ISSN 1357-4442Editor: Simon Denison

Issue no 34, May 1998


Looking for history under the branches

Numerous clues to all periods of man's past can be found in woodland, reports Chris Gerrard

Woodland, you might think, is the last place to stumble across any archaeology. But in almost every wood in the country it is, in fact, possible to find tell-tale signs of human activity from past centuries - either the signs of woodland management and related enterprises, or of pre-woodland settlement and field systems.

Fieldwork in woods is easier in the winter months, when the vegetation is less energetic. A good and relatively simple approach is to enlarge an Ordnance Survey map to 1:1250 or even larger, and to walk through the block of wood several times until you get a feel for the direction and scale of any earthworks or other features, annotating the base map with measured sketch sections where possible. As you go, it is always worthwhile to check upturned tree boles and spoil around the base of recently planted saplings for pottery and other finds. More detailed fieldwork is often not possible in woodland. Geophysics, for example, is hampered by vegetation and tree roots.

Clues to former woodland management are likely to be all around you. Until recently, virtually all woods were managed either for timber or coppicing. Some older trees grown for their timber, often oak, may survive. Telling the age of a tree is a specialist skill but a mature oak is often 150 years old, often older. Girth is a better indicator of age than height, but there is no standard girth-to-age ratio as the growth of any tree depends on local growing conditions. The age of a tree when felled is given by a count of growth rings across a stump, and it may be possible to estimate the date of felling by the amount of regeneration. I have seen some huge trunks regenerating from an ancient stump, suggesting a felling of long ago.

In a managed woodland, the timber may have been sawn on site, and sawpits can sometimes be identified as depressions, often near major tracks. The trunks were slung over pits to allow one woodman to get underneath the tree while his mate stood on the log above.

Large tree stools may indicate former coppicing, where crops of poles were grown and harvested for fencing, basketwork and fuel. Coppicing largely died out in the first half of this century - with the wider circulation of man-made materials - and formerly coppiced trees may survive with an abandoned crop of poles, each originally the width of a thumb, but now grown to the thickness of an arm.

Except where dry stone walls could be erected, almost every old woodland contains traces of banks, usually topped with hedges, which demarcated coppices from other areas of woodland, and which were built to prevent grazing animals from nibbling new growth. Each area was typically 20-50 acres in extent. The banks are hard to date, but the sinuous and larger scale examples may be medieval, whereas straight edges tend to be characteristic of an enclosure landscape. At Minchinhampton Common in Gloucestershire the woodbanks survive where the woodland has disappeared. Clarendon Park, near Salisbury, is a good example of a medieval wood where these banks can still be seen.

Fieldwork may also reveal evidence for industries attracted to woodland because of the ready availability of cheap fuel. Ceramic production centres of all periods, but typically Roman and medieval, can have a woodland-edge location where clay supplies were at hand. A medieval example is at Laverstock on the edge of Clarendon Park. Signs to look for are waster heaps - the remains of defect pots. Similarly, the combination of local supplies of sand, water and wood was irresistable to the glassmen of the medieval Weald. Yorkshire lead smelters preferred their fuel dried in kilns first and called it `white coal'. There are over 80 white coal kilns of post-medieval date in Ecclesall Woods near Sheffield, which now survive as small keyhole-shaped depressions in the ground.

Iron smelters, on the other hand, used charcoal made by stacking up an inverted basin of poles which could be covered with turves or grass, before burning for a number of days. To retrieve the charcoal, the turves were removed and dumped on the side, often leaving a kidney-shaped mound today next to an area of bare earth or nettles. Scuff the ground with your boot and you may find copious quantities of charcoal and ash - though few artefacts, as wood colliers, at the very bottom of the social scale, carried few possessions. Spoil heaps and hollows for the extraction of stone, clay, and ironstone may also be found close by.

People often want to know the age of a wood. Occasionally it is claimed that a patch of woodland is a relic of Britain's original woodland cover - the so-called `wildwood' that regenerated after the end of the last Ice Age. Research has revealed an extensive suite of beetles, flies, pseudoscorpions, snails, slugs, lichens, mosses and fungi which appear to be confined in varying degrees to `ancient woodlands'. Some plants, known as `ancient woodland indicators' (AWIs), and with entertaining names straight out of the TV programme Blackadder like toothwort, giant fescue and yellow archangel, are also typically slow to colonise recent woodlands and dislike disturbance.

The sad truth, however, is that there are, as yet, no hard data to prove any model of woodland continuity except perhaps in the Scottish highlands and on the remotest and most inaccessible of crags and cliffs elsewhere. One of the few ways to establish the long-term development of a wood is to obtain a pollen core from inside its bounds. This has been done very rarely and, with the exception of some Scottish pinewoods, the cores always reveal disturbance. At Sidlings Wood in Oxfordshire, for instance, where dated pollen sequences have been taken from well documented medieval woodland rich in AWIs, the results indicate long periods of prehistoric clearance and regeneration. The general consensus is that there are, in all probability, no demonstrable cases in lowland Britain of woodland which links back to primary post-glacial woodland.

It may be true that some populations of AWIs represent genetic lineages extending back to the time of the wildwood, but the woodlands they inhabit today are unlikely to have continuity extending much before the birth of Christ, in many cases much less. It is likely, in fact, that many long-established woods have their origin in the post-Roman period. Woods, perhaps surprisingly, quite commonly contain earthwork remains of prehistoric and Roman settlements and field-systems - the signs include lynchets and banked trackways - suggesting a certain amount of regeneration of woodland after that date, although whether the total amount of woodland increased during the period remains uncertain.

It is a myth, though an oft-repeated one, that Henry VIII cleared the majority of English woods for his ship-building programme. The only substantial replantings of modern times have been the shooting belts and ornamental plantations of the 19th century, and the impact of coniferous commercial forestry. Not all the answers about a wood's history can be found in the wood itself. Further information is stored in the record office and local history library, and the best place to begin is with Ordnance Survey and other historic maps. Taking each of the historic maps in turn, it should be possible to see how the woodland has changed shape, where boundaries have been created or removed, and how this sequence compares against any earthworks in the wood. Zigzag edges to a wood, for example, are typical of ancient woods and reflect intakes of farmland, called `assarts', in the medieval period.

Documents are an important source of further information. Post medieval and early modern planting records may still exist as well as medieval sources. Printed collections of medieval documents will introduce you to a world of woodwards and keepers trying to keep track of theft, stray animals and trespass, to the endless repair of hedges and fences and to common rights such as pannage for foraging pigs or firebote, the right to take deadwood for the fire. Much of this detail may feed back into fieldwork where the sources locate specific buildings such as warrener's lodges. The county Sites and Monuments Record (SMR) is another important source of information.

Woodland archaeology needs friends. The number of sites investigated in any detail is pitifully small. Yet, in a sense, woodlands are monuments like castles and long barrows and have complex and exciting stories to tell.

Dr Chris Gerrard is a Lecturer in Archaeology at King Alfred's College, Winchester. Much of his work on woodland has been done with David Clements, a consultant ecologist.

Further reading: Oliver Rackham, The Illustrated History of the Countryside (Weidenfeld, 1994), and The Last Forest (Dent, 1993). Good woodland trails combining archaeological and ecological interests are at Ecclesall Woods near Sheffield and Micheldever Wood near Winchester.

Return to Table of Contents | Return to CBA Homepage

When land first became private property

The spread of enclosed settlements from 1000BC marks the beginning of individual land ownership, says Roger Thomas

Throughout earlier prehistory, people in Britain lived in settlements that were open to the landscape, unenclosed and undefended. Then, in the centuries around 1000BC, a major change occurred. Enclosures, both large and small, began to appear everywhere. People dug deep ditches, and piled the soil up in banks around their settlements, or surrounded them with palisades.

In south-eastern England, some small circular enclosures known as `ring- forts' date to the 10th-9th century BC. The early hillforts of Wessex date to the 8th-7th centuries. In the Scottish Borders, open settlements seem to have been succeeded by enclosed ones in the 8th century. All over southern and eastern England during the 1st millennium, and along the Welsh border, hillforts and enclosed farmsteads were built. In southwestern England we find `rounds' - perhaps 3,500 in Cornwall alone (see BA, February) - and in Wales, we find their equivalent, `raths'. In northern Britain and southern Scotland palisaded enclosures occur widely.

Not all areas were characterised by enclosed settlement, and in some parts of the country (eastern England is one) they do not become common until the second half of the millennium. Open settlements are also less easy to detect archaeologically, and may therefore be under-represented in our records.

The general shift from open to enclosed settlement, however, is undeniable, and seen from a long-term perspective, the change seems sudden and startling. What revolution of human relations could have led to such a vast and ubiquitous change in behaviour?

Archaeologists have long been aware of this change, but the underlying reasons why it occurred have been little discussed. Traditionally, enclosures have been interpreted in functional terms - the role of ditches, banks and palisades as defensive features, as barriers to keep wild animals or stock out of living areas, or as corrals to keep stock in.

More recently, a number of prehistorians have interpreted enclosures in social terms, suggesting that the physical boundary of an enclosed settlement may have symbolised the social division between the group who lived inside and the world beyond. None of these interpretations, however, explain why enclosed settlements became common in the 1st millennium BC when they had been rare previously.

The change, in my view, can be linked to the first emergence of a concept of land as a form of property, following the intensification of agriculture in this period.

During the 4th, 3rd and 2nd millennia (the Neolithic to the middle Bronze Age) traces of structures and other settlement features are mostly insubstantial - where they are found at all - and it seems likely that occupation was often short-lived. Although the societies of this period practised agriculture, it is quite possible that settlement patterns were actually quite fluid, with areas of land being cleared and farmed for a period before the group moved on to another area. The major constructions of the period were ceremonial monuments, such as causewayed camps and henges.

Many of the enclosed settlements of the 1st millennium, on the other hand, contain substantial round-houses, and the settlements give an impression of greater permanence than before. Many were occupied for several centuries, suggesting an enduring link between their inhabitants and the locality in question.

The first half of the 1st millennium was a period of agricultural intensification, especially of arable farming, with the introduction of new crops such as hulled barley and spelt, replacing naked barley and emmer wheat. An extension of agriculture onto heavier soils seems likely, and an increase in alluviation in river valleys has been linked to increased soil movement following more widespread ploughing. New forms of land division were introduced - good evidence can be found in the linear ditch systems of Wessex, for example on Salisbury Plain (See Regions feature) - and increased effort may have been devoted to maintaining the fertility of arable fields. It has been postulated that the period was one of population growth.

Against this background, anthropological work on kinship structures and their relationship to patterns of agriculture and land-holding may provide an explanation of why people began to build enclosed settlements. In his book Production and Reproduction, the Cambridge anthropologist Jack Goody contrasts two different systems of agriculture, each with its own system of kinship. In traditional African societies based on hoe agriculture, land is used at a low level of intensity and there is no shortage of land for cultivation. As a result, there is no strongly developed concept of land as property and the pattern of marriage tends to be `exogamous' (outside the group) because the need to keep landed property within the group is not an issue.

By contrast, in pre-industrial Eurasian societies based on plough agriculture, land is used more intensively and there is only a limited amount of cultivable land available. This leads to a clear concept of land as private property and the pattern of marriage tends to be `endogamous' (within the group or social class) because of the need to maintain the family's hold on its valuable land and property. Goody sees an evolutionary progression from the first of these kinds of system to the second.

Work by Maurice Bloch, Professor of Social Anthropology at the LSE, on two groups of people in Madagascar examined similar issues at a local level. The Zafiminary are mobile `swidden' cultivators. They use land at a low level of intensity and, for them, the principal constraint on agricultural production is the availability of sufficient labour, rather than of sufficient land. The kinship system of this group lacks any sharp distinctions between `insiders' and `outsiders'. The Merina, by contrast, are rice-growers who invest considerable effort in terracing land and irrigating it for cultivation. Because of this investment of effort, land is regarded as valuable property. In order to keep land within the group, the kinship system has strict marriage rules and sharp divisions between `insiders' and `outsiders'. As with Goody's work, there was evidence that the first system could evolve into the second if land for cultivation became scarce.

These observations can be applied to prehistoric Britain. In the 4th to 2nd millennia, land was used at a low level of intensity. There was no shortage of land to cultivate (rather, the shortage may have been one of labour to work the land) and consequently there was no strongly developed concept of land as a form of property. This would have been reflected in the kinship system. However, in the late 2nd and early 1st millennium, agricultural intensification and, possibly, population growth may have led to cultivable land becoming scarcer and, thus, to land becoming valued as private property. This could have led to change is the kinship system and, in particular, to the emergence of new distinctions within the system between `insiders' and `outsiders' because of the need to keep the ownership of valuable land within the group.

While the larger enclosures of the period probably reflected the shared identity of larger groups, the smaller enclosed settle ments of the period are likely to have been the homes of single extended families, representing single land-holding groups. A stronger association of people and places may have developed in the period, with the growth of the notion of `our' home and `our' land.

Other characteristics of enclosed settlements of the period support this suggestion. Often, considerable effort was put into elaborating the entrances and flanking façades of enclosures, and `special' or `ritual' deposits of pottery, animal bone and metalwork were often made in the enclosure ditches. All of these things served to emphasise the distinction between the inside of the enclosure and the world outside. Special deposits in pits inside enclosures may have been connected with fertility, and may have symbolised the link between the inhabitants and the land which they occupied.

The gradual transition, from the Neolithic onwards, from a situation in which cultivable land was relatively abundant, and labour to work it relatively scarce, to one in which the converse was true, must have been one of the major transformations of later prehistory, and one which underlies many of the changes that we see in the archaeological record at this time.

Roger Thomas is an Inspector of Ancient Monuments at English Heritage. A longer article on enclosures by the same author appeared in the Oxford Journal of Archaeology (July 1997)

Return to Table of Contents | Return to CBA Homepage

Discovering the lost kingdom of Radnor

A forgotten kingdom `between the Severn and the Wye' was a major player in Norman Wales, writes Paul Remfry

The modern perception of Welsh History in the Norman period has tended to be focused on Gwynedd, and on the struggle by the princes of North Wales for supremacy amongst the Welsh. In fact, between the Norman Conquest of England in 1066 and the collapse of independent Gwynedd in 1283, Gwynedd enjoyed hegemony only under Llywelyn Fawr in 1217-1240 and Llywelyn ap Gruffydd in 1263-1276.

Outside these two brief episodes, the Welsh were far more likely to turn to their own local lords and princes for protection and advancement. Throughout the 12th century there were four principal competing Welsh `kingdoms' - and of these, one has only recently been recognised by modern research. The four were Gwynedd in the north and north-west, Deheubarth, arguably the most powerful of the four in the centre-west and south, Powys, lying roughly between Shrewsbury, Chester and Aberystwyth, and south of Powys, the newly-recognised Cynllibiwg, lying between Shrewsbury, Aberystwyth and the River Wye.

Cynllibiwg (pronounced, roughly, `Cunthlibiuc') has suffered greatly from scholarly neglect. This kingdom, centred more or less on the old county of Radnorshire, was assumed to be an insignificant place. Its borders, its longevity, even its name were uncertain. Now it can be seen to have been a major player, held in high esteem by Henry II, overlord of all the Welsh princes and King of England for the second half of the century.

The `rediscovery' of Cynllibiwg was the achievement of an independent Herefordshire-based historian, Bruce Coplestone-Crow, during the 1980s, but his work has not yet been widely published. He noticed that a number of variants of the name existed in documents dating from the 9th to the 13th century, all referring to the same broad area. The 9th century Historia Brittonum referred to `Cinlipiuc' or `Cinloipiauc'; the Domesday Book of 1086 has `Calcebuef'; and the 13th century Red Book of the Exchequer - an English royal list of 11th-13th century landholdings - has `Kenthlebiac' in a reference dating to the 1070s. In the late 12th century, the writer Gerald of Wales knew this land as Rhwng Gwy a Hafren, `Between the Rivers Wye and Severn'. Cynllibiwg is a modern spelling.

In 1093, this kingdom was carved up by Norman forces led by Earl Roger of Shrewsbury, Ralph Mortimer of Wigmore, Ralph Tosny of Clifford and Philip Braose of Radnor. That might have been the end of the story if it had not been for the collapse of Norman authority in Wales brought about by the death of King Henry I in December 1135.

The descendants of the old kings of Cynllibiwg immediately fomented a rebellion and by 1148, with Angevin aid, expelled the Mortimers and their royalist allies from most of the lands between the Wye and the Severn. With the accession of Henry II in late 1154, two brothers remained as princes, Cadwallon ap Madog and his younger brother Einion Clud. Cadwallon ruled the larger and richer cantref (roughly equivalent to an English shire) of Maelienydd, in the area around the modern town of Llandrindod Wells, and Einion the more southerly and eastern `mountainous land of Elfael', lying between Radnor and Hay-on-Wye.

In 1165 Cadwallon and Einion Clud combined forces and marched with the rest of independent Wales to join the massed Welsh army under the leadership of Owain, King of Gwynedd, at Corwen in northern Powys. Here, mainly because of thoroughly inclement weather, they managed to humble the great army of Henry II. In 1175, dignified by the titles of Princes of Wales (recorded in the 12th-13th century Chronicle of the Welsh Princes), these two brothers journeyed to Gloucester for a conference with Henry II, with many of their fellow rulers of central and southern Wales.

The chronicle, following the custom of ordering names according to their order of rank, listed the brothers, at this conference, second only to the great Rhys ap Gruffydd of Deheubarth. After the brothers came the mass of other `princes' of South Wales, - those, for example, of Glamorgan, Caerleon, and Upper Gwent.

Two years later Einion Clud met an untimely end, probably returning from Rhys ap Gruffydd's `eisteddfod' at Cardigan over Christmas 1176. With his brother out of the way Cadwallon annexed Elfael. Earlier the same year, Cadwallon had founded a new abbey at Cwmhir, in central Maelienydd, and the lands he granted to the abbey from his domain indicate its extent. These spread from the forests south of Montgomery to the headwaters of the Severn near Mount Plynlimon in the old kingdom of Arwystli, then south to the Wye and down that river to Clifford castle on the bounds of present-day Herefordshire. The extent of this principality is confirmed by the land grants of Cadwallon and Einion and their descendants. The very act of foundation of a monastery, too, appears to confirm the stability of Cadwallon's rule.

Cadwallon went for a further conference with Henry II at Geddington in Oxfordshire in 1177, this time with his fellow rulers of central and northern Wales. He was recorded by a contemporary English chronicler, the Monk of Peterborough, as King of Elfael (`Rex Delwain') - a very rare example of an English usage of the word `king' to refer to a Welshman. An English royal official, Roger Hoveden, later copying this work substituted this title with the phrase `Cadwallon ruler of Elfael' (`Cadewalanos regulus de Delvain'). In the chronicle, the Welsh princes are given the following order: first came Rhys ap Gruffydd of Deheubarth, then Dafydd ab Owain, king of a very shrunken Gwynedd, then Cadwallon, followed by three rulers from the fragmented former `kingdom' of Powys. Finally came `the other nobles of Wales'.

Cadwallon was assassinated in September 1179 by agents of Roger Mortimer. Cynllibiwg was split between Cadwallon's two sons and Ranulf Poer, the Sheriff of Hereford. Henry II, however, was furious, and locked Mortimer up for two years in Winchester Castle. Roger Hoveden wrote that the King's response was unprecedented - to take such action against a Norman noble on account of the murder of a Welshman suggests that the Welshman was no cypher.

After their invasion of Wales in the 11th century, the Normans built numerous castles which, after the Welsh revolt of 1135, were left in Welsh hands. What became of them? It is often said that the Welsh distrusted castles, and it is the case that Welsh castles fell repeatedly to Normans, not so much through direct assault as through lack of effective resistance. Again and again when threatened, Welsh garrisons would abandon their castles for the manoeuvrability of the open field.

Many, however, were taken over and maintained by the Welsh. In Deheubarth, Rhys ap Gruffydd had for his capital the great castle of Cardigan, further buttressed by major castles at Dinevor, Llandovery, Aberystwyth, Nevern and Rhayader. In north Wales, Dafydd ab Owain had his main seat at Rhuddlan castle and other strongholds at Degannwy and Basingwerk.

In Cynllibiwg there lay four major castles, Dinieithon and Cymaron in Maelienydd, Glan Edw (Colwyn) and Painscastle in Elfael. I would argue that Cymaron, now a remote earthwork site half way between Llandrindod Wells and Knighton, was the capital of Cadwallon's domain - partly because after his assassination, Ranulf Poer seized the castle for the Crown. The site, however, has never been excavated. Little is known of Dinieithon, near Llandrindod Wells, which is last mentioned (not as a castle) in 1179. Glan Edw, south of Radnor, was possibly maintained during the period, as it was later destroyed while in the hands of the Braoses in 1196. Painscastle, north of Hay-on-Wye, seems to have survived for many years in the hands of Cadwallon's descendents, various Marcher lords and latterly the English crown.

Cymaron Castle, meanwhile, remained in the hands of the sheriff until 1182, when along with Dingestow and Radnor castles it once more fell to Welsh forces. Finally, according to the Chronicle of the Welsh Princes, Roger Mortimer came in 1195 and drove them away, refortifying the castle for himself. This event marks the end of the independent `kingdom' of Cynllibiwg. During the early 13th century it became a puppet state run by Llywelyn Fawr of Gwynedd.

Cadwallon may in fact have built some castles of his own. Two possibilities would be Buddugre castle high above the River Eithon near Llanddewi Ystradenny and Crug Eryr castle blocking the mountain pass into Maelienydd west of Radnor. It is clear that Crug Eryr belonged to the princes of Maelienydd - Maelgwn, one of Cadwallon's heirs, greeted the Archbishop of Canterbury here in March 1188.

In style, both Crug Eryr and Buddugre appear Welsh. Both sit high on hill-tops, whereas Norman foundations of the period were all in low-lying positions. Buddugre, moreover, has a natural, possibly defensive, town enclosure below, with a rectangular crop mark some distance from the motte which may possibly mark the site the llys, or local royal hall. No excavations, however, have taken place here either. Only with further archaeological research will the full story emerge about the remarkable and overlooked Welsh kingdom between the Severn and the Wye.

Paul Remfry is an independent historian, author of The Castles of Radnorshire (Logaston, 1997) and of many booklets on castles. The Castles of Breconshire (Logaston) will be published this year.

Return to the British Archaeology homepage

Return to the CBA homepage

© Council for British Archaeology, 1998