Abandoned Landscapes - The Deep Structures of the Archaeological Landscape
Submitted by M_M_Robinson on Monday, 03 July 2006 Page Views: 7135
Multi-periodCountry: Wales
CHAPTER 1
THE ARCHAEOLOGY OF THE HILLS: A SELECTIVE OVERVIEW
THE DEEP STRUCTURES OF THE ARCHAEOLOGICAL LANDSCAPE
Before beginning our exploration of these uplands some general observations on the archaeological landscapes and remains encountered are necessary. These are divided into two parts:
Firstly, I have sought to model some of the 'deep structures' - the environmental, geographical, economic and social constraints which have created repeated patterns in land-use and settlement. The second theme is that of chronological change, in which the broad economic, social and cultural transformations from period to period are outlined in relation to the surviving archaeological evidence, and more flesh is put on the bones of the 'deep structures' summarised in the first section. Readers will note the emphasis on the Bronze Age and the Post-Medieval/Modern periods which are both archaeologically rich and also happen to be my main areas of interest [note 1].
The deep structures of the archaeological landscape (convergences)
Pastoralism and Cultivation
Due to high rainfall, elevation and often poor soils, the defining characteristic of farming in the uplands of North Wales has been livestock husbandry (mainly cattle and sheep) based on systems of seasonal movement. There are several variants which may have been utilised, but each was a complex and sustainable response, maximising the carrying capacity of an often harsh environment (see Ward 1997 103-4). Key features of upland stock husbandry are based on the natural social behaviour of the animals, modified by the guidance and influence of humans. It has been observed in relation to shepherding practices that '..customs that may be thought to have only cultural significance probably have a much deeper biological origin.' (Ryder 1983 4); undomesticated sheep for example have home ranges within which they move on a daily and seasonal cycle. Thus there are common elements of practice from earliest domestication to the present day (see Pryor 1996).
Nomadism
Nomadic pastoralism may be applicable (at least in part) to the late Mesolithic, Neolithic and Early-Middle Bronze Age - a period of several thousand years. Within this regime, each social group would have had defined routes and stopping places. Small-scale cultivation may have also taken place at some of the temporary sites (Carrier 1932; Ryder 1983 648-). The apparent rarity of settlements and houses from this early period may be explained by their temporary nature. Ritual sites (e.g. tombs, cairns, circles, standing stones), and evidence for 'trackways' may be associated with the cycles of movement between designated grazing grounds along with their temporary settlements [note 1a].
Transhumance
The most specialised form of nomadism is transhumance, in which there is a permanent winter base. The variant suited to North Wales is known as 'vertical' transhumance, where the seasonally-used upland tillage and pasture grounds lie within a home territory usually only a few hours journey away from the permanent (winter) farm. It is recorded in the region from the 10th century AD but was arguably practiced from the Late Bronze Age (or even earlier), and can perhaps be associated with the appearance of stone-built upland settlements and hillforts (Allen 1979, 48 and see Carrier 1932) [note 1b].
Both nomadism and vertical transhumance were primarily associated with subsistence production of milk and cheese rather than meat and wool production for the market. The good quality upland grazing required for cattle was available at only a limited number of favoured areas, often to be found in the central portions of the major upland cwms where there was relatively flat, sunny and well-drained land. Structures of various periods frequently clustered around and along such locations, becoming 'persistent places' (see Williams 1981 60). As well as cattle, sheep and goats would have been important to each community as they could utilise poorer and steeper ground, but they may have required a separate system of husbandry to cater for their differing needs [note 2].
There is relatively little detailed evidence for the practice of transhumance in North Wales. However, ethnographic evidence from this area and elsewhere in Britain and Europe evokes something of the lost human stories played out in the occupation and house sites amidst these hills. The summer migration from the winter farm (hendre) to the summer dwellings (hafodai) would take place after the spring crops were sown, with the return around haymaking time so that stock could graze on the stubble. The traditional dates were 1st May and All Saints Day (Davies 1977 50). Members of the community who participated varied from place to place; sometimes everybody was involved or perhaps just the young men, women and the elderly. Families could migrate separately or as part of a communal event. The social and symbolic importance of the migration is commonly emphasised, with descriptions of rituals of leaving and arriving accompanied by music, singing, noise and general excitement. Feasting could also take place on arrival. The accommodation in the uplands would be in a hut, smaller and more basic than the home farm. Repairs, especially to the roof, would often need to be carried out on arrival. The following description of an early 20th century alpine summer hut is perhaps applicable to many times and places:
'A doorless opening gives access to a single room of which the hearth and suspended cauldron occupies one corner, while another is furnished with a little straw to be used as sleeping accommodation... Rough stools or benches of wood or a few simple logs serve as seats. Some shelves for cheeses complete the outfit. In wet weather the unplanked floor becomes a pool of liquid mud...' (Carrier 1932 216-7)
The distance of migration could also vary; if it was relatively short, frequent returns could be made to the home base. In North Wales the journeys must have usually been of this sort as no high valley pasture grounds are more than a couple of hours walk from the lowland. The summer period could be spent at a single hafod, or include moving from one hafod to another as the grazing was exhausted. If connected to separate holdings and tenancies, hafodai tended to be isolated from one another but if communally organised, the huts would be clustered. In either case the site needed to be on dry ground with nearby running water which was essential for watering the stock, for butter and cheese-making and for washing the implements. (Sayce 1958, 44-5).
Daily life followed a closely-prescribed routine. The cattle would be milked 2 or 3 times per day, probably at early morning and evening. Butter and cheese-making (some kept for winter food) would occupy much time. Cattle commonly spent the night in folds - for protection from raiding and predators, and to facilitate counting or inspection. It would appear that temporary or ephemeral shelters (probably hurdling) were the norm in both nomadism and transhumance. Most of the stone folds and enclosures we find in the hills date from the last two or three centuries, and the majority of these relate to 'static' regimes of shepherding (see below). There could be a variety of different folds and enclosures - pairs or double pairs to keep calves from cows and kids from goats (Sayce 1958, 43), shelters for bad weather or for sick animals (Carrier 1932 218-). The stock might be self-reliant, going out to pasture themselves in the morning and returning at night, or they may have needed herding. Pasturage boundaries and territories were usually well defined (e.g. by township in the Medieval period), and would be regulated by the herders. There would be no need for fences or walls as limits would be well known from landmarks and other markers. Cattle would usually not wander beyond the habitual areas anyway (Sayce 1958, 47-8-).
The summer migration was an important feature of the lives of the individuals involved. Pure pleasure in being in a beautiful and healthy upland environment, the greater social autonomy and distance from authority, and the relative freedom for women are all recorded responses to the experience (Sayce 1958, 47). Another socially significant aspect was the frequent (and sought after) contact between communities normally living a good distance apart. For example the high Berwyn was shared by people from the Dee valley to north and the Tanat valley to south, and they would meet, socialise and exchange news at the hafodai (Sayce 1958, 74-5).
Pennant's well known late-18th century description of transhumance in and around the Snowdon massif encapsulate elements of the economic system and lifestyle which would probably be broadly recognisable to the people of the uplands back into Medieval times and in some respects well into prehistory [note3]. As a rare eyewitness account it repays quotation in full:
'This mountainous tract scarcely yields any corn. Its produce is cattle and sheep, which, during summer, keep very high in the mountains, followed by their owners, with their families, who reside in that season in Hafodtai, or summer dairy-houses, as the farmers in the Swiss alps do in their Sennes. These houses consist of a long low room, with a hole at one end, to let out the smoke from the fire, which is made beneath. Their furniture is very simple: stones are the substitutes of stools; and the beds are of hay, ranged along the sides. They manufacture their own cloaths; and dye their cloths with Cenn du y Cerrig, or Lichen omphaloides; and another Cenn, the Lichen parietinus; native dyes, collected from the rocks. During summer, the men pass their time either in harvest work, or in tending their herds: the women in milking, or making butter and cheese. For their own use, they milk both ewes and goats, and make cheese of the milk, for their own consumption. The diet of these mountaineers is very plain, consisting of butter, cheese, and oat-bread, or Bara Ceirch: they drink whey: not but they have a reserve of a few bottles of very strong beer, by way of cordial, in illness. They are people of good understanding, wary and circumspect; usually tall, thin and of strong constitutions, from their way of living. Towards winter, they descend to their Hen Dref, or old dwelling, where they lead, during that season, a vacant life.'(Pennant 1883 325-6) [note 4]
Cultivation in the Uplands
Although crop growing is now unusual in the area, in pre-modern times small-scale production as part of a mixed subsistence economy was the norm. Along the coastal lowlands, in well-drained tracts of the main valleys and on the lower slopes of the hills where we now see only sheep and cattle grazing, small arable plots near settlements and farms would have been a common sight at all periods probably from the Neolithic onwards [note 5].
Crops (oats) were also grown on the upland fringes up to perhaps 350m and higher in some favoured locations. Where shelter and aspect created a conducive micro-climate, we can find signs of long-term (and possibly repeated) encroachment of mixed farming onto the moorland. Prehistoric examples include the southern slopes of Tal y Fan, the eastern side of Cwm Pennant (Hebog), above the Nantlle valley, Mynydd Egryn in the Rhinogau, and around Moel Faban (above Bethesda). There are also many Post Medieval farmsteads in this marginal zone.
Such regimes may sometimes have been in response to environmental improvement (as has been suggested for the Bronze Age, when intensified activity perhaps associated with settlement occurred in areas such as the north Carneddau moors). However, assarting in the mountains is more likely to have been spurred by population pressure or economic stress; some of the 18th and 19th century homesteads were established in the most unprepossessing, even desperate, locations and were not sustainable in the long-term.
Establishing and maintaining arable regimes in the mountains must always have been a struggle. Firstly, the surface had to be prepared by 'denshiring', where the organic top surface was pared off, burnt and then spread around (Williams 1981 55-6). The labour involved was regarded as the toughest of all land work. Evans (1948 103-4) describes this process on the 'rough hill land' of the Denbigh Moors in the 19th century:
'The mountains of Cwm Eithin were low and grass-covered and the small farmers used the land almost rent-free at the time... But the ploughing of such land was the problem; there was no plough powerful enough to penetrate the tough turf ... There was nothing for it but to strip the turf first of all... The implement used was the push or breast plough... When the pieces of turf were dry they were burnt. In early winter the ground thus prepared would be ploughed in the usual way; oats would be sown in the following spring, and a good crop usually matured by harvest time.' [note 6]
The result could be a period of good production; he continues: 'I have seen oats higher than my head on some parts of the hill land, a better crop than could be grown on many places on the lowland.'. But there was a price to pay:
'This would go on for a few seasons, until the soil was exhausted. The good crops, however, lasted long enough to enable the landowners to raise the rents, with the result that some of those who had laboured to cultivate the mountain land were in the end worse off than those who had been content to use the hill as pasture for sheep according to the usage from time immemorial.'
Manuring of the fields was practiced in an attempt to extend and enhance the fertility of the soil as was carried out in the Llanuwchllyn district of Merioneth, where:
'... all dung was carried to the upland fields in hand barrows no more than a hundred years ago. Each farm in that mountain-locked locality specified a day for manuring its fields, and on that day neighbouring farmers and their families met at the specified farm with a large number of [two-man] barrows. The carrying was performed in a relay service...' (Geraint Jenkins 1976 55).
An alternative to intensive manuring was to leave the land to recover for a period of perhaps 20 years after which it could be brought back into productive use again. This would encourage a regime of temporary/shifting cultivation, as is recorded from the 14th century in the poor upland areas (Williams 1981). Both manuring and shifting cultivation must have been employed in prehistoric crop production in the uplands .
Settlement Territories and the Landscape
The structure and appearance of the landscape:
During a journey from lowland to upland, the following environmentally-determined land-use zones would typically be encountered, perhaps recognisable with increasing frequency from the time of the earliest forest clearances in the mid Mesolithic onwards:
Lower floodplain meadow suitable for summer grazing and haymaking.
Enclosed pasture up to 250-300m suitable for year round pasture, previously also used for small-scale arable and hay production. This zone is often divided into small rectilinear or irregular drystone-walled fields (?17th-19th century) and most settlement is located here. The winter farm (hendre) lies within this zone below 200m.
Ffridd or enclosed mountain pasture in a band between c.250/300m to c.400/500m. Usually heavily stocked in summer and previously also providing a hay crop. The existing boundaries are usually large or very large rectilinear enclosures of 18th/19th century date, with the upper boundary sometimes marked by a contouring intake wall. This is the zone where most surface archaeological sites of all periods can be seen, often towards the upper boundary. Hafod place-names tend to occur at c.300m.
High mountain pasture from 400/500m where sheep now graze in summer. These are areas where a deree of intercommoning may have taken place in the past. As noted above, the lower boundary will be a long intake wall. In addition the high ridges and watersheds may nowadays be walled or fenced, with cross-boundaries dropping directly downslope. Many settlement sites and sheepfolds survive in this zone although relatively few are found above 500m. [note 7]
Applying these notional zones to the real landscape, we can identify two broad types of upland terrain. Firstly the hillslopes facing a coastal plain, estuary or a major river valleys e.g. the western Rhinogau, the north slopes of Cadair Idris. These can contain very complex and rough ground, and be intersected by small valleys and streams, although the naturally occuring terraces or shelves on the slopes provided attractive settlement locations (e.g. the Nantlle valley, Myndd Egryn). Such areas tend to display a lack of clear natural boundaries.
Secondly there are the large 'closed' upland cwms which feed onto the coast, estuaries or major valleys. Here we might find relatively fertile and well-drained land below or beside the wide valley mouth, usually accommodating a modern settlement (perhaps Medieval in origin), and maybe overlooked by a hillfort on the adjacent high ground. Further up, the valley floor is often flat and poorly drained, leaving only the sloping sides for usable pasture. As the cwm narrows and the sides steepen we eventually reach the headwall which frequently provides easy access to the high mountain pastures. In the case of some of the large glaciated valleys of the higher massifs (e.g. Carneddau, Snowdon) they eventually narrow into intimidating rocky cwms with small glacial lakes, although even here houses and shelters may be found. There are also several instances where the valleys are two-tiered, with an extensive 'hidden' valley above the steep headwall of the lower valley. These latter places are particularly evocative to enter; some have Late Neolithic/Early Bronze Age monuments and seem to have been significant boundaries or connection points (e.g. above Llanrhaeadr in the Berwyn or Eglwyseg above Llangollen, both described in later chapters). Boundaries in the valley terrains are usually clearly defined by the natural topography, which also heavily proscribes stock movement.
Local Territories
How might these different terrains have been divided and utilised by families and communities in the past, and are there elements of 'continuity' and 'tradition' across periods [note 8]?
One of the most striking characteristics of this region is the sparsity of man-made boundaries or formal land divisions until the enclosures of the last couple of hundred years. I’m not aware of any prehistoric boundaries or major land divisions, and only a couple of Medieval 'ranch' boundaries (both of which are apparently associated with either royal or religious estates). We must infer that even if some come to light during future survey, the general situation was that natural features were utilised to form a framework recognised within and between communities (e.g. the sea, watercourses, mountains and watersheds).
Some insight into these ‘traditional’ but now intangible divisions can be gleaned from the archaeology. Looking first at localities, we might assume that the basic settlement unit was a single homogenous group or extended family of 15-20 people. From the Neolithic into the Bronze Age, when pastoralism or transhumance may have been dominant, territories may have been conceived primarily in terms of seasonal route-corridors rather than closely-defined property. The Neolithic tombs may signify a communal focus within large ancestral territories, and the accelerated marking of the landscape in the Bronze Age (especially pass-marker and ridge-top cairns) surely indicate increasing concern with boundaries and possibly sub-division of earlier lands.
From the Late Bronze Age and Early Iron Age we are possibly on slightly firmer ground with the appearance of the small hillfort or defended enclosure, probably based on extended family groups, located at intervals on the upland fringes. These may offer us a sense of the contemporary territorial organisation and spacing (although they probably only represent one element in the contemporary settlement pattern). The first point is that their geographical distribution corresponds with areas of earlier activity, indicating that 'core' areas already identifiable in the Neolithic continued to be the focus of the human landscape. Secondly, whilst to some extent their siting on prominent, defensible places can be seen as practical, it is striking when you stand on the ramparts of these little hillforts how often they overlook the lowland and appear to control access to the upland. Perhaps they stood at the centre of family or clan territories, each incorporating a gradation of catchments (and settlement). This is most apparent in the Carneddau where the mouth of virtually every upland valley is overlooked by its own defended settlement.
The local territorial pattern in existence by the Late Bronze Age may have continued until the Post-Medieval period. Local centres of power sometimes persisted below the hillforts, as well as markets and settlements (e.g. Aber), and several modern villages, superficially of Medieval origin stand in the shadow of hillforts (see below).
A Case-study of a Local Territory: The Brenig Valley
The nationally important and comprehensive archaeological study of an upland locality carried out in the Brenig Valley during the 1970's in advance of reservoir construction provides a model for the hidden complexity of the historic landscapes encountered throughout this book (Lynch 1993). The valley lies within the large moorland area of Mynydd Hiraethog (Denbigh Moors), with the main watershed at Bwlch Du, from which the Afon Fechan flows south to join the Afon Brenig at a wide central basin. Most of the work dealt with the sites and monuments towards the head of the valley. Pollen analysis from the peat revealed a very long-term pattern of land use, indicating grazing of the uplands with episodes of more intensive use and some agriculture in the lower valley. Mesolithic pits and hearths, perhaps related to year-round occupation, were found beneath and beside a Bronze Age barrow. Neolithic occupation appeared to be sparse, with a few Late Neolithic flints, and a macehead from Hafod Lom (lower down the valley).
However, the author suggested that from the Late Neolithic and Early Bronze Age there may have been an expansion of settlement into the uplands spurred by the exhaustion of lowland soils, an improvement in climate and population increase. The first indication is a 'marker' cairn (apparently without a burial) beside the pass at Bwlch Du (dated to c.2650 cal. BC) which may have been a community/territorial boundary at the junction of two ecologically contrasting zones. Tree cover became reduced, and sedge and grass took over, and during the Early Bronze Age a cemetery developed, focused on the then centuries-old marker cairn. The excavator concluded that this 'sanctuary' was conceived in outline as a single entity from the beginning rather than as a growing thing, although it was in operation for c.500 years and had three main phases of use. Features included a ceremonial 'ring' cairn, various barrows and cairns and a small kerb cairn in an adjacent valley (dated to 550 cal BC.), beneath which was a circular timber structure. In all only 23 individuals were found covering the whole period (representing a continual population of only one person over 500 years: Obviously most of the dead were being disposed of in ways not archaeologically detectable e.g. exposed or in rivers). Occupation soils and beaker pottery were found underneath a platform cairn.
The Iron Age was represented by a possible hut circle (interpreted as a seasonal dwelling) associated with pottery of the second century BC, found during the excavation of a Post-Medieval hafod. Further across the valley is the large and well-known Medieval enclosure of Hen Ddinbych. A group of Post-Medieval hafodai were excavated, providing the type-site for the region (see below). The 19th and earlier 20th centuries perhaps witnessed the most intensive period of use, with several smallholdings including Hafoty Sion Llwyd. Husbandry was mainly sheep-based but small fields around the smallholdings and some evidence for ridge and furrow indicates cultivation [note 9].
Favoured Locations and Special Places
As was discovered at Brenig, within the upland zone of each locality we find that the same kind of location or place is preferred for a variety of activities across periods, usually determined by geographical and practical considerations. Significant factors could be aspect, drainage and water supply. The reuse of existing sites may also have been prompted by the ready supply of building materials or because it had already been cleared and levelled. To take the case of the high valleys of the Carneddau, favoured locations here accommodate a cluster or string of sites - prehistoric houses and enclosures, possible Medieval hafodai, Post-Medieval stock pens and more recent complex sheepfolds. After some practice it becomes possible to look at a valley or hillside and predict the vicinity in which sites will be grouped.
Above-ground sites at favoured locations only represent the tip of the iceberg. Earlier occupation beneath or near later sites is encountered with remarkable frequency during archaeological excavations in the region (for example, the Neolithic house at Moel y Gaer hillfort, Bronze Age huts beneath Late Iron Age activity at Crawcwellt, earlier settlement beneath the Bronze Age kerb circle at Cefn Caer Euni (Lynch 1986 83-7), and several sites at Brenig (see above).
What might the buildings and settlements have looked like? It is easy to assume that there was a 'type' for each period and use, with remains of stone huts and houses indicating fully stone-built structures. However, I suggest that for seasonal buildings in particular, stone would have been supplemented by temporary timber elements on a pragmatic basis, particularly in earlier times [note 10].
As well as sites conducive to economic or settlement activity, certain natural features in the prehistoric landscape are likely to have been regarded as special or significant places where rituals, sacrifices and votive offerings took place (perhaps concerning issues such as human and animal fertility, death, and status and boundaries). This phenomenon occurred across prehistoric Europe from the Mesolithic onwards, diminishing with the growth of larger-scale and more centralised societies of the Roman and later periods in which ritual tended to occur in man-made monuments (e.g. temples, shrines and churches). Sites and practices would obviously have changed, but archaeological and ethnographic evidence from Britain, Ireland and Northern Europe show that certain kinds of place were repeatedly used over a very long time-span. Perhaps most important were watery places, primarily rivers, but also pools, lakes and bogs. The late Iron Age/Roman period bog bodies are the most popularly known evidence, but Neolithic axes, Bronze Age and Late Iron Age metalwork are commonly found in watery places. Rivers are also increasingly recognised as the main repository for the dead. Distinctive landforms such as mountains, cliffs, prominent or unusually-shaped outcrops or rocks could also be regarded as sacred (Bradley 1998; 2000; Edmonds 1995 133; Horden and Purcell 2000).
In a spectacular landscape such as North Wales, hundreds of places compatible with the above types can be found, many with multiple characteristics (such as the glacial cwms of the Carneddau). However, tangible evidence is very limited, and without some criteria for selection the quest to identify them is pretty meaningless. Based on patterns from elsewhere and some limited but direct evidence from our region, we should probably be lookingfor places in the vicinity of the core areas but in peripheral, transitional/marginal locations, especially those near natural/early routeways and perhaps passes and watersheds. Legends may also indicate significant sites, some perhaps containing folk memories of prehistoric origin. Single finds or hoards may be indicative and, although Neolithic tombs seem to avoid them, some special landmark features will have been formally 'adopted' by cairns or hillforts [note 11].
Larger Regions
The communities of each locality or micro-terrain did not exist in limbo; each would have formed part of a wider social and cultural network. There will always have been larger groupings, alliances, transient power structures and sometimes political and ruling elites, but are there any discernible recurring connections or relationships between the small-scale units which resulted in distinctive geographical territories?
In an important and provocative paper, based to some extent on his work on the Dartmoor Reaves, Andrew Fleming hypothesised the existence of long-term landscape units or 'large terrains' of more than 100sq. km originating in the British Isles as early as Mesolithic times (1998-). A different tack was taken by Waterbolk (1995 31-3), who started from the premise that for agricultural societies, a minimum population grouping of 250-500 is needed to remain viable, an imperative which stimulated agglomerations of perhaps 12-20 individual settlements. Unless a grouping of this scale exists, he argued, whole areas may remain largely unoccupied for significant periods. These larger groupings were termed 'adaptive areas' or 'nuclear regions'.
Both these approaches resonate with the evidence from North Wales. Taking Fleming's 'large terrain' concept, we might expect the Mesolithic territories to be 'laddered' regions, with the home base on the coast, beside a main estuary, allowing movement up the river valley into the interior and uplands on a seasonal basis. Hunting territories would presumably be defined by the mountain ridges and watersheds. The Vale of Clwyd, from the excavated site near the coast at Rhuddlan into the Denbigh Moors might form a regional archetype, whilst others might be postulated based on the estuaries and major river valleys (e.g. Conwy Valley, Ogwen Valley, Traeth Bach, Mawddach, Dysynni).
If we then look at the settlement pattern from the Neolithic onwards we find four distinctive zones, rich in archaeological remains, which I have termed 'core' territories, each of which could be a shadow of an earlier 'large terrain' (see for example Bowen and Gresham 1967, 24; Lynch 1974; Powell 1973). They have good access to the coast or major valleys and upland grazing, and are approaching the scale suggested by Fleming. They are:
1. The Conwy valley, with a grouping of Neolithic tombs, Bronze Age monuments (on the moors above) and hillforts.
2. An inland territory along the Dee Valley between Cynwyd and Llandrillo which has three tombs as well as later monuments (and links over the Berwyn hills to the Tanat Valley).
3. The coastal tract between Traeth Bach and Afon Mawddach including the western slopes of the Rhinogau, which has a similar variety of terrain and monuments.
4. Finally there is the land between Afon Mawddach and Afon Dysynni (taking in the northern slopes of the Cadair Idris massif), abounding in prehistoric and later settlement, although the Neolithic component seems to be absent.
Turning now to Waterbolk's theory, whilst it would be unwise to apply it uncritically, it is interesting that each of our distinct 'core areas' might have been able to sustain population levels and settlement numbers broadly in accord with his 'nuclear regions'. Furthermore, a conspicuous characteristic of the four zones is that they represent islands of archaeology in an otherwise largely empty sea. Perhaps the evidence is telling us that the introduction of the Neolithic ‘package’ (whether by colonisation or acculturation) was not a piecemeal and slow process but was introduced rapidly by (or adopted by existing) sustainable integrated communities within defined (possibly pre-existing) territories. In simplistic terms it may be that they were established and settled early but population never increased sufficiently to replicate new societies in the voids of the interior. Instead, only thinlyspread occupation developed outside these mother settlement areas, connected by the umbilical cords of trackways and rivers.
Continuing this line of thought, although these little societies appear geographically isolated by sea and mountain, their cultural affinities, technology and possessions demonstrate that they were integrated into major systems of communication and transport. Throughout prehistory their coastal/riverine locations were probably as important to their existence as was the subsistence productivity of their territories. Similarly, it is of profound significance that each is associated with a network of definable tracks and paths linking them to places both near and far (see Barker 2001). This aspect is considered in more detail in the introduction to Routeways, below.
See here for contents page, or here for the bibliography.
Note 1: For a more thorough, scholarly and balanced picture than is presented here, I refer the reader to Lynch et al 2000 and Arnold and Davies 2000 which are currently the standard works of synthesis on the archaeology of Wales (albeit they don't cover the last few hundred years).
In presenting this overview I've sought to incorporate recent mainstream thinking on British prehistory (up to 2002), but I'm aware that some elements are open to dispute. Firstly I've made some concessions to the phenomenological approach (although hopefully from an empirical perspective). I’m aware my view that Bronze Age hillside and valley cairns frequently lie on temporary settlement sites is contentious, as is my general emphasis on seasonal movement as an abiding and fundamental feature of upland economy and settlement, even in prehistory. This debate on seasonal versus permanent settlement has been recently reviewed (and fully referenced) by Ward (1997 104-7). In essence, workers such as Stephen Briggs question the traditional view of the dominance of transhumance, pointing to the importance and ubiquity of permanent settlement associated with arable cultivation even at high altitude.
Upland land use was undoubtedly complex and dynamic, affected by social change, population pressure, climate and environmental shifts, and I'm sure that the number of examples where permanent settlement can be indicated will continue to increase (e.g. through aerial photographic evidence for associated cultivation). However, my current view is that most 'occupation' sites of all periods above c.300-350m were primarily seasonal (albeit sometimes with episodes of permanent occupation). Writing about Medieval sites in the Northern Carneaddau, Hooke recognises that whilst many upland sites may have been permanently occupied 'some, however, lie at such heights and amidst such inclement terrain that it is difficult to believe that they can ever have been anything other than seasonal hafodydd' (1997, 89).
Note 1a: Shifting nomadic pastoralism can still be witnessed in Western Europe, for example in the Pasiego valleys of Cantabria. In this remote wet upland area the hillsides are dotted with rectangular stone buildings; each family has up to half a dozen houses and moves from one to the other with their stock and possessions as the grazing becomes depleted. The cabins are substantial rectangular stone structures, often set into the slope, with cattle byres at ground level and hay-lofts and/or living quarters above.]
Note 1b: Ward urges caution in adopting this model too simplistically. Although his comments are related to the Medieval and Post-Medieval period in the Black Mountain of South Wales, they perhaps have a broader geographical and chronological relevance:
'a) that transhumance is only one of a range of adaptations which may have been necessary for the exploitation of this area;
b) that the practice of transhumance may have involved a number of different social and economic models;
c) that usage of satellite outstations may have involved more than a repetitive seasonal cycle' (Ward 1997 108-).
Note 2: We can still witness the continued use of persistent places and also the complimentary grazing of cattle and sheep in many alpine regions. For example in the Pyrenees, near the stream of Estaube (east of the Cirque de Gavarnie ) the valley broadens out to provide lush grazing for cattle. This zone is clearly the natural locus for cattle husbandry; the cowherd's shelter (Cabane d'Estaube) is still in good repair, and nearby is a modern rectangular cattle enclosure. A ruined rectangular drystone stock shelter (similar to many found in North Wales) which sits against a small rise in the base of the valley, probably represents the predecessor of the modern enclosure. Beside the adjacent path are the remains of two circular structures which would be interpreted as prehistoric huts in the UK uplands. Sheep graze higher up the valley and on the steeper slopes. Lower down the valley, a few hundred metres south of Lac Des Gloriettes, is a zone of broken rocky ground also grazed by sheep. In this location is a disused stock enclosure built against a natural outcrop and a small ruined shepherds hut, again, similar to those found in our region.
Note 3: This seems to be a description of archaic and dying practices. At this very time major social and economic changes were taking place in the uplands, such as the decline in transhumance and production of sheep products for the market. In earlier times cattle were more important and production was more concerned with meeting a balance of subsistence requirements.]
Note 4: Horizontal or 'grand' transhumance of sometimes hundreds of miles (best known from the Mesta flocks in Spain, Balkans or Italy) is generally seen as a Medieval development and doesn't seem to have been characteristic of our region. However, in the 19th and earlier 20th century there was very significant migration and transport of sheep (mainly ewes and lambs) from the upland farms to overwinter in lowland zones 1-3 days journey away. In Snowdonia itself there was a divide along the line of the Nant Ffrancon, with sheep from the hills to the west wintering in two areas near the coast - in Eifonydd and around Clynnog, and those to the east wintering on the far side of the Conwy between Llanrwst and Capel Garmon (Davies 1935, 26-7), and it has been suggested that this practice was 'a continuation of ancient transhumant migrations of cattle herders, which existed in Wales as late as medieval times' (Davies 40). If these are genuinely early practices (i.e. prehistoric or Early Medieval) they have profound implications for the archaeological study of the region. Similar seasonal stock migration between separate lowland and upland communities (Agistment) is attested from the 15th to 17th centuries in Northern England, and there are indications that these too were the continuation of 'traditional' arrangements (Winchester 2000 93-8-).
Note 5: There is solid documentary evidence for extensive Medieval arable cultivation in the Conwy valley (Hooke 1997), and wheat, barley and rye were grown here in the 17th century (Williams 1981 73). Good examples of prehistoric cultivation fields and terraces can be found on the coastal slopes of the Rhinogau (see later chapter).
Note 6: Based on his grandfathers account books for 1815, relating to ffridd at Garreg Fawr Syrior, near Llandrillo (c.0038-), Evans estimates it would take 2-3 weeks to deturf an acre of such land. (Evans 1948 103, 104-5).
Note 7: Such divisions are not confined to our region. For example, 16th and 17th documents for Eskdale, Miterdale and Wasdale in the Lake District show a threefold division of upland land zones 'first, and most jealously guarded, were the lower fellsides, the steep (and now generally bracken-infested) banks immediately behind the farms' which offered good grazing for the milk cattle, then 'the moor' for summer pasture for other large stock including horses, and thirdly 'the high fells', reserved for sheep (Winchester 2000 110-111). These zones are broadly equivalent to 'enclosed pasture', 'ffridd' and 'high mountain pasture' . Similarly, the Bronze Age Dartmoor Reave system of land division is divided into three such zones (Fleming 1978 and 1983)].
Note 8: Comparison with studies of pre-modern European local settlement territories (called variously micro-environments, neighbourhood areas, settlement chambers, or community areas) is helpful in conceptualising the patterns in our chosen area. Neustupny and Kuna developed a model for prehistoric agricultural settlement in central Bohemia characterised by community areas 2-3 km apart, each with an oviod territory encapsulating a stream, covering an area of 4-8km sq. or larger. Each area has a number of activity areas - habitation (consisting of a very small number of households with a total population of c.15-20 individuals), storage, mortuary, cultivated fields, and pasture. These territories appear to have been very stable over long periods of time (Kuna 1991, Neustupny 1991).
Waterbolk studied the archaeological landscape of areas of Holland from the Middle Bronze Age to the 19th century, and identified persistent settlement territories. The actual settlements were c1.5km apart on good land, and on poor ground up to 5.5km distant. Within similar geographical environments there was a high degree of standardisation in the location of settlements and the division of land etc. (Waterbolk 1995 31-3, and see also Bintliff 2000 148 on the reuse of favoured sites). At the Araglin and Monavar valleys in the Monavullagh mountains in Ireland, a Bronze Age landscape with a core area of c.4km sq. at between 200-300m ASL was estimated to have a population of perhaps 15-20 adults, perhaps divided into 3 or 4 families, each with their own fields, huts and animal enclosures (Moore 1995 191, 237).]
Note 9: The long-term picture of cumulative but punctuated clearance and usage of the upland environment was also found at the important site of Llyn Morwynion, near Ffestiniog, summarised as follows: 'A series of clearance and regeneration episodes are discernible in the early to mid bronze age. Although there is some recovery of woodland during the early iron age, a further expansion of clearance continues during the later iron age and Romano-British times. By the medieval period, the open grassland and blanket and valley mire landscape of today, with grazing as the main landuse, had been created.' (Caseldine et al 2001 32).]
Note 10: The seasonally used buildings (kharkas) of the Nepal Himalaya provide a remarkable insight into the ingenuity and variety of constructional form which can occur, often in contemporaneous use within the same settlement. At one end of the spectrum are temporary earthfast constructions made of arched poles in wigwam fashion, with stone houses with 'permanent' wooden plank roofing at the other. Between these types are many examples of buildings with timber superstructures sitting on dwarf drystone walls, or even with single courses of flat stones at ground level. Some might have a stone gable or side wall with the remainder of timber. All buildings have complex internal timber divisions (e.g. stalls for baby yaks, human sleeping areas and storage compartments). Roofing can be wood shingle, traditional bamboo matting or modern material such as polythene sheeting. Lean-to storage areas and dog kennels could be attached outside, with tethering stakes for yaks, piles of manure and trampled areas nearby (evidence for tethering of animals has also been found around the Early Medieval hafodai on Bodmin Moor).
(10 cont.) What I found very striking was that apparently disused stone foundations would be turned into habitable and snug abodes each year by the addition of a withy framework covered in matting. This leads me to the speculation that some of the stone house and hut foundations in our area may have undergone a similar process each summer. It is also interesting to note that long- disused and overgrown stone-built kharkas can be seen which are very similar in size and appearance to the remains of Medieval/Post-Medieval longhouse sites in Wales.
Note 11: This is a much bigger subject than can be tackled here in any detail, but the following is a personal selection of places or locales which appear to meet some of the criteria set out above: The main rivers where they pass through core areas are obvious candidates e.g. the crossings of the lower Conwy, and the Dee between Corwen and Llandrillo. However, it is surprising how little metalwork has been recovered from the rivers (and indeed from the lakes) in the region to date.
(11 cont.) In the Carneddau we might consider the quartz outcrop on Creigau Gleison overlooking Llyn Cowlyd and also Bwlch Cowlyd itself. Bera Mawr and Bera Bach are huge piles of rock like megalithic monuments (which parallel the tors of South West England.). Yr Aryg is a smaller version of these. All lie close to a main route across the range. Aber falls is a spectacular natural feature in a heavily used landscape. Ffynnon Caseg and Llyn Dulyn are examples of atmospheric high lakes above well-used cwms. In the Northern Carneddau, sites may include Maen Amor (a massive and prominent crest-top boulder at 741744), Penmaenmawr Mountain, the vicinity of Y Teiryd (confluence of three streams overlooked by Dinas).
Moelfre (above Penmaenmawr) is interesting because the legend of the stone women of Moelfre relates to the hill rather than the circle below it (Styles 1990 28-9).
(11 cont.) In the hills above Llangollen the spectacular stony watercourse dropping down from near the watershed sanctuary beside Creigiau Eglwyseg looks interesting. In the Berwyn, Afon Rhyader is a spectacular natural feature in an area of archaeological significance. Cadair Idris offers several possibilites such as the peak itself, Llyn Cau, and to the north west, Llynnau Gregennan (atmospheric lakes near a routeway and standing stones). To the south west of Cadair Idris is Craig yr Aderyn, a majestic rock outcrop with a massive scree slope overlooking the Dysynni, topped by a small hillfort.
(11 cont.) Possible places in the Rhinogau are the split stone at 6419, and the natural outcrop on which Bryn Cader Faner stands beside a natural routeway at the end of a narrow valley with some small lakes nearby. This was possibly a vantage point for Mesolithic hunting/fowling parties. The area of Gwern Einion is also an intriguing location with natural stone features, a Neolithic tomb nearby and a bog with at least one Bronze Age votiveoffering.
(11 cont.) For Snowdon one would obviously cite the summit of Yr Wyfdda itself (marked by a now-destroyed large cairn), and Bwlch y Saethau (pass of the arrows), which drops down to Lyn Glaslyn (Llyn Ffynnon Las), the home of the mythical Afanc (Ross 2001, 96-7; Styles 1990 25-6, 86-7). In the Glyders I would point to Tryfan for its animal-like shape and texture and the anthropomorphic stones on the summit known as Adam and Eve. Cwm Idwal is associated with the legend of Idwal and is overlooked by Devil's Kitchen with little Llyn y Cwn above (Styles 1990 58-60).
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.
We would like to know more about this location. Please feel free to add a brief description and any relevant information in your own language.
Wir möchten mehr über diese Stätte erfahren. Bitte zögern Sie nicht, eine kurze Beschreibung und relevante Informationen in Deutsch hinzuzufügen.
Nous aimerions en savoir encore un peu sur les lieux. S'il vous plaît n'hesitez pas à ajouter une courte description et tous les renseignements pertinents dans votre propre langue.
Quisieramos informarnos un poco más de las lugares. No dude en añadir una breve descripción y otros datos relevantes en su propio idioma.