Abandoned Landscapes - Chronological Change: Mesolithic to the 16th Century Wales

Submitted by M_M_Robinson on Friday, 07 July 2006 Page Views: 5733

Country: Wales

**Conway Mountain** submitted by TimPrevett : The entrance of the citadel part of the hillfort, viewed from the outside. Obviously, this rampart would have been considerably more tidy when in use, with the entrance (just left of centre) topped with wooden walkway and guarded by gates. (Vote or comment on this photo)

PART 1
CHAPTER 2
THE ARCHAEOLOGY OF THE HILLS: A SELECTIVE OVERVIEW
CHRONOLOGICAL CHANGE: MESOLITHIC TO 16th CENTURY DAY

We have already seen that the Mesolithic hunter-foragers manipulated their environments, cleared land and probably managed herds, and their strategies for the diverse environments of coastland, woods, cleared upland and mountain were probably very effective.

Table of periods

Dates Cal. BC (Lynch et al 2000 1 and Manley et al 1991)
Mesolithic = 8000-4400
Earlier Neolithic = 4400-2900
Later Neolithic = 2900-2300
EBA = 2300-1400
MBA = 1400-1100
LBA = 1100-700
IA = 700-AD43
Roman = AD43-410 (omitted as incomplete)
Early Medieval = 410-1283 (omitted as incomplete)
Later Medieval (Post Conquest) = 1283-1536
Post-Medieval = 1536-18th century
Modern = 18th century to present

Mesolithic

Whilst there are no visible sites or monuments of this period, in a sense the Mesolithic landscape is still accessible to us as it forms the template on which the later landscape was modelled. Current thinking is that North Wales as a whole supported a population of only a few hundred people occupying a series of territories based perhaps on the main estuaries (Lynch 1993 16-17; Brassil 1991). These communities are thought to have adopted a seasonal cycle of resource exploitation, with the young men making summer-season hunting, fowling and fishing forays along the river valleys into the uplands perhaps following red deer or wild cattle [note 1].

There is evidence for the deliberate and controlled use of fire in the uplands which could be to create woodland clearings thus encouraging browsing animals, and to clear linear routeways and lines of sight to monitor game movements. It perhaps represents the beginning of the management and control of herd animals although probably not domesticated stock. Such clearances of the natural tree cover coupled with an increase in rainfall during the middle Mesolithic is thought to have begun the creation of the peat moorland environment which covers so much of the area today (see Moore-Colyer 1998, 27 - Moel y Gerddi Ardudwy (625323) and the Berwyn (Caseldine 1990 35; Brassil 1991, 49; Olding 2000 78)

Using evidence from the Black Mountains in South Wales, Tilley identified the kind of location preferred for upland activity - the edges of small lakes, river terraces, places with views across valleys, near passes and the heads of valleys (Tilley 1994 111-7 cited in Lynch et al 2000, 30, and see also Bruck and Goodman 1999 12-13). This period may also have seen the origins of the ritual deposition of artefacts at special places in the landscape, probably intended to mediate between natural and human environments (Bradley 1998 xv1-xviii).

Sites in the uplands are rarely discovered as they tend to consist of scatters of flints indicating the location of temporary hunting camps or kill sites. Where they have been found it is by accident or during the excavation of a site of later date. An important example is from the Brenig uplands where pits and hearths were found beneath a Bronze Age monument. Interestingly the finds assemblage indicated year round occupation rather than a 'hunting kit' (Lynch 1993 32, see also Llyn aled Isaf. in Lynch et al 2000, 300 and Brassil 1991) [note 2].

The Earlier Neolithic Period c.4400-2900 BC

The economic changes of the Earlier Neolithic appear to have entailed continued and more widespread upland clearance, probably to sustain largely mobile (nomadic) communities dependent on domesticated stock, small scale agriculture and also hunter/foraging (Caseldine 1990 46).

The Neolithic cultural 'package' or 'repertoire' also introduced large communal burial monuments, several of which survive today and have been used to define three of the four 'core settlement areas' set out above. The chronology, details and cultural affinities of the various tomb types have been discussed elsewhere (Lynch et al 2000 63-77; Lynch 1976 and 1989). Products from the Graig Lwyd axe-making centre (probably first utilised in the Mesolithic) were being widely distributed (RCHMW 1956 xli-lvii). There are several isolated axe find-spots many of which were probably ceremonial offerings at special places on or near the corridors of movement. A probable Early Neolithic timber long house has been discovered at Moel y Gaer hillfort (north west of Mold at 2169, Britnell 1991, 55). The fact that this isn't associated with a 'core' area demonstrates that settlement must have been more widespread than is indicated by the distribution of known tombs, and indeed some communities probably didn't build them.

We don't at present know whether the transition from Mesolithic to Neolithic was the result of 'indigenous' population gradually adopting the new cultural package or if they were more or less replaced/displaced by colonisers (perhaps from Ireland). The coastal and estuarine zones adjacent to upland grazing were the favoured zones for both systems so one can argue the case either way [note 3]

The Later Neolithic, Early and Middle Bronze Age. c.2900-1100 BC

It is generally thought that during this period there was a reduction in mixed farming regimes with a greater emphasis on pastoralism. An apparent improvement in climate by 2-3 degrees meant that the uplands became more attractive, and before 2000 BC the vegetation cover was usually sedge/grass moorland rather than forest, although there are continuing signs of soil degradation (Caseldine 1990 55; Lynch 1991; Moore Colyer 1998).

Current ideas for settlement in the British Isles at this period play down the idea of a fixed, settled landscape in favour of residentially-mobile households with pastoral and shifting agriculture without specific 'homes'. Lineages would hold tenure over a large area and move around this, coming together periodically in large groups at special sites such as tombs, sanctuaries and flint mines) (Bruck 1999). In this model there would be no central or paramount settlement site, and occupation sites would be ephemeral and difficult to identify archaeologically. This strikes a chord with the situation in North Wales, where there is a paucity of evidence for settlement or house sites dated to this period.

Only one Early Bronze Age settlement site has so far been confirmed in North West Wales, at Llyn Morwynion, near Ffestiniog (738425). This lies at about 400m OD and consists of a stone hut circle c.6m wide, dating from between 2,400-2000BC. Nearby, a stony bank (wandering wall) another hut circle and some stone platforms (thought to be the remains of other huts) were found. Later sheepfolds had been built on two of the platforms (Caseldine et al 2001). Although there was no evidence for early cereal production at this site, it may be that at other locations the improved climate and expanding population encouraged mixed-farming 'assarts' in the uplands - similar in some ways to the failed cultivation episodes of the 18th and 19th century - which were unsustainable for more than a few years due to soil exhaustion [note 4].

There is little tangible evidence for the organisation of land and the territorial scale of society as there are no identifiable fields or boundary systems. In terms of technology, metal was introduced into the region after c.2400 BC. There is a very early hoard from Moel Arthur in the Clwyddians (adjacent to the pass which has a stone circle, barrow and standing stone at its foot on the eastern side (1768)), and North Wales was in the forefront of metalworking experimentation in Britain. The internationally significant copper mines at the Great Orme started production around 1800 BC [note 5].

Monuments

In contrast to the lack of settlement and subsistence evidence for this period there is a wealth of religious, ritual and burial sites - sanctuaries, cairns and barrows, standing stones and ceremonial routeways. For clarity I will consider each category in turn:

Sanctuaries

‘Sanctuary’ is a term I use in connection with one or more sites which indicate a significant focus for religious or ritual activity [note 6]. Stone circles and cairn circles are commonly present with standing stones and cairns frequently in proximity, and more rarely stone rows and avenues. There is also a recurring association with human remains, some arguably originating from sacrificial ceremonies. They seem to appear in the Late Neolithic and may continue, with important modifications, additions and reuse into the Late Bronze Age. There is no magic formula about their location and distribution (Burl 1999), although they tend to be confined to the 'core' areas where there is a tradition of settlement, activity and monument building (e.g. Moel Goedog, Maen penddu, Druid's Circle, Tan y Braich). However, they do not focus on the much more ancient Neolithic tombs.

Although many now seem isolated and secluded, in fact they were located with accessibility in mind, often being beside or close to pre-existing and significant routeways (e.g. Brenig, Moel Goedog, Druid's Circle, Tan y Braich), routeway junctions (Maen Penddu) and perhaps near territorial boundaries (Afon Disgyfna, Brenig, Maen Penddu) (See Bradley 1998 121). Indeed, the concept of linear sanctuaries, where monuments are clustered along the line of a 'trackway', possibly for processional purposes, seems to be applicable in at least three cases (Druid's Circle, Moel Goedog, Cadair Idris).

Another fascinating characteristic is the pairing of monuments, where similar ritual sites reflect each other on either side of a boundary, in an ambiguous relationship signifying both separation and unity. The examples I can think of are the pair of cairn circles on either side of the Moel Goedog trackway (sited so that only one is visible from each direction of approach, the stone rows of Hafod y Garreg and Hafodty Hafod Dre on either side of the watershed in the Denbigh Moors, and the Moel Ty-uchaf and Tyfos cairn circles on opposite sides of the River Dee, but I'm sure there are more instances.

Some sanctuaries may have been deliberately founded on an occupation site, perhaps associated with an important ancestor e.g. Brenig cairn 51 and the kerb cairn, Cefn Caer Euni (Lynch 1993 and 1986, 83-7) and also the Hengwm circles (Crawford 1920 107-111) [note 7].

A number of sanctuary sites are easily recognisable from the upstanding monuments whilst at others, perhaps an isolated standing stone or denuded cairn may in some cases be the only visible indicator for a long history of feasting, ritual, ceremonies for the dead, burial and other communal purposes before and after the construction of the monument. The recent discovery of preserved fragments of worked timber at an Early Bronze Age sanctuary site at Trostrey, Monmouthshire gives a glimpse of what has been lost in structural terms. These appeared to form part of an elaborate wooden structure reminiscent of a small temple, with carved horn-like terminals and indications that it had been rendered with white and red pigments (Mein 2000)].

Sanctuary sites probably had a variety of uses and functions. Small ones may have served single family groups in a small locality. However, their frequent proximity to inter-territorial routes and boundaries indicates they may have had a wider function, perhaps connecting dispersed family groups of the same lineage (possibly replacing a role of the earlier communal tombs). The larger sites such as the Druid's Circle complex may have been 'neutral' places serving a whole 'core area' where different groups could gather for special ceremonies and for inter-communal and long-distance exchange of prestige goods, stock and marriage partners.

The way in which certain monuments seem to reflect or incorporate surrounding landforms, creating a 'power of place' is another interesting feature, although how much this is a modern construct is difficult to evaluate (Bradley 1998 122-3). Examples of this are the circle and row at Rhos y Beddau, which may mirror the shape of the valley in which it sits, and Bryn Cader Faner which appears to reflect the main Snowdon range behind it. Stone 'rivers' or 'waterfalls' near ritual monuments may also have been significant in some contexts (e.g. Rhos y Beddau and Cerrig Pryfaid circle), or even have had their own independent significance [note 8].

Cairns

Cairns and barrows are the most commonly encountered prehistoric monuments in these hills. Traditionally they have been interpreted as a type of burial site which marks a transition from a broadly egalitarian Neolithic society, symbolised by the communal burials in the megalithic tombs and chambered cairns, to a more hierarchical social structure in which individual leaders, perhaps chieftains of warriors - were commemorated in single burials. Nowadays there is a recognition that although a prominent personage (e.g. a local leader or 'shaman') may have been the primary burial, this is only one element in often complex monuments, perhaps used for burial and other ritual activities over several centuries, forming part of the wider communal geography. Where excavated they are commonly found to have been built on house sites (perhaps belonging to the deceased individual or their family), and special features such as stake rings and ritual pits and deposits are frequently found below the cairn, as well as secondary burials.

When you look at cairns or barrows on a map or notice them in passing they can appear to be randomly scattered, with many along the high ridges. However, when seen in their landscape setting, general patterns emerge in their location which can tell us something about the communities which erected them. Many, perhaps the majority, were on low ground and have now been cleared or ploughed out. In the uplands, where survival is better, three main types can be identified:

1. Ridge/summit cairns
2. Pass/route-marker cairns
3. Upland valley or hillside cairns.

Ridge/summit cairns.

These form a small but prominent group and are thought to be the earliest type, annexing prominent positions as statements of territorial presence or control. With the expansion of population and improved climate in the Late Neolithic or Early Bronze Age encouraging more intensive settlement and usage of the uplands, ways of negotiating, establishing and legitimating access and grazing rights to the uplands were needed. The appropriation of the main watersheds by families and clans may have been part of this process. Once the highest and most dominant locations were occupied, perhaps by the founder of a local lineage, a process of infilling of less significant locations followed to form linear ridge-top cemeteries (e.g. the Berwyn ridge, Cadair Idris).

In an important case study of the Denbigh Moors, Lynch identified three widely-spaced prominent cairns, which although not big, are intervisible, and are also visible from many other cairns. She suggested these were founder-cairns, possibly connecting territories, with the implication that there was concerted formal marking of the landscape across a wide area. (Lynch 1993 36-7 and also see Moore 1995 211-2). This seems to me a profound observation: The master-cairns are not primarily looking down on their territories and on the living, but across to the founding dead of adjacent communities in an interlocking regional cemetery on a scale of tens of miles. As with the instances of pairing of other ritual monuments (see above) they signal both separateness and connection. For obvious topographical reasons I think it unlikely that this category of monument was associated with house-sites although the row of three on Moel Faban might be an exception [note 9].

Pass/route-marker cairns

These are a very important but little recognised grouping. They may well be as early as the ridge top cairns (the only example excavated in modern times was at Brenig, which was dated to 2,650 BC) and have served a similar function as territorial signifiers. No burial was found at the Brenig cairn but other examples certainly display either indications of a cist (as at Bwlch Ehediad) or of robbing (e.g. Wilderness barrow above Llangollen). I suspect that future investigation may demonstrate that several were founded on house-sites, but given the high and often exposed locations these must have been temporary summer residences at the high point of the nomadic or transhumant seasonal cycle, where they would have offered good locations for overseeing of the stock.

It is striking that a good proportion seem to be on routeways (as might be expected if they were originally house-sites) where they could be seen by the trader, traveller or visitor as well as the locals. As special, perhaps liminal, places they may have been the focus for meeting, rituals and feasting.

Interestingly they are rarely sited on the actual crest of the pass but favour one side of the watershed. The consequence is that they would only have appeared imposing (e.g. as a skyline features) to those approaching from that side. This may be because they had to be within a specific territory rather than on the (possibly sacred) nomansland of the actual boundary. Where passes lie at the interface between territories rich in monuments and emptier zones, the cairns are often on the former side, and may thus have marked the threshold to those leaving for the outside world (e.g. Bwlch y Ddeufaen, Brenig, Bwlch y Rhigyr, above Pont Scethin, Cwm Cowlyd) [note 10].

Upland valley or hillside cairns.

Surprisingly, the majority of visible upland cairns are not on the tops or passes but are located in valleys, cwms and on hillsides and moors. More certainly than the other categories, these must usually be associated with settlement, and many probably overlie specific houses. If so, this gives us a magical insight into the distribution pattern for the invisible settlements of the period. Such data can be interpreted in various ways, but the apparent lack of associated fields and enclosures leads me to think they generally mark the sites of temporary summer stations. The fact that they sometimes lie in sequence up the valleys and into the high cwms, suggests each site may have been occupied for only a few days or weeks each year until the grazing was used. [note 11]

Sometimes we find persistent use of these locations in the form of nearby stone-built settlements (presumably of later date) (e.g. north of Foel Dduarth in the North Carneddau, Cwm Cowlyd, the Hengwym cairn), whilst many others are stranded in nomansland with no sign of persistent use, perhaps reflecting changes in husbandry practice and/or population levels.

The concentration of cairns and cists on the moor above the sanctuary zone around the Druids Circle area at Penmaenmawr is perhaps something different again. Many of these are beside the main trackway or its subsidiaries, and possibly reflect a denser, static population, or an ancestral gathering place favoured for burial.

Appearance of the cairns.

What did they the cairns look like in their heyday? Most certainly not the denuded, vandalised remnants which survives today. There are a number of types, some rare like the Kerb cairns which may have had a border of upright, recumbent or leaning stones (most evocatively at Bryn Cader Faner and Hengwym, although there is dispute about their original appearance), or have small flat stones laid in a continuous ring (e.g. near Druid's circle). Kerb cairns also often have inner rings of concentric stones.

Cairn-circles such as Moel Ty-uchaf are another unusual and impressive class of monument. They are sometimes interpreted as denuded cairns but its more likely that the stones were intended to be seen rather than buried. However, the majority of cairns are now undifferentiated heaps of rubble. Remnants of stone burial cists, originally containing ashes in a vessel or loose, can frequently be seen protruding from the centre, but their presence was not invariable. In terms of size, they tend to cluster conveniently into three groups with circumferences of 30, 60 and 90 yards (Hemp 74-6).

It is thought that the barrows in the southern chalk lands of England when newly constructed would have been fresh, gleaming white chalk standing out from the surrounding green turf and visible for miles. How might the visibility and impact of the upland cairns have been enhanced? Perhaps they were adorned with totems, offerings, or organic decorations now undetectable. But a key feature for many cairns, especially those on ridges or viewpoints was not their colour, material or contrast, but their distinctive silhouette against the sky when viewed from some distance. The most likely form for many examples was probably similar to the large reconstructed cairns on Drygarn Fach in the Elan Valley of mid-Wales which have been rebuilt as cylinder shapes 3-4m high with an outer coursed drystone retaining wall infilled with rubble (SN863584). These make most imposing monuments. Robbing or collapse of the retaining wall leave the usual stone pile which is all that survives of many cairns. A variety of other shapes and finishes probably also existed in accordance with local custom.

Other Monuments: Stone circles, standing stones, rows and rock markings.

Stone circles are frequently present at sanctuaries, and have been interpreted as arenas in which the community would gather for ceremonies (not necessarily funerary) . Consequently, a connection between circle size and the local population density has been postulated. If true, the generally small circles of North Wales would indicate relatively low populations, as we might expect (Burl 1999.xx). It may also be that each stone was set up by particular families and could actually represent individual ancestors (cf. Pryor's interpretation of causewayed camp ditch segments).

Stone rows and avenues appear to have developed as ceremonial elaborations to circles in the later Neolithic, eventually becoming architecturally distinct forms. They are very rare in these uplands; the portal stones at the Druid's Circle are thought to shows the origin of the form (Late Neolithic), whilst the site at Rhos y Beddau is a very unusual example of an avenue associated with, but detached from, a stone circle (Early Bronze Age). Waun Oer (beside the Cadair Idris trackway) is a short row (probably Middle Bronze Age), whilst at Bwlch y Ddeufaen are a pair of standing stones (Late Bronze Age). (Burl 1993, 268).

Single standing stones are usually thought to be Bronze Age. In the past they were often thought to be the remains of denuded cairns or route markers. However, where excavations have been carried out on them elsewhere in Wales they tend to reveal a much more complex and important ceremonial role. Features which have been discovered around them include pits (some containing human remains), large and small free-standing wooden posts, ritual deposits including charcoal, human remains (also found in the stone-socket itself), and soils brought from elsewhere. Some stones are associated with cist burials, or with circular ditches or stake-and post-settings (?possibly houses). Thus we can begin to see that the isolated monument poking above ground today was only part of a succession of features and activities at the site, perhaps taking place over several hundred years (Williams 1988 23-34). Based on this evidence, a role as simple route markers can be dismissed. It is true they frequently occur on routeways, but this is usually on well-defined stretches close to other monuments where there is no need for guidance, whereas ironically, on the wide moorland tracts where markers might be needed they don't occur. Comparison with wayside shrines in later societies may be more appropriate [note 12].

Rock Markings are very rare in the region. The most interesting but elusive are a small group of less than half a dozen in the area above Llanfairfechan. Previously known as 'Arrow Stones' because it was thought they were caused by the action of sharpening metal arrowheads on them (Bezant Lowe 1912 166-8), nowadays such features would be interpreted as a simple forms of rock art, marking the landscape, probably on or near places of passage. Three occur near a well-attested ancient trackway and another one is on a possible route from this track to the coast at Llanfairfechan [note 13].

Late Bronze Age and Iron Age c.1100BC-43AD

In this long period there appears to have been a climatic deterioration, leading to marsh in the valleys, and the spread of heather, acidity, wetness and peat growth in the uplands (Caseldine 1990 67; Moore Colyer 1998; Bowen and Gresham 1967,127). However, pollen studies indicate that this was also the main period of woodland clearance, suggestive of increased usage and population (Caseldine et al 2001 31). The panoply of ritual monuments described above (including cairns) cease to be constructed, and the only ritual activity we can find is the offerings of metalwork, usually in marshes or peat bogs.

The evidence for economy remains sparse, but at last we can clearly see surviving settlement remains: Hillforts and defended enclosures containing round huts and houses begin to appear from c.850 BC, and its likely that some of the many undated stone-built settlements and round houses found in the uplands are from this period. In an apparent inversion of earlier traditions, as ritual evidence becomes invisible, settlement sites became more tangible and permanent.

Hillforts are the most well-known and visible of the prehistoric settlement sites encountered in North Wales, and have been associated with the economic pressure caused by a 'retreat' from uplands (Moore Colyer 1998, 19). However, we should remember that they are part of a trend apparent throughout much of Atlantic Europe, frequently linked not only with resource or population pressure, but the growth of warrior elites, status competition and cattle raiding between communities.

Our hillforts can be roughly divided into large and small. The large ones (which may be later in date), presumably relate to larger local polities, perhaps under the control of powerful local chieftains (e.g. Craig Rhiwarth above Llandynog, Pen-y-gaer, and the group between Penmaenmawr and Conwy). The smaller ones, usually defended by a single good drystone wall on well-chosen naturally defensible sites, have been interpreted as defended farms belonging to single family groups and dependants (Bowen and Gresham 1967, 133; Hogg 1972). Although hillforts and defended enclosures are distributed across the settled part of the landscape, it is generally assumed that they formed only one element in the settlement hierarchy. This issue will be examined below.

Their topographic location is often very illuminating. Again and again we can see that they are located on prominent high ground 'controlling' passes and routes, both on land, river and sea, providing vantage points for defence and territorial authority, and as displays of power and status to varying degrees. More mundanely but perhaps as importantly they could also enable monitoring of stock over a wide area.

Superficially, the introduction of these defensive and practical settlements seems to signal a profound change, but in a sense they can be seen as the descendants of the earlier communal monuments. Indeed, many are within territories rich in symbolic monuments, which in many cases were probably constructed by the ancestors of the fort-builders. Like the various Early Bronze Age sanctuaries, their association with communication routes signals an integration with a wider societal network with extensive trade, display and travel. Across Britain there is a traditional connection between hillforts and fairs, which were controlled or 'neutral' meeting places at certain times. Their ceremonial role is also increasingly recognised with the discovery of shrines, and of special deposits in pits incorporating human or animal remains (perhaps associated with food production rites) (Bradley 2000 152-3). Furthermore, some sites may have had an inherited ritual association related to much earlier usages e.g. burial monuments, metalwork finds (for example the important Bronze Age hoard from Moel Arthur in the Clwyddians, the single find from Dinas Bran, and the tumuli on Pen-y-gaer in the Carneddau). Some hillforts may also have had specialised/exclusive roles such as Bryn y Castell which Peter crew has identified as an important iron production centre.

As well as the hillforts, a variety of other settlement sites were occupied during this period. Many small settlements with 1- 4 huts and attached simple enclosures have been identified on the upland fringe (although some may prove to be Early Bronze Age) (see Griffiths 1950; Manley 1991). These perhaps represent usage by a single family group. Another type consists of 12-19 huts sites with larger and multiple pounds and enclosures (e.g. Cwm Caseg, Pant y Griafolen). The huts are isolated or in small clusters, and there are separate rather than communal stock pounds. This suggests they were not based on communally integrated groups but were perhaps occupied by several distinct family groups. The presence of some small plots has led to the suggestion that limited cultivation was practised.

Other sites of note include Crawcwellt in the Rhinogau, where occupation of Bronze Age date was found, followed by Late Iron Age iron production (Crew 1998, Crew and Musson 1996). On the fringes of the Denbigh moors at about 400m, remains of a hut inside a small enclosure has been dated to between 1200-800 BC. This little complex has been interpreted as a shepherd's or cowherd's hafod (Manley 1990).

What was the relationship between the various different types of settlement, and were some or all occupied only seasonally? The evidence is currently inadequate to provide clear answers. Certainly not all sites would have been in existence contemporaneously, and many will have changed their character during this long period. Many hillforts were in a position to directly exploit both upland and lowland environments and so could have provided a year round base. Some of the larger hillforts may have had both upland and lowland out-stations used at the appropriate season by a proportion of the population. Evidence from elsewhere in Britain suggests that the elite level of society had separate estates away from the hillforts on the lowland [note 14].

Whatever the role of the hillforts and defended sites, for much of this period many families and communities distributed across the uplands probably had little connection with them, living in undefended sites, some of which were permanent and others (especially the isolated hut circles and settlements in the high cwms) occupied only seasonally [note 15].

Roman - omitted

Early Medieval - omitted

'High' Medieval (post-Edwardian conquest) 1283 to 16th century

In general terms this period witnessed the transition from tribal society to a feudal social structure in which the king, lords and ecclesiastical institutions held land divided amongst tenants, and then the development of a more market-based and capitalised economy. Population expansion led to more intensive use of the the lowlands with further woodland clearance and drainage as well as some encroachment of permanent settlement onto the uplands although there was a major blip resulting from climatic deterioration in the 14th century (Edwards 1997, 7). Cattle remained the main stock animal but goats and sheep were also common. The 14th century saw the development of cattle production for consumption in England, resulting in large-scale droving along specific routeways (usually west to east) (Williams 1981 51). The expansion in the cloth trade also stimulated the production of wool for the market rather than for subsistence purposes. Despite these important social and economic changes, transhumance still seems to have been the preferred method of stock husbandry for both cattle and sheep.

There is considerable debate about the classification and definition of Medieval settlement and building types in the region. However, for present purposes I will simplify things into two types of occupation site we might expect to find in the uplands - the long hut (frequently associated with permanent settlement) and the hafod (seasonal dwelling).

The long hut was the main house-type of the subsistence farmer in the 13th and 14th century in date, many being abandoned in the later Medieval period (e.g. after the Black Death). On the lower ground, settlement is likely to have been in scattered hamlets, where much of the visible evidence for which has have been destroyed by later buildings or cultivation. However, presumably in response to population pressure, many were built on the upland fringe where their remains can be seen singly, in small groups or occasionally larger settlements. In some cases the fields were not attached to each dwelling but were part of a complex pattern of dispersed holdings, a typical farm having lowland arable in small plots of 1-2 acres totalling 10-15 acres plus upland grazing. Upland examples are usually stone-founded, are long and narrow, and frequently have internal dividing walls (some being true longhouses accommodating both people and stock). Those constructed on level platforms cut at 90 degrees to the slope are known as platform houses and are thought to be distinctively Medieval. They may be protected on the upper side by a U-shaped bank and often have a small irregular enclosure around the front end [note 16].

Dating and categorisation of long huts is not a straightforward matter. Simple rectangular structures continued to be built into the 18th and 19th century. Even those originating in the Medieval period may have been reused or adapted as hafodai or for storage in the following centuries. Not all were permanent dwellings; some may have been cowhouses, milking sheds or stores, and higher examples could well have been hafodai (see for example Ward 1997 101).

Mountain hafodai remains of this date tend to be small, rectangular and close to streams (sometimes hugging the edge of the watercourse). Some structures would be for milking or storage. Surviving examples tend to be on the open mountain, rather than on existing upland farms with hafod place-names (excellent sites are to be found in Cwm Brynog, Cwm Anafon, and Cwm Pen-llafar). As with long huts we should be cautious with regard to dating as similar structures could be earlier, later or modified and reused. Also they are often in favoured occupation areas so can sometimes be difficult to find amongst later structures, shelters and sheepfolds.

As we have seen, major early land boundaries in the uplands of North Wales are rare, and those which exist are thought to date from this period. They survive as stony or earth banks (presumably originally topped with a hedge or fence) and can be divided into two types: Firstly, ridge-top banks apparently associated with large lordly, royal or monastic cattle ranches. These have been identified at Moel Eilio and Bwlch Ehediad (see Snowdon and Perfect Line chapters). Secondly, there are 'head-dykes' which contour lower hill-slopes, separating the cultivated/improved grazing land from the rough upland grazing. These are less rare than the estate boundaries but are still not commonly encountered. Sometimes they may form part of a developing system of land division based on improvement and encroachment, representing a forerunner of the more ubiquitous and systematic stone-walled upland boundaries of later centuries [note 17].

Note 1: Jacobi (1980 191-8) identifies two kinds of site; the main occupation sites in the lowlands, and summer hunting camps in the high uplands. The latter are associated with the hunting of male red deer (the culling of which doesn't affect the breeding population). As with all upland stock regimes the limiting factor is the availability of winter feeding grounds in the lowlands.]

Note 2: Looking slightly further afield can help illuminate upland usage in this period. Excavations at Waun Fignen Felen in the Black Mountain of South Wales at a height of c.485m uncovered a site used during the early and later Mesolithic and also probably later in prehistory (Barton et al 1995). Located in marshland at the edge of a lake attractive to water fowl and game, it was a 'persistent place'. Flint and stone scatters suggest it was repeatedly used for small hunting parties albeit for periods of perhaps only a few minutes. The excavators paralleled this site with a lakeside site in North Germany where a birch bark mat was found - possibly representing a hunting stand or bird hide. Using ethnographic examples they go on to emphasise the importance of topographic features in defining hunting territories, and the role birch bark or elk hide canoes in river travel. Another relevant example is a hunting camp from Scotland dating 7200-6700 BC found at over 2,000ft on Ben Lawyers which enjoys wide views of the area, located on a route between Glen Tay and Glen Lyon which is today still used seasonally by herds of deer. As with many such early sites its discovery occurred during the excavation of a later site, in this case a 16th century shieling (British Archaeology 2001).

Note 3: Jacobi 1980 16-17 notes that the winter ranges of the deer herds were also the agriculturally attractive areas, and if these were destroyed or taken over the herds will have gone.]

Note 4: We simply don't yet know the dating, history and usage of most stone-built prehistoric upland settlements and in the absence of scientific dating a case can be made for almost any date: It has been suggested some may even be Neolithic (Crew and Musson 1996). On Dartmoor and other upland areas of Britain many similar sites are usually attributed to the Early Bronze Age, but in our region some are likely to be Later Bronze Age, Iron Age or Romano-British. A complicating factor is that upland settlement tends to occur in a limited spectrum of naturally favoured locations, and may contain elements and reused features from many centuries. Nor should we forget that the visible sites may hide earlier or contemporary timber remains.

Note 5: An insight into the sophistication and practicality of apparel and personal equipment of the period which would have been available to the local people of this time can be obtained from the Late Neolithic glacier corpse who died (apparently after being shot in the back by an arrow) at the top of the pass across the main ridge of the Alps, between his ?native settlement and the summer pastures. He possessed well made and warm clothing and footwear, including a long grass 'cloak', a back pannier (or snow-shoes), fire making kit in his waist pouch, copper axe, longbow and arrows, food supply (dried meat), small flint and bone tools, and spare parts and repair equipment. (Spindler 1994, Barfield 1994). His kit is characterised by the use of a wide range of wood, materials and animal products.

Note 6: The term as developed by Frances Lynch is normally reserved for major ritual complexes; however, I find it useful to apply it more loosely, broadly and at times conjecturally.

Note 7: One theory is that in some cases stone monuments represent the final 'hardening' of a site (previously of wood) and its transfer to the realm of the ancestors (Parker Pearson and Ramilisonina 2000, 149; and see McMann 2000.]

Note 8: Examples of broadly comparable upland sanctuaries from elsewhere include the Kilmartin Valley in Argyll, where monuments include a Neolithic cursus, a cairn cemetery, standing stones and a stone circle. Of special interest in this case is the discovery of a timber circle and other posts (possibly for totems) surrounding a now-infilled pool (British Archaeology 2002).

(8 cont.) On the plain below the hilltop Neolithic passage grave cemeteries at Loughcrew in Ireland there is a grouping of Late Neolithic and Early Bronze Age ritual monuments. These consist of a stone circle, standing stones and cairns which are thought to extend the earlier boundaries of this great site. Mc Mann (2000 347) writes that 'The structures and traces we see today may be the remnants of what was once an expanded, interconnected network of features related to myth and the cosmos as well as geography.'

(8 cont.) One of the most detailed recent studies of an upland sanctuary complex is at a well preserved Bronze Age settlement area in the Araglin and Monavar valleys in the Monavullagh mountains in Co. Waterford. Three sanctuary sites were identified, the main one covering a large area and consisting of cairns and standing stones, developed over a period of perhaps 500 years in the 2nd Millennium BC. A 'ritual geometry' in the siting of the various monuments was proposed, linked to themes such as the calendar and the sun. The use of 'visual fields' was employed to identify a very careful design connecting groups of monuments into conceptual structures. It was suggested that the sanctuary might have developed a wider political significance in the region, with people travelling to the site perhaps for ceremonies processing around the complex or to attend funerals. The similarity with the Brenig valley sanctuary is noted (e.g. proximity of a pass, developing pattern of monuments, intervisibilty of monuments) (Moore 1995, Lynch 1993).

(8 cont.) Unfortunately I have had little success in applying the idea of visual fields to the sites in North Wales: Whilst monument locations were obviously purposefully and carefully chosen, the choices appear to have been based on a repertoire of largely topographically-determined preferred locations as set out above (such as beside routes, cross-roads, territorial boundaries, prominent positions or special locations), and if placed on the site of an earlier dwelling, the choice may have been pre-determined. No doubt this field will benefit from further research.

(8 cont.) Ethnography and folklore relating to prehistoric ritual sites is unlikely to have a direct connection with their early usage, but I found the following description of a practice associated with a standing stone eerily plausible: 'at Holne, on Dartmoor, the young men , before daybreak on May Day, assemble and seize a ram lamb on the moor. This they fasten to a certain granite pillar, kill it, and roast it whole. At mid-day they scramble to get a slice of it, to secure good luck for the ensuing year. The day ends with dancing, wrestling, and so forth.' (Vaux, J. E. 1902 'Sacrifice of Sheep and Lambs' in Gossett 1911 310-11)]

Note 9: Evidence for a high status for the early ridge cairns is limited. and the only gold object from a cairn in the region was from Carnedd-y-ddeliw at 2250ft (RCHMW 1964 lxi). A few examples are very large which may reflect the social importance of the individual commemorated by the monument (e.g. Carnedd Dafydd)

Although it has largely been lost, there is a long British tradition of celebrating special times and events with beacons, and anyone who witnessed the millennium celebrations at an elevated rural location will understand the cohesive power that several distant fires in the night can have (especially if enhanced by alcoholic beverages). I therefore hypothesise that a network of bonfires based on founder cairns may have formed a feature of inter-communal festivals. I'm not aware of any archaeological evidence for this but it has been demonstrated how traces of funeral pyres leave almost no trace unless buried (McKinley 1997 esp. 134, and see Ross 2001, 30 for bonfire traditions of Wales).

Note 10: A less exotic explanation is that if they relate to earlier house-sites then the practical consideration of getting out of the wind may be relevant. Some significant passes are not apparently marked (e.g. Tal y Fan and possibly Bwlch Cynwyd, Cwm y Llan). Although I have not undertaken the research needed, I suspect pass-marker cairns are a feature in other upland areas. For instance, on the ground it is striking how a cairn and adjacent standing stone are sited just below the pass beside the modern hill road between the Neolithic passage graves at Carnbane East and West at Loughcrew (see McMann 2000 336 fig. 2)].

Note 11: examples include Nant y Bache (above Llangollen), Eglwyseg moors, Cwm Rhayader (where there is a cairn above the waterfall and a cairn and sanctuary higher up at Rhos y Beddau). Nearby access to water for stock would be important in such a context: Many are near a source, those which are not may have different origins.

Note 12: Comparisons with recent practice in Madagascar illuminates the potential diversity of functions which such features can take on. The island has a centuries-old and continuing tradition of erecting standing stones for a variety of reasons in different areas - they may be cenotaphs, boundary markers between group territories, commemorations of important events, lines of stones between palaces, and symbols of the ancestors. A common feature noted was that their erection was associated with social unity and togetherness (Parker Pearson and Ramilisonina 2000 147-9).]

Note 13: I'm not aware of any recent work on these features and I have only been able to locate one of them on the ground. They are detailed in RCHMW 1956, 129; Bezant Lowe1912, 166-8 and Arch. Camb. 1864, 315.

Note 14: Alcock suggests transhumance in the territory of the Deceangli; the three large hillforts in the Clwydian range in the middle of their territory used for summer pasture, with their main settlements on the river plains of the Alwyn and Dee (Alcock 1965 188-9)].

Note 15: It is widely accepted that many of the Medieval/Post Medieval rectangular buildings in the uplands are summer dwellings so I see no reason why even the more extensive prehistoric sites in broadly similar locations are not of a similar nature. For example the 16th century hafodai at Nant y Criafolen in the Brenig valley consisted of a group of seven buildings with their own associated enclosures (Allen 1979).

Note 16: See Gresham 1954 on platform houses. See also RCHMW 1964 clxxviii-ix. There is an extended row of them on the south slope of Clip y r Orsedd (709747), and larger groupings on the ridge dividing Cwm Pennant and Cwm Ystraddlyn.].

Note 17: Possible examples include stony banks below Bwlch Maesgwym (area of 574555, see Llanberis chapter), on the moors above Llanfairfechan is a distinct turf-covered bank around 707737), and south east of Llidiart y Parc in the North Berwyn at c.133418. They might be paralleled with the Medieval head-dykes of the Lake District, described by Winchester as '...the most fundamental feature in the upland landscape.' and were the 'hinge' of the transhumance system, keeping stock off the lower ground in summer allowing it to recover and to produce hay and other crops (Winchester 2000 53). As in North Wales, the land below the intake was individually owned or tenanted and that above was held in common. (Winchester 2000, 29-30. 52-4).

See here for contents page, or here for the bibliography.

Note: This is part of an original series of articles written by archaeologist M M Robinson.

Contributor Bio for M_M_Robinson: Based in North of England, contributing since 2006. Occupation: Archaeologist.

<< Abandoned Landscapes - The Deep Structures of the Archaeological Landscape

Feature Articles

Abandoned Landscapes - Chronological Change: 16th Century to The Present >>