Digging up art
Bronze bog hoard
tv in ba
Editor Mike Pitts
Smoking out the past
Peter Fowler reports on what could turn out to be significant prehistoric remains revealed by a forest fire near the south coast of France, in the Languedoc
After months of drought, on August 4 2003 a careless cigarette started a fire among pine trees on the Causse Méjean. The fire seared through 25 sq km of dense, coniferous forest, moving so fast that it left trees standing but burning so hotly that all vegetation was consumed. A Paul Nash-like landscape of stark black verticals against a black ground was created. Over New Year 2004, it became a striking black on white landscape in sun-sparkled snow. Four months later, rain had washed away the carbonised surface soil and the trees stood against the white of eroded Carboniferous Limestone. Artistic effects apart, the fire exposed and made accessible a hitherto unknown archaeological landscape: welcome to post-pyro archaeology.
Charles Thomas [CBA President 1970-73] and I have been fieldworking the Causse Méjean since 1992. We had already found and published some of its impressive archaeology while writing a book about the area. Just when we thought we were finishing, the situation was transformed by the fire and a locally-based English IT expert, Michael Waygood. With him and his artist wife, Danny, I first inspected a burnt area in detail on the ground in late August: we were lucky enough to walk straight into a hitherto unknown, and apparently extensive, ‘ancient’ field system, marked out by lines of stones on a hillside about 1,000 m above sea level. Quite an altitude for such evidence, and this in a country which does not have, more correctly is not known to have, much in the way of ‘ancient fields’.
We had not previously examined this area because of the denseness (and prickliness – juniper!) of the understorey. Thick pine foliage above meant nothing archaeological showed on air photographs. We had effectively written it off. The ease with which we had come across these remains, however, clearly implied there could be further evidence awaiting us among the blackened trunks. There was.
Over New Year 2004, Priscilla Boniface and I found a trackway and walls linking our August discovery with similar remains on the next hill to the north. We were now looking at not just a field system but a landscape, which we called ‘Le Sauvage’, the nearest available name. We also discovered an apparently separate block of rectilinear fields down below (‘Rivalte’). We bought our own set of GPS equipment and over Easter Michael Waygood and I began recording. What I describe here is based on metrical survey of roughly a square kilometre, perhaps a 10% sample of the whole.
Méjean occupies the southern edge of the Massif Central, some 100 km inland from Montpellier and the Mediterranean, immediately west of the Cévennes. It is part of an extensive calcareous landscape, now divided up into separate plateaux (‘causses’). Some 30 by 20 km in area, the Causse Méjean is like an island, sharply defined by circumambient canyons. Its surface gently undulates mainly around 950 m above sea level. It is wooded in the west where the fire occurred, rising across open sheep pastures to 1247 m. The superficial emptiness of the landscape is echoed by demographics: 1.4 persons per sq km, the lowest in France.
The newly-discovered ‘ancient’ landscape covers at least 1.5 km sq near Mas-St-Chely, incorporating tracts of uninhabited ground between 1,000 and 1,100 m asl. The main visible archaeological features are in stone, notably walls, cairns and trackways, with possible remains of a few rectangular buildings and burial cairns. Long axial walls, roughly straight and up to c 450 m long, are reminiscent of Dartmoor’s Bronze Age reaves. With walls of various shorter lengths, they make conjoined enclosures, characteristically rectangular but often with a rounded corner. On lower ground down a 20º slope, the Rivalte block of 20 narrow stone-walled fields, with lengths of c 150 m, utilises a 20 m unit of width present, though not so ostentatiously, in Le Sauvage too.
Particularly idiosyncratic are integral revetments built across erosion gullies, apparently to create damp areas and perhaps even briefly hold water. Other revetments were built along the contours and across slopes within fields, characteristically detached from other structures. The most obvious feature in the landscape in addition to the walls, however, are supposedly clearance cairns (locally ‘clapas’) – thousands of them. Most were built deliberately, not just heaped up.
This organised landscape is now limestone. Soil has washed off the slopes, surviving in places but characteristically only a few centimetres deep. The contour revetments seem to be late and may represent attempts to stem the loss.
We can resolve extent and morphology by more fieldwork. Date, however, is harder. A relative chronology is indicated by such as a blocked or secondary entrance; but an absolute chronology is lacking. Some ‘clapas’ are clearly recent (18th-20th centuries); at least morphologically, some are demonstrably ‘early’ and many ‘ancient’. Despite the overall excellent state of preservation, everywhere are signs of wear and tear. Numerous breached walls and revetments, for example, bear witness to rushing water as distinct from long-term soil erosion. Furthermore, nearly all the walls have either collapsed or been robbed or both. Thousands of tonnes of stone, laboriously cleared off the land and built into walls, have subsequently been removed.
The recurring 20m width and the attenuated, narrow shape of some fields could suggest a Roman date. They could be prehistoric with Roman or Medieval insertions: judging by extant but pre-modern hedged-field patterns in the region, Le Sauvage and Rivalte might well be Medieval. But if this landscape were in Britain, it would be Middle Bronze Age (c 1,500-1,000 bc).
We have recorded a corner of a foreign field system which is not English. We have discovered a great spread of pre-modern landscape in a remarkable state of preservation and we do not know where to place it between 2,000 bc and ad 1800. This is embarrassing. As far as we are aware there is little if anything in French archaeology locally, and perhaps even nationally, to guide us. Can anyone help? I shall happily risk having oeuf on my visage by saying that I think this landscape is essentially of 1,000 bc +/- 500. We can be confident that Le Sauvage/Rivalte paysage d’autrefois, whenever d’autrefois in this case was, can legitimately take its place among the great, visible ancient farmed landscapes of western Europe.
Peter Fowler (firstname.lastname@example.org) is a landscape archaeologist and former professor at the University of Newcastle upon Tyne. For more on Méjean see Antiquity 73, 1999, 411-19. It is hoped to publish a full account in Landscapes 2004
CBA web:Jan/Feb 2005