Chapter SevenStalking with Stones on Exmoor
Chapter SevenHard and smooth pinkish brown sandstones generally underlie much of the northern part of Exmoor moor, with grey flaky shales and slates covering much of the area to the south (Edmonds and Williams 1985; Edwards 1999, 2000). This geological structure is broadly reflected in the topography consisting of two principal west-east ridges: one running along the coast and a central ridge that includes the highest land, The Chains and the highest point at 519 m, Dunkery Beacon (Figure 7.1). Replicating the geological axes, a watershed runs roughly east-west across this part of the moor. Rivers or streams flowing north do so swiftly, sometimes through rocky gorges, for only short distances to the northern coast. The local term for these is ‘water’—for example, Farley Water, Hoaroak Water, Badgworthy Water—which infers that they are too small to be properly called rivers and too large and powerful to be referred to as streams, flowing as they do in very deeply incised valleys created in ancient periglacial conditions (Straw 1995). There are some thirty named rivers and
FIGURE 7.1 Exmoor, showing places mentioned in the text.
waters marked by the Ordnance Survey crossing Exmoor averaging about 10 miles in length making a total of 300 miles of significant watercourses together with hundreds more miles of unnamed streams and tributaries (Allen 1978; Bonham-Carter 1991: 81). The rivers flowing south generally have less steep gradients and wind through the confined flat valley bottoms in which alluvial sediments have built up over millennia. The rivers are far longer, connecting the moor with the English Channel. The moor is named after the river Exe, which rises in its centre before flowing south to reach the sea beyond Exeter in East Devon (see Chapter 6). This dual directionality makes Exmoor distinctive both in terms of its own geography and its riverine and coastal connectedness to the outside world.
The northern boundary of Exmoor is created by dramatic sea cliffs, which are also the highest in England. The coastal hills have a distinctive ‘hog’s back’ shape, at first steeply dropping away in a long seaward slope and then plunging vertically to the sea below forming a small vertical sea cliff over which coastal waterfalls plunge (Arber 1911). These cliffs form a formidable barrier to the Bristol Channel with few natural harbours or landing places, creating a distinctively sharp northern edge to this upland world. There is only one area of flat coastline, the 2-mile sweep of Porlock Bay between Gore Point and Hurlstone Point. Here there is a massive curving shingle storm beach, with its pebbles distinctively graded in size (smallest to the east), with brackish inland salt marshes behind. The southwest boundary of the moor is well defined by a third ridge with steep south-facing slopes. Elsewhere, to the south, west, and east, the moor lacks any distinctive edge; instead it slips away, merges, and blends into the surrounding undulating hilly landscapes of North Devon and West Somerset.
Only small boggy patches on Exmoor are associated with the upper parts of the valley systems—nothing like the extensive and treacherous bogs that occur on Dartmoor and Bodmin Moor. Today, the high moors are treeless, and about one fifth of the higher central moorland area is occupied by Molina Caerulea (purple moor grass), a coarse perennial species forming dense tussocks and growing up to 0.7 m high and in wetter areas deer sedge with cotton grass being abundant. Elsewhere, the moor is a mixture of heather, ling, and gorse, with bracken on the drier hill slopes. In spring and winter, the moor is a mosaic of stark contrasts between the bleached dead grasses, which are commonplace on the upland ridges, the brown hues of the dead bracken, and the blackness of areas where heather is dominant. (Sinclair 1970). Briefly, in August and September, the heather transforms into a striking purple carpet locally studded with bright yellow gorse. Parts of the northern coast and the stream and river valleys are thickly wooded in their lower courses, particularly along the course of the Lyn and in the area between Dunkery Beacon and Porlock on the eastern side of the moor, where the extensive Horner Wood consists largely of stunted and often crooked sessile oaks.
The rocks are obviously and dramatically exposed in the coastal cliffs, but elsewhere there is an almost complete absence of outcropping rocks across the moor. There is only one notable exception, the Valley of the Rocks—but even this is anomalous, being situated close beside and running parallel to the coastal cliffs at Lynton. This extraordinary location is in fact the now dry valley of an ancient river and is the only place where dramatic sandstone rock stacks, or tors, occur (Figure 7.2a). The main river system flowing north off the moor, the Lyn and its tributaries, has created a series of rocky boulder-strewn gorges with many waterfalls in its lower courses before reaching the sea at Lynton. In inland areas, the bedrock is only sparsely revealed as small crags along the valley sides, occasionally jutting out in a series of parallel outcrops, like ribs (Figure 7.2b). Along the Barle, Exe, Badgworthy, Farley, and Oare Water valleys there are also a series of rocky valley floor knolls. Extensive frost-shattered scree slopes along valley sides occur in the northern parts of the moor and on the steep slopes, where these drop down to the coastal cliffs. In general, the farther north you go on Exmoor, the greater the frequency of these scree slopes and rock outcrops, but, for the most part, it is the absence of surface rock exposures that is the defining characteristic of Exmoor.
The high moors of Exmoor are utterly exposed to the winds, whatever their direction. This exposure prevents trees and even shrubs growing at the higher altitudes, where the vegetation consists mainly of heather, gorse, bracken, and purple moor grass. The high moors provide only rough grazing for livestock, and arable cultivation is restricted to pockets of coastal lowland, notably the Vale of Porlock on the eastern boundary. Substantial woodland is confined to the valley systems. In the absence of trees or rocks, there is no protection or cover as the wind scours the open expanse of land. The prevailing southwesterly winds are generally mild, but when the winds blow from the east or the north, it is bitterly cold, chilling to the bone. The only refuge is down in the stream and river valleys, where, not surprisingly, all the contemporary settlements are located. The absence of farmsteads or electricity transmission lines over the moor is evidence of how hostile and exposed the high moor is. Besides the scouring winds there are frequent sea mists and fogs that envelop the high hills, reducing visibility to 50 m or less. The contrast with the surrounding lowlands that may be bathed in sunlight while the moor is shrouded is dramatic. On a clear day, views from Exmoor are particularly extensive to Dartmoor, south Wales, Bodmin Moor, the Mendip Hills, and the East Devon hills. The English Channel and the Bristol Channel can both be seen from the highest points, but when the dense fogs and mists descend, the moor closes in on itself and is shut off from the outside world. Exmoor becomes an interior introspective world of localised geography.
FIGURE 7.2 (A) Castle Rock in the Valley of the Rocks, Lynton; (B) Swincombe Rocks, Challacombe.
In all weathers, the sky is the dominant element of the high moor, with the two elemental planes of land and sky rubbing against each other, one static, the other dynamic, forceful, and ever unpredictable. Because Exmoor is exposed to the maritime climate of the southwest, the weather is constantly changing. The light can change in an instant, one moment being diffuse, the next intense and focussed, shifting from a brilliant clarity to thick obscuring mists in which it is unwise and dangerous to venture from the security of the valleys to the indifferent emptiness of the open moors. Broad vistas across and off the moor become lost in a dense and impenetrable shroud of cloud, making orientation and a sense of place impossible to realise. The hills may be obscured in clouds for many days at a time before they appear again, opening up to the wider world beyond. The sunlight slanting through the clouds may occasionally individually highlight such features as part of a ridgetop or deep-sided gullies, clefts, and valleys, which otherwise are not distinctively dramatic in this landscape. So, in a largely undifferentiated landscape, it is the changes in the light that create the spectacle of the landscape. One moment a particular hill is brought into prominence, and then it fades away among the backdrop of other hills as the light passes from it. Such changes in visual experience alter one’s perception and experience of the landscape, which become foregrounded or backgrounded, as in Gestalt experience.
Exmoor opens out and closes in on itself on a regular basis. The fogs and mists most often shroud the high hills, leaving the valleys free; at other times, they descend to the rivers and stream courses, covering the landscape in a soft grey and dripping blanket, encouraging the growth of tree ferns and unusual hanging lichen growth on the trees in the deep valleys.
Perhaps the single most important defining feature of Exmoor is rain—in terms of frequency, total volume, and, on some occasions, sheer intensity. Because of the mean altitude of the landscape, Exmoor has a distinct microclimate, whereby it can be sunny in surrounding areas and yet raining heavily on the moor. Not surprisingly, facing the direction of the maritime winds, the southwestern edge of the moor has the highest rainfall. Annual rainfall on The Chains reaches over 2,000 mm, whereas in surrounding areas it is half that amount. Between Dunkery Beacon and the Vale of Porlock, the rainfall drops by half in as little as a few miles (Pearce 2001: 35). The whole upland area of the moor acts as a huge sponge, retaining considerable amounts of water that occasionally reach their limit with dramatic results. The saturation can become so great that the moor resembles a huge lake or reservoir with water running off the highest ground like a continuous sheet, accumulating in the narrow valleys and producing catastrophic flood events such as that which occurred in August 1952 at Lynton, when the escape of floodwater was exacerbated by landslips and blockages. This reservoir effect is created by the absorbent qualities of surface peat and the impermeable geology below, and it enables the rivers and streams to be perennially fed from the rain collected and released from the high moors. The heavy rainfall also fosters peat development, which rises and shrinks depending on the moisture content of the ground. This shrinking and rising of the peat may obscure, or reveal, many of the lithic monuments, whose visibility is also affected by the degree of vegetation growth during the summer months.
Because of the high wind speeds, deep snow is relatively rare on the moor but especially along the coast. In the recent historical past, terrible blizzards occurred in winter, filling the lanes and roads with snow and making movement in to or out of the moor impossible and isolating farms and villages for weeks (Burton 1969: 83ff.; Eardley-Wilmot 1990: 178ff.). Winters can be severe. As recently as 1963, the cold began on 23 December and lasted until 3 March, with the mean day temperature on the Moor being -3 degrees centigrade during those seventy-one days (Burton 1969: 84). Blackmore, in his novel Lorna Doone, drawing on historical records of the 1676 winter, describes the scene dramatically: 'There was nothing square or jagged left, there was nothing perpendicular; all the rugged lines were erased, and all the breaches smoothly filled .... Not a patch of grass was there, not a back branch of a tree [in the Doone valley]; all was white; and the little river flowed beneath an arch of snow; if it managed to flow at all' (Blackmore 1997: 286-287).
Exmoor is the least visited of England’s National Parks, with very few tourists staying within it. The resident population is very low, with about only 10,500 people living within the Park boundary. The main settlements—Lynton and Lynmouth, Dulverton, Porlock, and Dunster—are all along the coast or are on the fringes of the moor, which still seem, despite modern transport links, peculiarly isolated, a point that has been commented on many times (for example, Pearce 2001: 11), and this was also the case in the historic and prehistoric past. There are few substantial upland settlements or field systems on Exmoor belonging to the prehistoric period, which stands in sharp contrast to Dartmoor and Bodmin Moor. Exmoor, with low population and no discernable economic resources, was never Romanised as were other parts of southwest England, although it does have two small military forts, suggesting some exploratory interest. It becomes a compelling view that the people of prehistoric Exmoor principally inhabited not the high moors but the uniquely named and characterful valleys, each with its own personality, as is the case today. These sheltered locations were also the places in which the elemental rocks of the landscape revealed themselves.
The lithic ‘monuments’ recorded here—stone circles, stone rows and geometric arrangements of stones, and stone settings—are all small, discrete, and difficult to find. They are often entirely hidden by long rushes throughout the year or moor grass during the summer and autumn. Often they remain invisible until you reach them. In a sense, their very presence on the moor is unexpected and extraordinary, given the almost complete absence of surface stone on the extensive hills and ridges.
There are only two stone circles known with any certainty on Exmoor (Figure 7.3). The Withypool stone circle is situated on a gentle southwest facing slope half-way down a markedly rounded hill island bounded by the river Barle to the north and east and valleys with small streams to the west and south leading into the Barle. On the top of the hill, there is a large but low summit cairn out of sight from the circle. From it there are extensive views to the south, where the Sidmouth Gap is visible in the far distance, to the west as far as Dartmoor, and east To Dunkery Beacon. Views out from the circle are similar, except that the Sidmouth Gap can be seen only from the upslope northern arc of stones in the ring and is lost from view in the rest of the interior. Both the location and the stones themselves are discrete. All the surviving twenty-seven stones are 0.5 m high or less, and the circle may have had up to one hundred small stones. The most likely source of these is from the stream bottom to the west of the circle. Some of the stones in the northern and southern parts of the ring appear to have been chosen because of the presence of quartz veins (Table 7.1).
While the Withypool circle stands in splendid isolation, the Porlock stone circle is associated with a short stone row 50 m to the southeast and a small low cairn 4 m to the northeast. The stone row is aligned in the direction of the cairn but not the circle. This circle, like that at Withypool, is also situated on a gentle southwest-facing slope, with high land to the northeast restricting visibility in this direction. It is situated close to the head of a valley down which views are seemingly directed, with a stream that runs west and then north to join the Lyn river system and the Bristol Channel. The fourteen surviving stones are irregular sandstone blocks, the largest, about 0.8 m high, being in the southeast and northwest parts of the ring, which also has a small centre stone (Table 7.1; Fig 7.4). Excavations by Harold St. George Gray revealed packing stones around uprights as little as 0.1 m in height. At the geometric centre of the circle, about twelve small stone slabs were revealed but no charcoal or other finds (Gray 1928: 75).
It is interesting to note that these two stone circles, one situated in the southern part of Exmoor, the other in the northern part, are associated with water courses and river systems that flow, respectively, south to the English Channel and north to the Bristol Channel, thus symbolically connecting Exmoor to the outside world. There are no stone settings anywhere near the
FIGURE 7.3 Distribution of stone circles and stone rows on Exmoor.
Table 7.1 The Stone circles of Exmoor: morphology (for locations see Figure 7.3).
| Stone Circle | Diameter | Approx. No. of Stones | Highest Stone | ||
| Withypool | 36 m | 37–100 | 0.5 | ||
| Porlock | 24 m | 14–21 | 0.8 | ||
FIGURE 7.4 Stones in the northwest part of the Porlock stone circle.
Withypool circle and only one within 1 km of the Porlock circle—and since both these circles are surrounded by substantial areas of unimproved moorland, this absence may be significant (see discussion below).
Nine stone rows are documented on Exmoor (Riley and Wilson-North 2001: 24; Riley 2007). They are highly variable in terms of length, orientation, numbers of stones documented, and in terms of their topographic locations (Tables 7.2 and 7.3; Figure 7.5). Three of them, Culbone, Madacombe, and the White Ladder row are long rows exceeding 250 m, while the other six are short alignments, only two of which (Warcombe Water and Wilmersham A) are longer than 50 m. Six are single rows and three are double rows. The numbers of stones present varies between three and well over 160. In all cases the stones are small and virtually invisible from the surrounding landscape.
Table 7.2 The stone rows of Exmoor: morphology (for locations see Figure 7.3).
Table 7.3 The stone rows of Exmoor: topographical locations and visual fields (for locations see Figure 7.3).
| Stone Row | Topography | Visual Field | |||
| Culbone Hill | Runs up east (low)-west (high) slope Parallel with coastline | Extensive to north across Bristol Channel to Wales, east To Hurlstone Point, more limited to west and south | |||
| Cheriton Ridge | On nearly flat ridge top | Panoramic. Most extensive views from any stone row, E toward Lynton and valley head of Farley Water, N to coastal hills and down valley of Farley Water, W to Holdstone Down and S to The Chains | |||
| Furze Hill | On flat ridgetop. Valley head of Warcombe Water prominent. | Panoramic. Extensive views N to coast at Lynmouth, S to The Chains, W to Holdstone Down, E to Brendon Common | |||
| Madacombe | Across south (high)-north (low) slope | Limited in all directions. Overlooks Madacombe valley and valley head visible a short distance to the NE | |||
| Porlock | Across N (high)-south (low) slope | Restricted especially to north by rising ground. Extensive to W across Moor and up-slope to east and valley head | |||
| Thornworthy Little Common | Up west (low)-east (high) slope just above a break in slope to a N-S stream valley | Views extensive to N to coastal hills otherwise encircled by higher ground and hill slopes | |||
| White Ladder | Across S (high)-north (low) slope. Runs down to lowest point on the ridge but terminates just short of it | Restricted to S and E, more open to W, extensive to N | |||
| Wilmersham A | Up SW (low)-NE (high) slope. Axis of row runs down to main valley bottom mirroring natural gullies in facing hillside | Restricted by rising ground in all directions | |||
| Wilmersham B | Across SE-NE slope. Axis ofrow points toward junction of valleys | Restricted in all directions | |||
The most impressive of these rows in terms of the size of the stones is that on Culbone Hill. This row runs up an east (low) to west (high) slope running parallel with the coast with extensive northern views across the Bristol Channel to Wales. Here the widely spaced stones, up to 0.6 m high and 0.9 m broad, are set with their broad faces parallel with the slope, as if meant to be seen from the north or the south. It is closely associated with a series of cairns and barrows to the south and at its western and eastern ends.
FIGURE 7.5 (A) Thornworthy Little Common stone row looking west; (B) Part of the White Ladder stone row looking northwest. The large Setta barrow is visible on the skyline; (C) Looking down the Wilmersham Common stone row. The locations of the stone rows are marked by flags.
The short row of tiny stones on Cheriton ridge occurs in the middle of an almost flat ridgetop, gently rising to the south and affording panoramic views in all directions. There are no cairns in its vicinity, and no stone settings are visible from it. The alignment of three stones on Furze Hill also occurs on a flat ridgetop with extensive panoramic views and is a similarly discrete monument.
The row at Madacombe, by contrast, runs across the contours of a gentle south (high) to north (low) slope overlooking the Madacombe stream valley to the north. It is associated with intervisible cairns at its western and eastern ends. The Porlock double stone row also runs across a gentle north-south slope, is in close vicinity to the Porlock stone circle (see above), and is roughly aligned on a small cairn to the northeast of the circle. The single short row on Thornworthy Little Common runs up a west (low) to east (high) slope and is not intervisible with or closely associated with any other monuments (Figure 7.5a). The White Ladder double stone row (Figure 7.5b) is by far the longest and has a remarkably high number of small quartz blocks less than 0.1 m high, most of which are invisible except in exceptionally dry conditions when the peat shrinks. It runs diagonally down the side of a south (high) north (low) slope and is situated close to a dramatic ridgetop grouping of barrows. It is not aligned in relation to any of these barrows, only a few of which are visible from it.
On Wilmersham Common there is a unique row consisting of two separate interlinked alignments (Figure 7.5c). The longer of these comprises very small rounded single stones about 0.1 m high. The shorter alignment has pairs of similarly small stones. The short row crosses the slope, whereas the longer row runs down the slope. The transition point between these rows and both ends of the row are marked by significantly taller stones about 0.5 m high. This row is discretely located on the slopes of a basin defined by valleys with streams and dry gullies with rising ground on all sides restricting views across the wider landscape. It is not associated with any other monuments in its vicinity.
Each of the rows is unique when we consider their morphology and landscape settings in combination. They are widely distributed across central, northern, and southern parts of Exmoor. Half are associated with cairns and/ or barrows in their vicinity, but none is intervisible or closely associated with stone settings, suggesting that the stone settings constitute a discrete and special set of monuments radically different in both form and social significance.
The unique characteristic of the prehistoric landscape of Exmoor is the presence of numerous geometric stone settings (Figure 7.6). These were first recognised
Figure 7.6 Distribution of the stone settings on Exmoor.
in the early seventeenth century, systematically recorded during the first decade of the twentieth century (Chanter and Worth 1905, 1906) with the most comprehensive and detailed survey documenting fifty-seven surviving monuments (Quinnell and Dunn 1992; Riley and Wilson-North 2001: Appendix 1; since then another has been found: Riley 2007). Given the nature of the heathland vegetation and the small size of the stones, one surmises that there are probably many more yet to be located and that many others on improved land have been destroyed. These stone settings have no convincing parallels elsewhere in Britain and appear to be unique to Exmoor. They are presumed to be of late Neolithic or early Bronze Age date, but there are no radiocarbon dates, and their purpose has remained entirely enigmatic. Exmoor in this respect is both same and different: stone rows and stone circles are found elsewhere throughout many parts of the British Isles and are common on Dartmoor, where the most northerly of the rows is only 48 km distant from the White Ladder stone row (Burl 1976, 1993), but they are not settings of stones of this type. This fact may suggest that they are about internal relationships and particular types of identification of people once living on Exmoor with this particular landscape.
These stone settings consist of geometric arrangements of stones in patterns of variable form. Some appear to consist of triangular arrangements of stones, others rectangular or quadrilateral arrangements or quincunxes (if there is a fifth stone at their centre), parallelograms, and rhomboids of various dimensions. Burl refers to this ‘angular’ megalithic geometry as being ‘like a series of Euclidean exercises’ (Burl 1993: 89). The smaller stone settings may consist of only three stones. The larger ones may have up to fifteen. In many cases, individual stones may have been removed, making the original pattern or design of the stones impossible to ascertain. What appear to be solitary standing stones or pairs of stones today may originally have been part of much more complex geometric arrangements. Consequently, some appear to be random arrangements of stones simply because some are missing—collected over the centuries to create field boundaries and stone walls. Quinnell and Dunn estimate that within the twentieth century one tenth of the recorded settings have been totally destroyed, and of the remainder one quarter are less complete than when originally described by Chanter and Worth and others (Quinnell and Dunn 1992: 4). A concern with geometry and precise arrangements and alignments of these stones is quite clear in those cases where the original arrangements have been well preserved. In some cases, the stones are arranged in parallel rows leading Burl (1993) to attempt to classify them as very short, double, treble, or multiple stone rows, but this seems a rather contrived interpretation, given both the existence of long, single and double stone rows and the very different locations of the stone settings in the landscape (see below). Only three have ever been excavated (see below).
FIGURE 7.7 (A) Toppled slate central stone in the Chapman’s Barrows stone setting; (B) Central sandstone stone in the Brendon Two Gates Stone setting.
Common to all the stone settings, and a feature that they share with the stone rows and the stone circles, is the tiny size of the stones. A stone 0.5 m high is large on Exmoor (Figure 7.7). Such stones are not impossibly heavy to carry, and all the stones used to build a stone setting could be comfortably brought to the construction site by only one or two persons. Even the largest and most complex of the stone settings might easily have been erected by a small group of five or six persons in a matter of days. In sharp contrast to the many large Bronze Age cairns and barrows found across Exmoor, they required only a minimal mobilisation of labour. These settings, like the Exmoor stone rows and circles, are not in any sense monumental at all. Even today, when standing within a stone setting, one notes that many of the individual stones are concealed by vegetation—bracken, heather grass, and rushes—or engulfed in peat. Some of their locations are intervisible, but the stones themselves are not. How are these remarkable yet discrete monuments to be understood and interpreted?
All the known stone settings occur within the heart of Exmoor on high land 5 km or more distant from the coast (see Figure 7.6). The fact that none occur in the vicinity of the Bristol Channel appears to be a genuine feature of their distribution rather than being a simple product of differential destruction, since stone rows and standing stones exist near to the coast, as do areas of unimproved moorland, where they might be expected to survive. They are thus related to the inner part of the moor rather than the coast. Most (90%) occur in a rough west-east band approximately 20 km long and 3 km wide. Ten occur in the vicinity of the headwaters of the rivers Exe, Barle, and Quarme, draining the moor to the south to form one interlinked system flowing into the English channel. The majority (82%) are situated near to the headwaters of rivers and streams draining the moor to the north and the Bristol Channel—the West and East Lyn and their tributaries.
The landscape locations of the individual stone settings are distinctive insofar as they do not occur in any obvious way in pairs or groupings at a short distance from one another. Straight-line distances between the nearest stone settings in the central area of the moor, where they are most densely clustered, vary between a few hundred metres and 1 km, with the majority being situated 250–500 m distant from each other. This parallels their individuality in terms of the variability of their geometric forms and the numbers of stones used to construct them. Although located high up on the moor in a general sense, they are never located in the centres of ridges or on hilltops—or flat ground at the very highest points—but occur most frequently on gently sloping ground near the tops of ridges and hills. This means that, unlike with some of the stone rows, there is never a panoramic view over the whole landscape from any particular stone setting. The view out from them is invariably limited in one or more directions by gently rising land. In other words, they afford distinct viewpoints over particular and specific areas of the moor in particular directions. So, in terms of the wider landscape, they afford restricted views and have a directionality about them. This emphasis on directionality is replicated in the component stones, which are invariably arranged, in the larger and more complex settings, so that the stones run up and down the hill slopes rather than running along or parallel to the contours, a characteristic feature of some of the stone settings. Fifteen representative examples (see Figure 7.8) from across the central part of the moor are now discussed in more detail from the west to the east.
CHAPMAN BARROWS (FIGURES 7.8, 7.9) This stone setting consists of five thin slate slabs forming a quincunx. They now range from 0.1 m to 0.8 m high with a substantial central pillar. The broad faces of all but the easternmost stone cross the slope in a uniform arrangement. The setting is located toward the top of a gentle south-facing slope. Views to the north and the east are restricted by rising ground. They are very extensive to the south down the deeply incised Challacombe stream valley, where the outcropping ribs of Swincombe rocks 750 m distant (Figure 7.2b) are visible along the valley sides and to the west across Barnstaple Bay and as far as Dartmoor.
WOODBARROW (FIGURES 7.8, 7.10) This stone setting now consists of six stones and may originally have been a rectilinear setting made up of more stones. Two upright stones are 0.5 m high and up to 0.4 m broad. Four fallen stones are between 0.7 m and 1 m in length. All are of local slate. The setting is situated on a gentle north-south slope near to the top of a ridge. The east side consists of a row of four stones, three of which are fairly substantial rectangular blocks, the fourth a stump. The west side now consists of two stones that are markedly different in form, being lower and broader. Both are set with their broad faces end on to the slope. These stones are located just over 300 m to the northeast of the head of Yarbury coombe, a stream system forming a major topographic divide in the local landscape and flowing into the river Bray. There are extensive views across the coombe to the south and to the west and east across the moor. To the north rising ground limits the view.
LONG CHAINS COOMBE NORTH (FIGURES 7.8, 7.11) Three stones between 0.3 m and 0.4 m high, forming a triangle, are situated just above the break of slope down to Long Chains Coombe to the south on gently rising ground to the northwest. Just below the stones, the land plunges down to the bottom of the coombe. The easternmost stone in the setting is situated at the point on the slope at which the bottom of Long Chains Coombe first comes into sight as one walks down the slope. From the stones there are dramatic views up the length of the coombe to the southwest. The locations of the Exe Head stone setting 750 m to the south and another stone setting set high up
Figure 7.8 Locations of the stone settings discussed in the text.
FIGURE 7.9 Chapman Barrows stone setting. (Top) Central part of the stone setting marked by flags looking south over Yarbury Combe. (Bottom) Plan after Quinnell and Dunn 1992.
FIGURE 7.10 Wood Barrow stone setting. (Top) Visible stones marked by flags looking north up the hill slope. (Bottom) Plan after Quinnell and Dunn 1992.
FIGURE 7.11 Long Chains Combe (North) stone setting. (Top) Visible stones marked by flags looking south across Long Chains Combe and up to the head of the Hoaroak Water. (Bottom) Plan after Quinnell and Dunn 1992.
on the south side of the same coombe are visible but not the stones themselves. Views to the north and northwest are restricted by rising ground, but there are extensive views to the Bristol Channel to the northeast.
EXE HEAD (CHAINS VALLEY) (FIGURES 7.8, 7.12) This stone setting now consists of at least ten small upright and fallen stones 0.1 m–0.6 m high. They are all of the local pinkish-brown sandstone. It is situated high up on a
FIGURE 7.12 Exe Head stone setting; plan after Quinnell and Dunn 1992; visible stones marked by flags looking south toward the head of the Hoaroak Water.
of these rows on the southeast side form the original arrangement, with that on the northwest side being added later. Most of the stones are arranged so that their broad faces parallel that of the slope, whereas the possible additional gentle north-facing slope. To the north, there is an extensive view across the moor and the Bristol Channel to south Wales. By contrast, to the south the view is limited to a few hundred metres by rising ground and to the west by The Chains ridge about 500 m distant. To the east, it is extensive down the Exe valley as far as Almsworthy Common. From the centre of the stone setting, there is an extensive view down the line of Hoaroak Water flowing north, and the setting is located immediately above the head of this valley and the source of the stream. It is also only 250 m from the extensive boggy area forming the head of the river Exe to its west. The valley immediately below the setting to the north is joined after a short distance by the Long Chains Valley. There are small rock outcrops jutting out of the valley sides and patches of scree where the Long Chains coombe joins the valley.
The setting itself gives the impression of being two parallel NNE-SSW alignments of stones extending over 30 m. Some of these have their broad faces positioned down the slope, others across the slope or set diagonally in relation to it.
FURZEHILL COMMON V (FIGURES 7.8, 7.13) This setting now consists of seven upright stones up to 0.6 m high and 0.4 m broad, together with a series of four shallow erosion holes that probably mark the positions of other stones. The original form of the setting may have comprised three parallel lines of stones forming a rectangle about 20 m long and 10 m broad. The setting is situated on gently sloping ground on a west-east slope running down to the Hoaroak Water to the east. It is positioned high up on the western side of the stream valley on a shelf. The linear N-S axis of the setting runs parallel to that of the valley below. Immediately to the west, the ground rises, restricting visibility in that direction to about 10 m. Looking north, one sees the coastal cliffs, and there are extensive views south to The Chains ridge and to the east across Brendon Common.
HOAROAK (FIGURES 7.8, 7.14) This stone setting now consists of three upright stones 0.1 m–0.4 m high and three fallen slabs forming a pentagonal shaped setting. It is located high up on a west-east slope running down to the Hoaroak Water, which the setting overlooks. Views west are restricted by rising ground. They are extensive north to the sea along the Hoaroak Water valley, east to Brendon Common and south to The Chains ridge.
CHERITON RIDGE IV (FIGURES 7.8, 7.15) This setting consists of four short rows, each including four stones, all of which are sandstone blocks and pillars between 0.2 m and 0.7 m high. Eight are upright, the rest fallen. Positions of others are marked by erosion hollows. Interestingly, not all the rows are of the same length, and it has been suggested that the original form was diamond in plan (Quinnell and Dunn 1992: 10). However, it is possible that three
FIGURE 7.13 Furzehill V stone setting. (Top) Visible stones marked by flags looking east across the Hoaroak Water. (Bottom) Plan after Quinnell and Dunn 1992.
Figure 7.14 Hoaroak stone setting. (Top) Visible stones marked by flags facing southeast. (Bottom) Plan after Quinnell and Dunn 1992.
FIGURE 7.15 Cheriton Ridge IV stone setting. (Top) Visible stones marked by flags facing west toward the head of the Farley Water. (Bottom) Plan after Quinnell and Dunn 1992.
of these rows on the southeast side form the original arrangement, with that on the northwest side being added later. Most of the stones are arranged so that their broad faces parallel that of the slope, whereas the possible additional row has two stones with their broad faces positioned across the slope, two down the slope and the uppermost stone diagonal in relation to it. The setting is situated high up on a southwest-to-northeast slope above the Farley Water with extensive views to the north along it, south to the head of the valley and east across Brendon Common. To the west, they are restricted to a few hundred metres by rising ground.
FIGURE 7.16 Brendon Two Gates stone setting. (Top) Visible stones marked by flags facing south to the head of Hoccombe Water. (Bottom) Plan after Quinnell and Dunn 1992.
BRENDON TWO GATES (FIGURES 7.8, 7.16) This setting is a quincunx with one central pillar, the tallest, in the centre. There are two other surviving stones, an erosion hollow marking the position of a third and the remaining stump of a fourth, making up a rectangular arrangement around the central pillar. It is situated high up on a gentle north-south slope overlooking the head of Hoccombe Water. There are extensive views to the east down the valley, to the west and south across it. Rising ground restricts views north to a few hundred metres.
FIGURE 7.17 Plan of the Lanacombe I stone setting after Quinnell and Dunn 1992, with additions; see also colour plate 6.
LANACOMBE I (FIGURES 7.8, 7.17) This setting is located on a gentle southeast-facing slope near to the top of a ridge that runs northwest to southeast. It is one of a series of six settings located to the south of Lanacombe Hill. Views are restricted to the northwest by the rising ground but extensive to the west and east along the Lanacombe valley and extend about 1 km south to the top of Trout Hill. The stone setting was built overlooking the area where Lanacombe swings around to the north in a great arc, with a side valley running into it dramatically from the southwest. Just below the setting is a boggy area and a springhead producing three parallel rivulets running into Lancombe from the north.
The stone setting is made up of at least eleven upright or recumbent stones and two fallen stones set in a straggling NW-SE arrangement extending over 43 m X 18 m up the slope. They vary between 0.3 m and 0.65 m in height and 0.2 m to 0.4 m in width. All are a fine-grained pinkish-brown sandstone. The stones differ markedly in shape and height. Some have their broad faces placed in the same direction of the slope, others in the opposite direction—across the slope. As one walks through the setting, the arrangements of stones progress and change. There appears to be a definite pairing of stones in terms of the manner in which the stones are set end-on or face-on to the slope. The stones differ markedly both in shape and size, and all have their own individual characteristics. No attempt seems to have been made to collect similar stones, which would have been quite possible, or to erect them in a uniform pattern in relation to one another. One of the uprights is set in a very small low cairn. To its north, 11m distant, there is another small cairn (2.5 m in diameter and 0.2 m high). The original form of the setting was obviously quite complex, and it may well have been altered and extended in prehistoric times. Resistivity survey suggests that the setting may have been erected in an area where the soil was markedly thinner than elsewhere, perhaps even an area with outcropping smaller stones, a feature noted elsewhere at East Pinford and Tom's Hill (Gillings, Pollard, and Taylor 2007). A 2-m-square excavation around an exposed stone hole (H) revealed that it had been created simply by removing outcropping rock from a northeast-southwest aligned oval to a depth of 0.18 m beneath the contemporary ground surface. The base of the hole was levelled with a few flat stones to provide a level base for the upright, which was secured by small vertical and sloping packing stones (ibid.: 11-12). Two large pieces of worked quartz were recovered from the stone hole. A small excavation around a fallen stone (C) at the Lanacombe III setting revealed that this was not situated in an area with outcropping rocks or very thin soils. Here a small ramped posthole was dug and the pillar-shaped upright placed in it and then bedded in place and packed with small stones thrown up from digging the stone hole (ibid.: 25).
TROUT HILL II (FIGURES 7.8, 7.18) The setting is located high up toward the top of a west-east slope. Views out from the setting are limited to the west by rising ground. They are extensive up the higher reaches of the Badgworthy Water to the southeast and down the line of the valley to the north. The setting consists of three small uprights, one fallen stone, and an erosion hollow that contained a fifth stone. This is the remains of a quincunx that survived until
FIGURE 7.18 Trout Hill II stone setting. (Top) Visible stones marked by flags facing northeast. (Bottom) Plan after Quinnell and Dunn 1992.
about 1976, when an unexploded shell was detonated against the central stone (Quinnell and Dunn 1992: 43).
FIGURE 7.19 Great Toms Hill stone setting. (Top) Visible stones marked by flags looking north toward the Badgworthy Water. (Bottom) Plan after Quinnell and Dunn 1992.
GREAT TOM’S HILL (FIGURES 7.8, 7.19) This stone setting now consists of five upright stones and a fallen slab. The presence of erosion hollows suggests that the original form may have been of three parallel rows, each with three or more stones creating a rectangular shaped setting. Resistivity survey found that hollows J and K are impact craters rather than erosion hollows and that the setting might have been aligned along an area of shallower soil above the bedrock (Gillings, Pollard, and Taylor 2005: 7). The surviving uprights, which are of sandstone, are thin posts between 0.3 m and 0.5 m high. They are set across the slope with their broad sides facing down-slope. The setting is situated high up on gently sloping ground rising to the southeast, where views are restricted. They are extensive to the west, north, and south across the moor and in particular over the junction of the Badgworthy Water with Lancombe, where the locations of other stone settings are visible.
EAST PINFORD (FIGURES 7.8, 7.20) The setting now consists of a rectangle of six stones ranging from 0.3 m to 0.7 m high, with its long axis orientated east-west. Circular settings of small stones are visible around four of the uprights. The tallest stones are at the western and eastern ends of the setting, with those at the eastern end set at a noticeable angle to the others. There are numerous small stones in its vicinity, some resembling broken stumps. This is very unusual for Exmoor and might indicate that the original form of the stone setting was larger than that visible today, extending farther to the east. There are also substantial rock outcrops visible beside the Badgworthy Water, a short distance to the west. The flat surface of one of these is covered with a series of distinctive erosion hollows that resemble cupmarks (Gillings, Pollard, and Taylor 2005). The location of this setting is highly unusual, being situated low down in a hidden location in the landscape. It is situated on a very gentle east-west slope but in an almost flat area bounded by stream valleys and ridges and hills to the west, north, and south and rising ground to the east with restricted views in all directions. To the south of the setting, a stream head is visible, one of the sources of the Badgworthy Water.
WESTERMILL (FIGURES 7.8, 7.21) This setting consists of four stones arranged so as to form an irregular rectangle. Four of these are 0.6 m high and about 0.2 m thick, the fourth only 0.2 m in height. They are all slim sandstone posts. The broad faces of three of the stones face uniformly down the slope. The broad face of the fourth stone is set side on to the slope. The setting is situated toward the top of a gentle east-west slope running down to Sparcombe Water and in the vicinity of two very deep valleys running down to the deeply incised Exe valley 1 km to the south. Views out from the setting are restricted to the east by rising ground but extensive to the west along the Exe valley and to the north and south.
FIGURE 7.20 East Pinford stone setting. (Top) Looking west to the Badgworthy Water. (Bottom) Plan after Quinnell and Dunn 1992.
FIGURE 7.21 Westermill stone setting. (Top) Visible stones marked by flags facing southwest toward the Exe valley. (Bottom) Plan after Quinnell and Dunn 1992.
An area defined by three of the four stones was entirely excavated in September 1981. This excavation uncovered the old ground surface on which the stones were erected and identified the socket for a fourth fallen stone (now replaced). No further structural or artefactal evidence whatsoever was recovered nor was there any buried soil suitable for pollen analysis or charcoal (Burrow and McDonnell 1982).
ALMSWORTHY COMMON (FIGURES 7.8, 7.22) This is one of the larger and more complex stone settings on Exmoor. It is situated in the middle of a gentle NW-SE slope. The views out from the setting to the north and the northwest are restricted by rising ground to no more than a few hundred metres. To the east, the visual field extends as far as the ridge of Rowbarrow, studded with large cairns. Views are extensive to the northeast as far as the sea and to the west and southwest across the moor. The stone setting is situated a few hundred metres east of a springhead at the top of a coombe to the east forming part of the Chetsford Water stream system running north, and from it one can look down the stream valley. The setting consists of at least fourteen earth-fast stones and another that is loose on the surface. All are a soft pink-coloured, fine-grained local sandstone. The stones vary in height from 0.1 m to 0.7 m high and 0.2 m to 0.7 m in width and are 0.1 m to 0.3 m thick. The two largest stones in the setting are down-slope toward the lower end. Some of the stones are set so that their broad faces look down the slope. Others have their broad faces set side on to the slope, and a few are diagonally set across the slope. As elsewhere (see below), there appears to be some indication that individual stones, or pairs of stones, were deliberately set so that their broad faces occur at right angles to each other rather than in a uniform fashion, thus differentiating between them. The stones in the setting are all different in terms of both shape and dimensions. Some are flat topped; others have pointed tops; most are irregular in form but a few roughly rectangular.
This setting, first discovered in 1931, was described, somewhat fancifully, as a stone circle made up of thirteen stones with an outlier, forming three concentric ellipses. Quinnell and Dunn (1992: 37) suggest more satisfactorily that it in fact consists of four parallel rows of stones, each with a slightly different northeast to southwest orientation. Undoubtedly some of the original stones making up the setting are now missing, and others have been reduced to stumps.
PORLOCK ALLOTMENT (FIGURES 7.8, 7.23) This is an arrangement of six stones in two groups of three set above and below a large and unusually shaped sandstone block with a small cairn (6.5 m in diameter and 0.5 m high) nearby. The setting is situated at the end of a spur defined by streams flowing to the north and to the south. The setting overlooks the confluence of these streams, which flow north, forming the Weir Water. Views down the
FIGURE 7.22 Almsworthy Common stone setting. (Top) Visible stones marked by flags facing west toward one of the heads of the Chetsford Water. (Bottom) Plan after Quinnell and Dunn 1992.
Figure 7.23 Porlock Allotment stone setting. (Top) Looking west down the Weir Water. (Bottom) Modified plan after Quinnell and Dunn 1992, with additions.
valley to the northwest are extensive but restricted to the east, north, and south by rising ground. The rare large stone—2.5 m long, 0.8 m wide, and 0.5 m high, with a bow-shaped edge facing up the slope—may already have been a significant place on the moor before the stones were erected in its vicinity.
Small cairns, very inconspicuous both in terms of height and diameter (0.1 m–0. 3 m high and up to 5 m in diameter) sometimes occur near to the stone settings or in very close vicinity to them. There are five cases in which there appears to be a direct association. These cairns are always to the east or the northeast of the stones except in the case of the Porlock Allotment stone setting, where the cairn is to the southwest (Figure 7.23). At Lanacombe I, one of the stones is set within an irregular cairn (Figure 7.17). This is interesting in terms of the association between the Porlock stone circle and stone row and a small cairn to the northeast of the circle. As is the case with the stone settings with which they are associated, these small cairns are utterly different both in terms of their size and their landscape locations from the numerous much larger barrows or cairns on Exmoor, most of which are presumed to be of early Bronze Age date. In total, they number around 350 and may be up to 35 m in diameter and as much as 4 m or more high. Such is the size and prominence of some that they have been used as landscape and boundary markers from at least as early as the thirteenth century c.e., which is why so many have names serving to differentiate a landscape remarkably devoid of other distinctive topographic features (Riley and Wilson-North 2001: 32). The principle ridges and summit areas of Exmoor, all the very highest points, are occupied by these barrows, which occur singly, in pairs, or in larger groups, including diffuse clusters and long linear arrangements along ridges and on nearby hill summits (Figure 7.24). These barrow groups are usually intervisible across the moor from one group to another. Some, such as Five Barrows (actually a group of nine), are skyline sited so as to be visible for up to 20 km away off the moor.
The cairns and barrows are clearly part of an early Bronze Age social and cosmological system found throughout Britain and indeed northern Europe. They represent the participation of people in a much more widespread and generalised system of beliefs and practices, whereas the stone settings appear to have a much more localised significance. The new emphasis on placing the dead in cairns on the hilltops of the high moor enabled these locations to command both distant views beyond the moor and across the interior landscape of valleys and water courses occupied by the living. The dead were being elevated away from the living to places in the landscape in which both distance and proximity could be encapsulated.
FIGURE 7.24 The Five Barrows on the skyline seen from the northwest.
It is interesting to note that, from the large hill and ridgetop cairns and barrows, the dramatic outlines of the valley systems are rarely visible except when seen from the ridge ends. One looks across the landscape from these monuments to other hills and ridges studded with their own barrows and cairns rather than down and along the valley systems. Importantly, as already mentioned, the stone settings and the large barrows rarely occur in close proximity. From the stone settings, these large ridgetop barrows are rarely visible. One of the few exceptions is the stone setting occupying the same ridge as the Chapman Barrows on the western side of the Exmoor, but even in this case the highest and most prominent barrow in the group is out of sight. This physical separation sets up a whole series of contrasts between the stone settings and the barrows as follows:
| Barrows Stone Settings | |
| Mass constructed from stones derived from | Specially selected blocks probably from the bottoms of water |
| surrounding area on high ground and quarry ditches | courses |
| Monumental with panoramic views across the landscape | Discrete with strongly directional views down water courses and to valley heads |
| Mass elements put together | Individual elements set apart |
| Round | Non-circular geometric arrangements |
| High visibility | Low visibility |
| Skylined | No skylining effects |
| Massive labour investment | Minimal labour investment |
| Permanent landscape marker | Transient in character |
| Commemorate the dead and dwarf the living | Dwarfed by the living whose relationships they commemorate |
| Suggestive of genealogical ties and distant relationships | Suggestive of individual relationships |
These contrasts are now explored further in a social interpretation of the significance of the stone settings, in an attempt to make some sense of both their geometric forms and their locations in the landscape.
Common to almost all the settings is their intimate relationship to rivers and watercourses in general and the heads of coombes in particular. In all but a few cases, such as East Pinford and Porlock Allotment settings discussed above, they are situated in high locations on hills and ridgetops but never on the very highest points of the moor, which afford panoramic views across the landscape. Because the stones are small, they are imperceptible from the surrounding landscape. Although the locations of nearby settings are often inter-visible, the stones themselves are not. The high landscape locations suggest that these were places from which one looked out across the landscape rather than identifiable places one looked to. The likely sources of the stones are the watercourses themselves, since these are primarily the only places where rock outcrops are exposed and loose stones could be collected. There is no indication that any of these stones were moved any great distance, and whether they are slate or sandstone is a simple reflection of the immediate local geology. Although the stones themselves are not visible from the landscape beyond, people erecting the stones and visiting the settings would have been visible from elsewhere, at least if they had been standing up. Exmoor is a very unusual case in that the people would have dwarfed the stones rather than the other way round. People’s temporary presence might have indicated the position of the stones to others in a fleeting way: here one moment, gone the next. On Exmoor, it was the people rather than the stones that formed the monuments. People revealed their location and thus brought the monuments into being. The stones themselves might have acted as aides-mémoire for the people who erected them. Through the stones one could remember persons and events and stories.
So why erect these tiny stones at all? Their significance has always baffled archaeologists, but the general assumption has always been that they must have had some kind of (unspecified) ceremonial or ritual significance, just as the stone circles and the stone rows did. However, suggesting this does not account for the specificity of either their geometric forms or their landscape locations. They appear to mark places with excellent vantage points along and across specific stream valleys and parts of the moor. Today we might marvel at the fantastic views afforded across the moor that has become romanti-cised and aestheticised through novels such as Blackmore’s Lorna Doone and Williamson’s Tarka the Otter and through numerous guidebooks and picture postcards. But this aesthetic is almost certainly not the experience of, nor is it relevant to, prehistoric lives. People were most likely using and looking out from these settings for a much more pragmatic reason, and the interpretation put forward is that they were specifically watching for game and red deer.
Unlike in other areas of southern and southwest England, there is no evidence of Neolithic monument construction on Exmoor. There is a complete absence of long barrows, megalithic monuments, and causewayed enclosures. A possibly late Neolithic henge monument on the western fringes of the moor (Grinsell 1970: 25) has recently been suggested to be more likely an early eighteenth- or nineteenth-century tree ring enclosure (Riley and Wilson-North 2001: 34). The two large Whit Stones, to the west of Porlock, have been suggested to be remains of a megalithic structure, but this is also extremely unlikely, so why this lacuna? Does it represent an absence of settlement or a different lifestyle in relation to place? The direct evidence for Neolithic occupation of Exmoor remains insignificant and ephemeral at best. The material consists of chance finds of artefacts: a few imported flint and groundstone axes. The former are likely to be from Beer Head in South Devon (see colour plate 1) or from the chalk downlands of Wessex; the latter are of Cornish origin (Grinsell 1970: 23). There is no evidence of agriculture; however, finds of more than a dozen leaf-shaped arrowheads do suggest the importance of hunting, as do finds of discoidal polished flint knives from Furzehill and Kentisbury (ibid.: 25; Riley and Wilson-North 2001: 20).
Environmental evidence for agriculture is slight. Pollen analysis has shown that tree cover existed across Exmoor before blanket peat developed on the hill and ridge summits and probably continued into the historic period on the steeper slopes and valleys. Former woodland cover on The Chains, where the peat has a maximum thickness of up to 3 m, consisted of oak as well as hazel, birch, pine, alder, and elm. At Warren Farm, direct evidence survives in the form of tree stumps preserved below the peat (Straker and Crabtree 1995: 47). Peat began to form on The Chains plateau area around the beginning of the third millennium b.c.e. and during the first millennium b.c.e. on Codsend Moors (ibid.: 45). A number of factors seem to be involved, including altitude, humidity, topography, acidification resulting in podzolisation and tree clearance, and use of fire to clear vegetation. One interpretation is that people and domestic animals provided a final stress on an ecosystem already under stress and that blanket peat initiation on Exmoor correlates strongly with a deteriorating climate (Merryfield and Moore 1974). Just how significant domesticates or human interference were in the disappearance of trees across Exmoor from the Neolithic onward is unknown. At Hoar Moor, the pollen record begins at the same time as the putative Mesolithic/Neolithic transition elsewhere. The pollen sequence indicates the presence of open woodland, with much of the land surface affected by mire development without any evidence for clearance phases or arable farming (Francis and Slater 1990: 22). The evidence seems to indicate that a hunter-gatherer way of life continued on Exmoor well into the Bronze Age. The partial skeleton of an aurochs (wild cattle) under the shingle ridge at Porlock Bay, dated to the early Bronze Age (McDonnell 1998), indicates that there must have been a viable breeding population in the area up to this time, long after its extinction elsewhere in southern England. This fact in itself seems to be a strong indicator of a very low human population density and limited interference in the landscape at this stage. Pollen analysis at Codsend Moors showed that peat development occurred there around 470 b.c.e., when additional climatic deterioration occurred with the onset of cooler and wetter conditions. There was no good evidence for arable agriculture in the earliest levels of deposition, but from the fourth century b.c.e., clearance of trees and the spread of grasses increased dramatically relating to livestock grazing and possible cereal cultivation (Francis and Slater 1992: 26–27).
Today the farming economy of Exmoor is pastoral. The bulk of the moor provides only rough grazing land for cattle and sheep at best. Arable land is found only off the moor in areas such as Porlock Bay to the east and in surrounding lowland areas to the south. Stock rearing took place over most of Exmoor in the historical past. From as early as Saxon times, a large part of central Exmoor known as the Royal Forest was protected under Forest law and used for summer grazing by sheep and cattle. Records show that at the end of the sixteenth century, 40,000 sheep, 1,000 cattle, and 400 horses were pastured annually in the Royal Forest administrative area of central Exmoor, with similar numbers in eighteenth-century estimates (Maltby 1995: 35).
The direct evidence for early Bronze Age settlement and agriculture is slight when compared with other upland areas in southwest England such as Dartmoor and Bodmin Moor. Only ten prehistoric field systems, twenty fragmentary field banks, and forty-five house circles, or platforms, are known from the entire area (Riley and Wilson-North 2001: 42). This may be compared with the presence of fifty to one hundred or more houses in a single settlement on Bodmin Moor (Bender, Hamilton, and Tilley 2007; Johnson and Rose 1994). Furthermore, many of these settlement areas and houses on Exmoor might well be of Iron Age or later date—none have been excavated. The available putative evidence for Bronze Age settlement and agriculture across most of Exmoor seems to indicate no more than scattered individual farmsteads and a few associated fields developed in a piecemeal fashion, suggesting no more than isolated pockets of agriculture. The only more extensive pattern of houses and fields with clear evidence of coaxial field systems of likely early Bronze Age date occurs on Codsend Moors to the southwest of Dunkery Beacon, covering a total area of only about 4 ha (Riley and Wilson-North 2001: 40–51).
Exmoor has long been famous for its wild red deer. It has the largest population in England and is the only area where they have remained continuously in the wild from prehistory to the present. Once widely subject in England to poaching and near extermination as a pest, they survived in Exmoor as royal game in Exmoor Forest. Annual counts suggest that there are around 2,700 within the National Park today, and they have spread to surrounding areas. To many people, Exmoor is synonymous with deer hunting. Hunting has always been the hallmark of an Exmoor identity and relationship to place. This is the last place in Britain where stag hunting has taken place, and it continues in some form despite the recent government-imposed ban on hunting with hounds. The hunt has vibrant local support with many hundreds of horse riders and motorised followers, and it still has a powerful hold on the imagination among some sectors of a deeply rural and conservative Exmoor community. The modern method of deer hunting with horses and hounds is recent. Until the eighteenth century, hounds were used to drive the deer out of their coverts, after which they were then shot with longbow or crossbow (Burton 1969: 55–56). In modern times, deer have been chased rather than driven. Big stags were hunted in the autumn, hinds in the winter and young stags in the spring, a pattern clearly related to the life cycles of the deer, with the calves being borne in June and the rutting season taking place in October. Deer are gregarious and for much of the year move around in herds of between ten and thirty animals, but herds of up to one hundred have been recorded. They live concealed for much of the time in the woods and are difficult to detect. During the winter, they mainly eat ivy in their wooded retreats but come out onto the open moor in spring and summer to graze, which is when hunting was most likely to have taken place during the Neolithic and the Bronze Age.
All the view—the slopes, the wood, the heather—was instinct with the presence of the wild deer; though sheltering in harbour from the heat, they were there. (Jefferies 1892: 44)
The argument made here is that these stone settings were frequented by parties of hunters and might also have provided places in which rituals connecting with hunting might have taken place. More specifically, they would have provided ideal locations in which one could hide and watch. These hunters, perhaps wearing deer hides, would have stalked from the stones, where they would have lain and waited for the deer. The stones could not have been intended to conceal people among them, since they are all so small. They rather marked places from which it was good to hunt at particular seasons and times of the year and according to the direction of the wind. If the stone settings were named places, a party of hunters might have been expected to go out and meet at a particular stone setting. Such hunting groups were likely to have been small, and, regarding the social symbolism and significance of the stones, one possibility is that each stone might have represented an individual hunter with the setting itself representing the hunting group and the social relationships within it. The well-preserved quincunx stone arrangements all have a central tall pillar with four smaller stones arranged around it: the leader of the hunt and the hunting party of five? Many of the settings seem to be composed of between three and eight stones. Sometimes there are significantly more—up to twelve or more at the Almsworthy Common, Lancombe I, and the Exe Head stone settings discussed above.
An implicit assumption has always been that the settings were planned and erected as a single event, but it might equally well be the case that stones were added to particular settings over time and the overall geometric form of the setting altered over time with the positions of the stones being changed. Perhaps this happened in relation to the success of certain hunting locations over time, or it might be linked to the changing generational structures of hunting groups and the significance of particular locales. Alternatively, the larger settings might be places where much larger groups of hunters met.
The settings undoubtedly marked significant places in the landscape, named places to which one could return, discrete markers of place and identity, places from which one watched and waited and made preparations for the hunt. The animal that has always been of paramount significance on Exmoor is the red deer, which has been hunted here from prehistory to the present. The stone settings provided, and still provide, ideal locations from which the movements of the Exmoor red deer can be observed. The acts of waiting for and sensing the deer in the surrounding landscape might well have developed strong social bonds and reciprocal relationships among the hunting parties symbolically expressed by the geometric relationships among the stones, providing material metaphors for the hunting group—alert, watchful, acting together, and having an intimate knowledge of place and the wider landscape.
The individual stones in the settings are shallowly set. The differences in the sizes of the stones are small. There is little indication of any ‘hierarchical’ relationship among them with some stones being much more important— larger and more powerful—than others. The stone settings required very little labour to erect and could have been easily erected by a small hunting group in a matter of a day or days. The geometric and symmetrical relationships among the stones might have provided excellent metaphors for the dynamics of the hunting group. Through the act of erecting the stones, such a group would have been making a claim to a place or a series of interlinked places or territories from which they hunted. It was not necessary that these places be seen from the wider landscape. Indeed, the essence of successful hunting lies in concealment rather than visibility. The stone settings themselves are highly unlikely to have been butchery sites or places where tools were made or fires were lit. The rituals associated with hunting, such as divination and a respect for the bones of the dead animal, would have been unlikely to have left depositional traces in the archaeological record, especially in an area with high soil acidity. All one might expect is small talismans associated with, or perhaps found during the course of, the hunt, such as small pieces of quartz or other unusual stones, and the few excavations that have taken place of the Exmoor stone settings, discussed above, have, perhaps not surprisingly, recovered very little. Deer butchery is likely to have taken place at the site of the kill, with the stone settings representing starting rather than end points for the hunt. These places were repeatedly used and returned to by generation after generation. They marked good places to be and to congregate, memorials to the past successes of the hunting group.
The stones might well have had a moral purpose in terms of expressing the solidarity among members of the hunting group, their social relationships, actions, and events. Recounting hunting stories would have involved recalling the names of the stone settings from which the hunt had commenced. There was a need to mark the place and materialise the significance of these places by erecting stones, because they added power and significance, historical depth, and moral authority. The unique character of each stone setting arrangement, such as the number and disposition of the stones and the manner in which they were set in relation to the hill slopes, enabled each setting to carry its own name and maintain its own identity as a named place. While the stone settings may have had a deeply personal and individual significance to the hunting groups using them, the stone circles and the stone rows are much more likely to have had a collective social significance to communities as a whole.
The stone settings also appear to have been erected in places where it was likely, based on previous experience, that deer would be found. The primary activities taking place at them would have been sitting and waiting and talking and looking out across the landscape. The attributes of the deer—speed, strength, grace, endurance, intelligence, stealth, and cunning recounted in numerous modern hunting stories (for example, Collyns 1862; Evered 1902; Goss 1931; Hamilton 1907; Jefferies 1892)—are precisely those that one might wish to emulate as a hunter. Becoming and thinking like the deer would be the key to a successful hunt, and through the consumption of meat this would become, quite literally, an embodied process. Among ethnographically documented hunter-gatherers, there is frequently a deeply reciprocal and spiritual relationship among the hunter, the prey, and the landscape (for instance, Ingold 2000: 12ff.; Jordan 2003; Tanner 1979), which might be objectified in the form and the character of the stone settings themselves.
In deer hunting, the direction of the wind is absolutely crucial in that one would have to be down wind of the likely places where the deer would emerge. Jefferies describes this well:
In front appears a coombe, overgrown with heather from summit to foot, and I stop suddenly. On the opposite slope are five hinds lying down, their heads visible above the heather, but too far for a good view. To stalk them it is necessary to go round the head, or shallow upper end of the coombe (a mile or nothing), and so get the wind to blow from them. Their scent is so quick that to approach down the wind is useless.... The hollow of the coombe carries the wind somewhat aslant just there from its general direction like a tube, else I think they would have scented me as it is. (Jefferies 1892: 37)
So, according to the wind direction, it would be appropriate to wait quietly at one stone setting rather than at another. Although the views out across the moor and along the watercourses crossing it are always restricted in one or more directions, most parts of the landscape can be seen by visiting a combination of them, and each stone setting may be suggested to have its dominant or primary view (see Figure 7.25). The predominant wind direction on Exmoor is from the southwest or the northwest and, so it is not surprising that
FIGURE 7.25 Dominant views looking out from stone settings around the upper courses of the Badgworthy Water.
the majority of the stone settings are located to the north and the east of the valley systems and combe heads that they overlook.
During the winter, red deer group together in coverts deep down in the heavily wooded valleys of Exmoor. They remain invisible, and hunting in such places is very difficult indeed, unless the animals are driven out from such places onto the open moor, which has always been the traditional practice. It is primarily in the late spring to the autumn that they venture out onto the high moor, to graze during the warmer summer months. So, it would have been from April to October that the stone settings might primarily have been visited and used by groups of hunters. This is the time of year when one is most likely to have long sunny days with excellent visibility. The stone settings were sited at locations where the deer were most likely to emerge from the valleys and pass onto the high open moor.
FIGURE 7.26 Small herd of red deer seen looking west from the Wood Barrow stone setting.
During our main periods of fieldwork in September 2007 and April 2008, we regularly saw small herds of deer from the vantage points of the settings we were surveying on an almost daily basis (Figure 7.26). There also might be a metaphorical relationship between the geometric and complex forms of the stone settings and stag antlers. With their impressive antlers, large stags have always been the symbolically most important and significant of the deer. There is a special terminology for describing the antlers, with a massive horn or beam curving upward and outward, with tines and points projecting out from it. The size varies with both nutrition and age. A male calf has no horns until a year old. In the second year, slender horns grow; in the third year, the brow rights are added, then the trey. The middle or bay rights come in the fifth year, and in the sixth year, points are added on top, with more points growing between the sixth and eight year, all the time the antlers increasing in size, length, and strength (Goss 1931: 77–90) (see Figure 7.27).
Figure 7.27 Full grown stag’s antlers and the terminology used to describe various parts (after Burton 1969).
The whole head of antlers is shed between the middle and the end of April and remarkably regrows and increases in size by late August or early September. This prodigious growth is quite remarkable, providing a ready source of metaphors for place, virility, strength, and regeneration. Impressive individual stags, identifiable from their antlers, were historically given proper names, like persons. Jefferies notes of the stag: ‘He not only lives in the wild, wild woods, and moors—he grows out of them, as the oak grows from the ground’ (Jefferies 1892: 100). We know that antler was widely used during the Neolithic to fashion a whole variety of tools, from digging implements to weapons, and was symbolically deposited in the basal segments of the ditches of monuments elsewhere in southern England. Most recently, antlers have been found in excavations undertaken as part of the Stonehenge Riverside Project (see Larsson and Parker-Pearson 2007), deliberately deposited in the basal segments of the post holes of the southern timber circle in the Late Neolithic henge of Durrington Walls, in the ditch of the Stonehenge Cursus, and in the ditch of the large Amesbury 42 long barrow.
The general arrangement of the stag’s antlers is linear with the points branching off to the sides at more or less regular intervals and clustering at the top (see Figure 7.27). They are symmetrically arranged on both sides of the head. The numbers of tines, or points, are variable, usually between two and fourteen, with the record for Exmoor being twenty (Burton 1969: 49; Jefferies 1892: 85). No two sets of antlers are identical, and perfect ones are rare. The overall geometric arrangement of the antlers, with the points branching off at intervals in different directions, is analogous to the regular and geometric arrangements of the stones in the settings running up and down the hill slopes and placed at different angles to the slope—some facing down the slop, others across it, and others diagonally to it. The individual stones in the settings most frequently occur in arrangements, or groups, of between three and five up and down the slopes, just as the points do on both sides of the stag’s head. Furthermore, the numbers of recorded stones in the settings never exceed fifteen. Most of the better preserved examples have between five and twelve stones, well within the range of the maximum number of points on a mature stag’s antlers. So it seems possible to suggest that individual rows of stones in the stone settings might represent the points on a stylized stag’s antler or, alternatively, the overall arrangement, a stylized stag’s head with points on the antlers.
Over time, some of the stone settings might be expected to grow in size and complexity, just as the size of the antlers on a stag’s head grow in relation to the stag’s age and maturity. Furthermore, the stream valley systems with their dendritic structure of branches at regular intervals, visible when one looks out from the stone settings themselves, also resemble the form of deer antlers, with the deepest and thickest parts low down and the tips reaching up and out across the high moor (Figure 7.28). The landscape itself can be conceived to be in the form of the stag’s head and antlers. Perhaps it is no coincidence that even in the present day this is the logo for the Exmoor National Park. The
FIGURE 7.28 Water courses resembling the forms of stag’s antlers. (Top) View along the course of the river Barle south of Simonsbath. (Bottom) View down the Badworthy Water from Trout Hill.
stone settings placed at strategic points along the valleys and overlooking them may then relate to the ancestral form and significance of stags embodied in the landscape and concomitantly their mythological and social significance, which has continued into modern times. Symbolically, the stone settings are grouped along and within the stag’s antlers, represented by the watercourses spreading out, over and up, and across the moor. The stone settings represent an intimate relationship with watercourses and valley heads within the interior world of Exmoor away from the coast. In the most general sense, they are within the body of the same moor to which they relate. This corporeal identity of Exmoor as a kind of body is further exemplified, as discussed above, by the occasional exposure of rocks, often in the vicinity of the stone settings, jutting out like skeletal ribs along the valley sides.
In prehistory it seems entirely likely that Exmoor was as marginal and peripheral to the social mainstream of events, beliefs, and values as it is today. During the Bronze Age, other moorland areas of southwest England, such as Dartmoor and Bodmin Moor, were centres of tin production and supported large and successful communities. By comparison, the people on Exmoor seem to have remained fairly isolated hunter-fisher-gatherers, well into the Bronze Age and beyond. The physical and cultural isolation of these populations resulted in a deeply conservative ideology, never fully adopting the influences and the values of the outside world and continuing local traditions commensurate with their landscape and its resources.
They had no need to fix their identity or presence their relationship to land in terms of constructing monuments or establishing permanent settlements, boundaries, or fixed settlements. The populations here clearly shared in a general repertoire of ideas concerning the erection of stone circles, stone rows, and large hilltop cairns found elsewhere in the late Neolithic and the early Bronze Age. But what is distinctive about Exmoor is the lack of concern with monumentality, except in relation to hill and ridgetop cairn construction, which is equally ubiquitous elsewhere in southern England. But although the early Bronze Age populations did construct cairns and barrows, many of these have a distinctive local morphology (Quinnell 1988, 1997; Riley and Wilson-North 2001: 34ff.).
The unique forms of the stone settings, with their various geometric arrangements of stones, also indicate a concern to create or to maintain cultural distinctiveness and difference. On the one hand, the tiny character of these lithic monuments can be explained simply in terms of the absence of suitable building stones—but, on the other hand, the desire to construct cultural difference and mark out a particular and different identity may well have been equally strong. After all, it would have been quite feasible to construct massive megalithic monuments in the Valley of the Rocks near to Lynton, where suitable stones abound.
Monumentality was neither required nor intended in the construction of the stone circles, stone rows, or the stone settings. As markers of place, the stone settings, unique to Exmoor, objectified or materialised the significance of that which was unique to Exmoor: the paramount significance of the hunt. Constructed at particular locales overlooking valleys and coombe heads, they were an intimate part of the symbolic, mythological, and social geography of the hunt. They solidified in the landscape places that otherwise would be unmarked and lost and memories that would be forgotten without being objectified in such a place: places at which to meet, places from which to observe and to hunt, places to recount stories, places in which one might honour and represent the deer.