More information :
[SD 49237281] HILL FORT [LB]. (1)
A Class A earthwork (i.e. promontory fort) in a perfect position on a prominent limestone hill, naturally defended on three sides by rock terraces and overlooking the sea to the west and the Ingleborough/Settle region to the east. The fairly level summit changes to a long slope on the north, defensively reinforced by three curving, but parallel, stone wall ramparts, originally about 10ft thick. These were constructed of unhewn stone with rubble cores and have various entrances. The inner wall appears to contain a 5ft diameter chamber within its thickness. Possible hut sites (3) are evident and the former existence of "innumerable small oblong barrows of earth and many sepulchral cairns, with two stone cists, cremation remains and pottery" is recorded. (2-3)
The northern hill slope is densely wooded and the stone, both artificial and natural, is so heavily overgrown that only general rampart lines could be traced; no individual features (entrances, hut circles etc) or constructional details could be discerned. Two rubble-stone built enclosures use the shelter of the ramparts and rock outcrop but are probably much more recent than the fort.
Published survey, 25", revised. (4)
SD 492 728. Warton Crag. Listed in gazetteer as a multivallate hillfort with wide spaced ramparts covering 2.6ha. (5)
Additional reference. (Not Consulted). (6)
SD 4922 7289. Warton Crag small multivallate hillfort. Scheduled RSM No 23643. A sub rectangular enclosure of c.3.2ha defended by a combination of rock scarps and steep slopes to the S and W and three stone ramparts to the N and E. The latter are 3m to 7m wide and up to 1.3m high, roughly parallel to each other and c.50-60m apart. Within the enclosure are the boulder foundations of three sub-rectangular huts constructed against a long low rock escarpment. Immediately outside the inner rampart a further two hut foundations are located against the same escarpment. To the S, below the main summit of the crag, faint traces of a bank and ditch have recently been observed along the edges of a limestone shelf. The alleged tumuli observed by antiquarian sources to the N of the outer rampart are not evident earthworks, but may survive as buried remains. The summit cairn is of modern construction. (7)
The hillfort at Warton Crag (SD 4923 7285) was visited by EH in May 1999 as part of the RCHME National SAMs Survey Pilot Project.
The whole of the site is wooded and at the time of the investigation much of the ground surface was also covered by dense vegetation making detailed survey impossible. Despite these conditions the course of all three ramparts - as shown on the 1:2500 survey made by authority 4 - could be traced. The inner rampart is the best preserved and comprises a broad spread of stone containing a small number of upright orthostats. This rampart encloses the summit of the hill, a relatively level area of 2.5ha; no remains of huts or other features associated with the hillfort were identified in this area.
The middle and outer ramparts are extremely poorly defined making it impossible to establish with any certainty whether these features are artificial or natural; it may be pertinent that on the 1st Edition 1:2500 OS map (8) the middle rampart is omitted entirely and that only the western end of the outer rampart is depicted, a section which is coincidental with a pronounced ridge of outcrop.
Two other features were observed, both post-dating the hillfort. A rubble-stone enclosure shown on plan (see authority 4) adjoining the exterior of the outer rampart (NGR SD 4933 7308) appears to be nothing more than surface quarrying of the outcropping limestone. The footings of a sub-rectangular enclosure (also noted by authority 4) located at SD 4933 7287, are built against the south-western end of a linear ridge of outcropping rock which is oriented NE - SW. The enclosure is situated immediately inside the inner rampart which probably formed one of its walls. The enclosure is sub-divided and may have continued to the north-east of the rampart but the earthworks are poorly preserved. The remains are likely to be stock pens of much later date than the hillfort. (9)
The monument is flagged in the latest Arnside and Silverdale AONB Statutory Management Plan as in need of positive conservation management (10a), and was visited with a view to determining the feasibility of archaeological survey.
The site is much as described by authority 9. The dense cover of scrub woodland across the whole hilltop precludes detailed observation and would also make survey extremely difficult at the present time. However, a fairly rapid perambulation located elements of both the inner and outer ramparts although no definite trace of the alleged middle rampart. The inner rampart is the best preserved (and was probably originally the most substantial) but has nevertheless been flattened and/or heavily robbed, and for much of its length survives as no more than a broad flat rubble platform up to 7m wide, overlain by occasional large boulders and in one place an upright stone slab that looks suspiciously like a gatepost. Much of the fort interior and wider hilltop is covered in limestone pavement, but in places this is missing, presumably broken up to provide material to construct the ramparts. Apart from the stock pens described by authority 9, no internal features were seen, but the presence of pavement over large areas of the fort suggests it was never densely or permanently occupied.
Although the site has always been described as an Iron Age hillfort, its scarp-edge location combined with the unstructured form of the ramparts and the unaltered nature of large parts of the internal ground surface must raise the possibility that it is, instead, a hilltop enclosure of Neolithic or Bronze Age date, similar to sites such as Carrock Fell in Cumbria (NY 33 SW 1) and Gardom's Edge in Derbyshire (SK 27 SE 37). (10)
The fragmented earthwork/stonework remains of the defended promontory enclosure described by the previous authorities was mapped from Environment Agency lidar images. Though the site is largely obscured by dense scrub and small trees the traces of the inner rubble rampart could be seen for its entirety making a sub-rectangular enclosure approximately 155m x 200m. The more fragmented remains of the outer circuit could be seen, the northern side largely complete, but the eastern and western ends could not be detected on the lidar images. Between the inner and outer ramparts were the faint and fragmented traces of the middle rampart described by the earlier authorities. The middle rampart remains were approximately 38m beyond the inner rampart and the outer approximately 60m beyond the middle circuit. Traces of two banks were also noted on the lidar images on the terrace below the eastern crags on the same alignment as the inner and middle ramparts. The lidar images were not detailed enough to allow any further inner enclosures and structures noted by previous authorities to be located within the enclosure to be identified. (11)
Warton Crag was targeted for follow-on ground investigation in Stage 2 of English Heritage's NAIS Upland Pilot Project (covering parts of the Southern Lakes, Western Dales & Arnside).
The multivallate enclosure is much as described by authorities 9, 10 and 11. Clear stretches of all three ramparts were identified and walked. The substantial inner (southern) and outer (northern) ramparts are both largely constructed of heaped limestone boulders with only occasional suggestion of stone facing. The middle rampart more clearly comprises inner and outer stone-facing with a rubble core and overall it is much slighter at only about 1.5m wide and up to 0.5m high. Recent clearance of vegetation around the area of outer rampart east of the modern boundary wall (for reasons of rare-species conservation), has exposed a long section of the large rubble bank, fortuitously reducing the impact of intrusive tree growth on this part of the monument. However, elsewhere the formerly more open area at the north end of innermost rampart is now becoming overgrown with young saplings and such like. (12)
Between June 2016 and January 2017, Historic England’s Historic Places Investigation Team based in York undertook aerial mapping and ground-based investigation of newly acquired, fine resolution (0.25m gridded), lidar data of Warton Crag. The lidar survey had been commissioned jointly by Historic England’s Heritage at Risk (HAR) Team in the North West and the Headlands to Headspace (H2H) Landscape Partnership Scheme, as part of ongoing moves to develop a conservation management plan for the scheduled (alleged) hillfort on the Crag summit which has been on Historic England’s HAR Register since 2012 (13a). The new lidar imagery successfully revealed far more of the course of all three of the monument’s walled circuits than had previously been possible to map from the air, and loaded on to a handheld mapping-grade GNSS unit also enabled each circuit for the first time to be followed precisely and examined in detail on the ground. The lidar survey covered an area of 0.96 sq km and extended across the flanks of the Crag in order to contextualise the site in its wider landscape setting (see NRHE ****-****), but ground investigation was confined to the scheduled area on the Crag summit only.
As noted by previous authorities, the inner rampart appears the most massive and best preserved. However, ‘rampart’ seems an inappropriate term to continue to use to describe any of the monument’s circuits. When looked at closely on the ground, the construction of all three circuits is in fact very similar: each originally seems to have comprised an inner and outer face formed of boulders or smaller upright slabs (orthostats) set only 1.8-2m apart, the space in between filled by other stones (many now missing) but where present of a size comparable to the facing orthostats. The term ‘wall’ seems preferable to describe such circuits, which cannot really be described as defensive in character. The implications of this observation for the function and date of the monument are reviewed in more detail below.
In a couple of places the inner circuit has a different earthwork form to that described above, however. For 225m or so in the middle of the circuit, the inner wall has more the appearance of a 2.9m-wide terrace levelled into the hillside, which here is exposed limestone pavement dipping gently to the north and east. The inner edge of the terrace appears cut into the pavement, the outer edge pushed out over it. It is difficult to account for this change of form. Superficially it suggests the wall has been cleared away and its course re-purposed for a terraced trackway, raising the possibility that in the 18th century (when the Crag was the subject of a Parliamentary Enclosure Award (13b)) the inner circuit began to be remodelled in connection with attempted improvement of the hill summit that was never completed. Indeed, about halfway along this stretch of wall course at SD 49287 72923 an upright stone circa 1.4m high on the line of the inner scarp is suggestive of an intended gatepost (as noted previously by Authority 10), but if so, there is no evidence that a gate ever hung from, or fastened against, it. An alternative and perhaps more likely explanation for the very different form of the wall in this sector is simply that it has been heavily quarried, and that the ‘gatepost’ is a piece of limestone pavement abandoned in the process of being levered up and robbed out. Alternatively it may have been intended as a scratching post for cattle. The southernmost 50m of inner wall between a localised limestone scar and the escarpment edge (Beacon Breast) is also of a different earthwork form, increasing in size and changing from a rubbly spread between facing stones to more of an earthwork bank circa 4.3m wide by up to 0.6m high. This may, however, be simply an impression caused by the accumulation of soil and leaf litter over and against stone-faced wall: the tops of large boulders visible within the bank are suggestive of facing stones, in which case the underlying structure and width of the wall is comparable to that of the circuit immediately to the north.
Two entrances through the inner circuit may be suggested, in agreement with observations made in 1789 when the monument was first described (Authority 2a). Each is set a little way back from the escarpment edge (Beacon Breast), one towards the circuit’s north-west end, the other towards the south-east; both are followed by modern footpaths. A third, much wider, break in the course of the inner circuit immediately west of the’ terraced’ section described above is likely to be due to stone robbing.
Immediately east of the north-western entrance in the inner circuit, a sub-rectangular recess or chamber, measuring 2.1m east-west by 1.5m north-south by 1m deep, is visible in the rear of the wall (as first noted by Authority 3). The state of the vegetation means it is impossible to determine from field inspection whether the recess is original or secondary, but in size and general position the feature has clear parallels with structures recorded at the entrances to certain hillforts in southern England and the Welsh Marches, conventionally interpreted as ‘guard chambers’. It is an interesting conjecture whether a similar chamber may once have existed adjacent to the south-east entrance in the inner wall but is now obscured by soil accumulation and leaf litter. However, functions other than guard chambers are possible for such features (13c), and their presence is certainly not sufficient reason by itself to classify the enclosure on Warton Crag as a hillfort.
In 1789, Hutchinson also observed two entrances through the middle wall and three through the outer (Authority 2a). The frequently poorer survival of these two circuits means that it is now impossible always to be certain whether breaks are original entrances or later disturbances, but the middle rampart has at least one good entrance close to its south-eastern terminus above Beacon Breast (in an area now infested by scrappy saplings, and therefore invisible on the lidar imagery) – and may originally have had as many as three – while the outer rampart appears to have four entrances about which we can be reasonably confident, arranged at fairly regular intervals along its length. Most of these are identifiable from having at least one side defined by a large in-situ or fallen transverse orthostat.
The original height of any of the walls is unclear. In most places they survive no more than a single stone (circa 0.5m) high; nowhere do they stand more than 1.5m tall and even at this one location (in the northern arc of the outer circuit, south-east of the modern gate through the field wall) where for a short distance a string of large boulders appears to overlie smaller stones, moss, lichen and other obscuring vegetation prevents fine observation of structural detail. However, the irregular shape and large size of many of the stones used to form the walls, the orthostatic construction technique employed and the relatively narrow width of each wall, all combine to make it extremely unlikely the walls ever stood much higher. Although it is possible they were originally surmounted by some form of timber breastwork, in practice it is difficult to envisage how timbers could have been fixed securely to the stonework beneath (further evidence that the monument was not built with defence primarily in mind, perhaps). The non-defensive nature of the enclosure is suggested, moreover, by the observation that in places the circuits do not follow the most defensible line in terms of maximising the advantages of the natural topography (ie they do not run immediately above steep, localised, scars in the limestone, but are set on level shelves that lie between parallel scars). This is true of the outer wall in particular.
A large oval depression circa 15m across east-west by 11.5m transversely by 0.8m deep towards the centre of the area defined by the inner circuit is best interpreted as a medieval or post-medieval dewpond (it is mentioned by Hutchinson in 1789 (Authority 2a)), while a series of small animal pens defined by boulder walls and built against the base of two of the localised limestone scars, appear on early Ordnance Survey (OS) mapping (13d) and have been noted by previous investigators (eg Authorities 3, 7, 9 and 10): three lie in line north of the south-eastern entrance through the inner circuit and two more outside the northern arc of the outer circuit. (NB Authority 9 is in error in suggesting the latter two are quarrying). All are now heavily overgrown by trees and the boulders covered in moss and lichen, making detailed field assessment impossible, but must be later than the walled circuits they abut.
There is no evidence of structures or occupation of any kind within the enclosure that can be considered to be contemporary with the walls. Indeed, large parts of the interior are today - and presumably were in later prehistory also - characterised by bare limestone pavement. (Evidence for a loess covering of northern English karst landscapes, including specifically at Warton Crag, prior to circa 8200 BC is presented in Vincent et al 2011, but does not bear on the question of how widespread that cover remained after this date (13e)). Such areas are clearly unsuited to being the sites of houses or animal pens because of the fractured and irregular nature of their surfaces, and must be considered to have been equally unsuitable to domestic occupation or the corralling of animals in the past. It is therefore extremely improbable that the interior was ever permanently occupied. It might be argued from this that the enclosure was instead constructed as a place of refuge or last resort, but the defensive shortcomings of the walls already outlined militate against this. Furthermore, there is no source of permanent water within the enclosure that can be considered contemporary with it (see dewpond above).
If not a place of permanent occupation, a defensive stronghold or a place of refuge, therefore, what is the Warton Crag enclosure? It has previously been called a hillfort probably as much as anything on account of it being defined by multiple walled circuits, but its lack of defensibility, small scale of the enclosing walls and multiple entrances militate against such a label: hillforts have ramparts that are more substantial and better constructed, and the majority typically have only one, or at most two, entrances. Indeed, there seems no obvious exact parallel for the monument, although the Gardom’s Edge enclosure in Derbyshire cited by Authority 10 is probably the best. This has recently been excavated and dated to the late 2nd millennium BC. Its excavators have suggested that that site was primarily connected with livestock and acted as a central place for the gathering of local herdsmen where animals could be sorted, exchanged or served, feuds settled or simply bonds between different communities renewed and celebrated (13f). In the absence of other parallels, this currently seems the best explanation and date for Warton Crag, too.
Full details of the aerial mapping and subsequent field assessment and edit of the lidar data have been published as an HE Research Report (13b). Photographs and other information are deposited in the HE Archive. (13)
|