natural increase and not augmentation by periodic immigration’ and that ‘the cultural base of Welsh communities in the first millennium BC … has its origins in a protracted phase of social and economic change and interaction affecting communities whose roots lay firmly within the region’. No doubt, in large part - but at some stage in the prehistory of the British Islands, this indigenous population learned a new language, ancestral to modern Welsh, and they learnt it directly. If not during later prehistory, then when? If so, can we detect the influence of inward migration in the archaeological evidence? The evidence of succeeding periods across Europe catalogues unremitting migrations, some so recent as to be familiar to living memory. There will certainly have been gift exchange and trade. What then are the mechanisms by which goods are traded? Historical analogy would require there to be an invariable aristocratic control or sanction. Gift exchange implies diplomacy, alliance and a social hierarchy.





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