Current Archaeology 110 from 1988

Click on images to enlarge

My thanks both to 'Current Archaeology' and its publisher Rob Selkirk (Current Archaeology) and as ever to PWD for their kind permission to reproduce this abstract from CA 110 in 1988. It's rather longer than my usual posts but there wasn't really an obvious way in which it could be split into instalments.  

"Crickley Hill 1969 – 1987 by Philip Dixon

Crickley Hill lies on the edge of the Cotswolds, with magnificent views over the valley of the Severn.  The excavations began by investigating an Iron Age hillfort, but soon a Neolithic causewayed camp was discovered, and later a Dark age stronghold.  This view shows the Neolithic entrance passage, looking west.

Twenty years is a long time to spend on a single excavation, particularly when the site is hillfort of modest size.  But the excavations at Crickley will soon enter their third decade, and seem likely to continue for some years yet.  The site has proved to be remarkably rewarding, with a long and complex history: furthermore, no ploughing has damaged the deposits, so even ephemeral traces of past activity can be picked up in the soil.

The project was begun in 1969 with limited aims.  We wanted to dig the entrance and part of the rampart of an Iron Age hillfort, and to dig a few trenches in the interior to try and assess when the fort was built, and whether it was built for permanent or temporary occupation.  The wider aims were unrealistic, given the small scale of work at first, but by the end of the third season in 1971, we had answered most of the questions with which we had started: by then two superimposed hillfort entrances had been explored, and a system of phasing arrived at – a phasing sequence still in use on the site. The Iron Age defences were dated by pottery and metalwork, and soon by radiocarbon, too, to the period c. 700 to 450 BC.  Within the banks, trial holes revealed occupation debris post holes, dense in places, of broadly the same two periods as the defences.

Our problems, however, were just beginning.  Towards the centre of the hill, well within the Iron Age defences, lay a much eroded bank.  Cuttings across this area revealed unmistakably Neolithic assemblages – flint tools and waste flakes in large numbers, and pottery akin to that from the Windmill Hill excavations. Below this earthwork layer completely separate Neolithic bank and ditch, and even at this early stage it was apparent that these two Neolithic banks had been subjected to a whole series of alterations and reconstructions.

The site was clearly sufficiently complex, and well enough preserved to warrant a more thorough examination than was possible in a series of individual cuttings.  So in 1972 we began a programme of stripping, in a regular grid pattern across the top of the hill.  Over the seasons since then we have excavated some 160 10 x 10 metre squares, an overall total of about 17,400 square metres.

At first we were concerned that our returns would diminish as we knew more about the site.  Theory, after all, suggests that a site may satisfactorily be investigated by a limited, often small sample. This has not been our experience at all. Almost every season has produced something new (and has required the modification of our interpretations of other formerly-believed aspects of the site).  Even minor details are proved to have an added significance when seen in the total context of our growing knowledge of Crickley.

Our earliest horizon is a causewayed-enclosure of two circuits of ditches, with at least nine entrances. In comparison with the majority of Neolithic interrupted ditch enclosures, that on Crickley is tiny, no more than two acres in size.  Its history is complicated:  the outer ring was built first; the larger, inner, ring was contemporary with the second phase of the outer ring.  Both were subsequently abandoned and restored on at least two occasions, and in flimsier phase of reoccupation the inner ring alone was rebuilt.  While the relative sequence is clear, the time-span of all this is still doubtful.  Almost all our Neolithic assemblages fit into a single phase of Neolithic material culture.  At the moment, we are proposing a date range from c. 3000 to c. 2600 hundred BC for these enclosures.

The final Neolithic enclosure, a single-ring affair with few causeways, was built after a period of abandonment.  Its narrow entrances, strong fences (and the plentiful evidence from leaf arrowheads and widespread burning) indicate that it was intended for protection, and fell only after a serious assault (see CA 76, May 1981). After the results from Hambledon and Carn Brea, Neolithic defences in Britain can no longer be denied.  Though this enclosure has occasionally been singled out as the “Neolithic Fort” at Crickley, in its layout of entrances, banks and fences, it is not at all unlike the earlier Crickley causewayed-enclosures: these too have now produced leaf arrowheads in significant contexts, and their double rings would clearly provide a defence in depth more effective even than the more obvious “fort ”.  On the whole, then, it seems likely that all these neolithic enclosures provided a protection – and a protection, probably for a settlement, for several phases of roads, fences and houses have been uncovered in the interior.

During the final phase of the Neolithic enclosure a small plateau at the western end of the hill had been segregated from the rest of the site by complex series of fences (see photo opposite). It was approached by a narrow footpath, blocked by two gates, and at the further side of the tiny courtyard stood a small rectangular building, with a hearth in the opening front of it, and with low cairns containing bones, flint and pottery at the extremities of the platform.  Apart from these deliberate deposits, the whole of the area was kept scrupulously clean.  The arrangements resemble those in the Danish Neolithic shrines of Tustrup and Herrup and the Crickley building is provisionally identified as the shrine of the last Neolithic settlement: it was burnt at the same time as the settlement itself.

Sometime now passed – long enough for the Neolithic ditches to silt to about one third of their depth.  The footpath leading to the former shrine was buried in a massive cairn of small rubble, and apparently marked out by upright posts.  On top of the shrine a circle of small stones set upright as a kerb surrounded a huge burning slab, seen opposite.  In its final form, with western circle, long mound and kerb-stones, and a marker or totem at its eastern end, the monument was as long as 100 m, and was isolated from the surrounding countryside by a small trench, cut into the infilled ditches of the Neolithic enclosure. We still do not know when this ritual use of the site came to an end.  Its beginning was probably in the late Neolithic or early Bronze Age; it is possible that it continued in use until shortly before the reoccupation of the site during the building of the Iron Age hillforts.

These early phases of the site are fascinating, but very ticklish to interpret.  The hillfort, in contrast may seem straightforward.  The area of the occupation was more than twice that of the earlier enclosure, and the defence is tall, and military-seeming, with narrow in-turned gatepassage and timber-laced ramparts.

The settlement inside was of two periods.  The first was the more unusual, for it contains several rows of rectangular buildings (see plan, p. 75). These have a very early date, possibly in the 7th century BC, and they remain virtually unique in hillforts in Britain. There was also a series of clusters of small square buildings from which have come thousands of charred grains of cereal.  After a period of occupation only modest length, the fort and its buildings were destroyed by fire.

They were then reconstructed in a completely new arrangement.  The houses were by now the more usual roundhouses, with a particularly large one just inside the entrance (for the chief?).  There will also some 4-post granaries, and the defences were massively improved.  Work on them was still continuing, and an outer ring of vallation had only just been begun, when the fort was once again destroyed by a fire which consumed the roundhouses and granaries, and involved all the timber work in the new defences.  The date of this, the final destruction of the hillfort, seems to have been in the fifth century BC.

At the moment we have no clear evidence of any further occupation of the site for nearly a thousand years.  A change of use, however, came at the end of the Roman period.  At the end of the fourth or in the fifth century AD two separate settlements were built on top of the ruined hillfort about 200 m apart. One was sited close behind the collapsed Iron Age ramparts, and consisted of rows of small slightly sunken huts.  When we first explored them, we thought they were going to be platforms of Iron Age round huts, but they were dated by late Roman material (including a military belt buckle) and grass tempered wares.

Two hundred metres to the west, on the very tip of the promontory was another settlement, enclosed by a strong fence, and entered through a wide gate with a guard house.  It contained a series of buildings, including a granary and at least one substantial framed house.  This too was dated by late Roman material and grass tempered wares.

However there were considerable differences between them in layout and cleanliness; all the finds came from the rather dirty settlement behind the rampart, which we are thus identifying as a peasant village. The settlement on the end of the promontory was much cleaner (fewer fines), and we are thus identifying it as an enclosed high status site.  Both presumably belonged to a sub-Roman British group, perhaps even the descendants of the occupants of one or other of the nearby Roman country houses.  These settlements in common with almost all Crickley’s occupations over three millennia, seemed to have ended through an act of war, perhaps in the late fifth or sixth century.

The destruction of the Dark Age sites at present seems to have been the end of serious occupation on our hill.  In the Middle Ages the top was used by grazing animals, and fragmentary hut on the side of the hillslope may have belonged to a shepherd.  When the agricultural improvements of the 19^th century came to the Cotswolds, Crickley was remote from the areas of arable farming around the villages, and so escaped that destruction by the plough that was so widespread in the region.  Until the start of our excavations it remained little visited, a place where (in Ivor Gurney’s poem about the hill)

The orchis, trefoil, harebells nod all day, /High over Gloucester and the Severn Plain."