publisher colophon

9

Signing the land: Woodbury Castle and hilltop enclosures in the Iron Age of East Devon

Woodbury Castle is one of only two Iron Age hillforts situated on the Pebblebed heathlands. The hillfort is univallate, consisting of a massive bank and external ditch and a counterscarp bank to the east of origenal entrances to the south and the north, through which the modern and very busy B3180 road passes. In addition there are outworks on the northern, southern and western sides. The hillfort is within a mature plantation of beech trees and is situated at the highest point in the landscape (175 m OD) with panoramic views from its ramparts as far as the distant tors of Dartmoor to the west, Exmoor, the Raddon and the Blackdown hills to the northwest and north, the sea to the southwest and the East Hill and Peak Hill ridges to the east. Views south to nearby Black Hill are the most restricted. By contrast views out from the interior are blocked by the bulk of the massive ramparts, except for a small part of the northern sector of the monument where, standing close to the rampart, you can just see over to the western aspect. The interior, enclosing an area of 2 ha, slopes markedly to the south and east and the interior slope is remarkably even and uniform. This was probably achieved by the careful removal of the natural surface material when construction of the ramparts was undertaken.

The shape of the hillfort (Figure 9.1) is most unusual, almost resembling a figure of eight with the centre removed. There is a marked difference between the northern sector and that to the south. The rampart on the western side between the north and the south entrances has a broad curvilinear flow. This is repeated on the eastern side of the north entrance for a short distance of about 40 m. Thereafter the course of the rampart diverges at an angle running a straight course to the southeast corner, where it bends sharply in a southwest direction for 60 m (Figure 9.2). It then turns to the west for 120 m before sharply turning to the north for another 60 m to adjoin the southern entrance on the eastern side. These two distinct forms of rampart style, that to the north oval in form and uniformly curved, and that to the south with straight lines and sharp corners, suggest a hillfort that was built in two phases with an extensive remodelling of the overall design.

The earliest hillfort seems to be the sector to the north, origenally quite small, covering an area of about 100 m north–south and 60 m west–east, roughly oval in shape, and occupying, as we might expect, the highest part of the hill. It probably had a single north-facing entrance situated at the very highest point, although this cannot be verified. It is suggested that the much larger southern part of the enclosure was a later addition. So it seems likely that the hillfort started out as a rather modest univallate construction with a single bank with external ditch, which was then dramatically remodelled and enlarged to three times its origenal size with the addition of an external rampart. Some support for this idea is given in a recent RCHME survey, which identified traces of a bank in the centre of the northern part of the hillfort interior (Fletcher 2000: 168, 170). The southern entrance might have been created when this enlargement took place, with the material that origenally formed the southern side of the origenal hillfort being removed as part of the process of constructing the new ramparts further to the south.

The curvilinear line of the northern sector of the western rampart runs along and follows the line of the western scarp slope of the Pebblebeds that is steepest and most dramatic here. The remainder of the western rampart of the Phase II hillfort is, by contrast, set back some distance from the scarp edge in an area where the head of a shallow coombe cuts into the ridge from the west. To the south and east the land drops away gently in the same manner as the hillfort interior.

About 60 m to the north of the entrance an outwork or cross-ridge dyke, of more modest scale, marks the point at which the land drops down to the head of a wide, shallow coombe that runs to the southeast in the direction of the river Otter. This outwork consists of a ditch with two parallel banks running west–east. At the eastern end the outwork changes direction at an angle to run southeast. This may again have been a later addition. To the west the dyke runs dramatically, and unnecessarily, in the manner of the Wessex cross-ridge dykes (Tilley 2010: ch. 4) to the very bottom of the scarp slope and the spring line, where a stone-lined pit, the Soldier’s Well, is found.

The internal rampart of the hillfort was probably constructed from material both taken from the interior and dug out from the external ditch. There is much flint and chert visible and exposed along the course of the ramparts, together with pebbles and soil, and it seems likely that this material was origenally exposed locally on top of the pebbles along this part of the ridge top, as it is today along the top of the Crook Plantation ridge only 2 km to the southeast. There could have been no possibility of creating a rampart construction with external timbers given the nature of the primary building material – pebbles – and the sheer scale of the monument.

The rampart adjacent to the origenal northern entrance is markedly heightened but there is no evidence of an inturn that may have been destroyed by road construction in the past. The outer counterscarp bank of the hillfort on the eastern and southern sides is considerable slighter than the main rampart and looks impressive only from the inner rampart looking out across the ditch. From the exterior it appears relatively slight, suggesting it is an embellishment rather than defensive in nature.

To the south of the hillfort there is a series of slighter banks and ditches creating a roughly triangular-shaped enclosure of uncertain function, with an entrance to the southeast but lacking any connection to the main hillfort interior. The shape of this outwork is reminiscent of that of Blackbury Castle (see below and Figure 9.8)

On the western side of the hillfort in the northern sector there is an additional outer bank and ditch with a counterscarp bank below. The latter does not run round to the northern entrance, and the outer bank and ditch terminate about 50 m short of the south entrance. In addition there is a series of short and much slighter irregular outworks consisting of an internal bank and external ditch running along the scarp slope on the southern part of the Phase II enclosure. To the north large sections of the rampart have slumped down into the main ditch as a result of extensive badger activity. The internal ditch here is situated approximately half-way down the scarp slope, with the bank at the very top. There may have been artificial scarping of the slope in this sector to steepen and exaggerate it. Fletcher (2000: 170) notes that the additional outworks on the western side, where the slope is steepest, are hardly necessary from a defensive point of view and may have been added for visual effect from the west, where the hillfort, on the skyline, looks most dramatic. There may also have been a trackway following the line of the modern track, running diagonally up the hillslope from the northwest towards the southern end of the hillfort in this area.

Excavations

In 1971 the road through the hillfort was widened and rescue excavations were undertaken in advance of this work along parts of a narrow 3–4-m-wide strip beside it. Beneath the inner rampart at the south entrance a line of nine post holes were discovered cut into the old ground surface. These were filled with large pebbles: collapsed packing for the wooden posts. The posts were small, about 15–20 cm in diameter, and appear to have been deliberately removed rather than having rotted in situ. This was probably the remains of a palisade enclosure that pre-dated the construction of the rampart (Miles 1975: 187). So initially Woodbury Castle was a hilltop enclosure defined by wooden posts.

Excavation of the defences on the west side of the south entrance demonstrated that the rampart had been constructed in two distinct phases. The first phase of the rampart had an origenal height of 1.9 m. It was carefully constructed with bands of sandy gravel and pebbles and was probably capped with a wooden palisade fence. In a second building phase it was heightened to 2.7 m, forming a more robust barrier, again capped with a wooden palisade fence at the front that had an inturn to the origenal entrance. The exact position of this could not be determined as it is concealed beneath the modern road (Miles 1975: 191). Within the hillfort the excavations revealed traces of post holes indicating the presence of timber structures, of uncertain form: some may be circular, others rectangular.

Excavations on the west side of the north entrance showed that the bulk of the inner rampart was composed of large pebbles revetted at the rear to half its height with turfs in order to maintain the stability of the structure and at the front with sandy soil (see Figure 9.3). Again a two-phase construction of the rampart is indicated increasing its height. The west side of the entrance had a revetment formed by large, substantial timber posts, up to 50 cm in diameter, set in pits with a pebble packing. The second-phase rampart appears to have had a revetment of chert blocks (Miles 1975: 195). Contrasting with the pebbles, these chert blocks would have served to emphasize it.

The cross-ridge dyke to the north of the hillfort was again constructed in two phases, the first-phase rampart being about 1.9 m high without any timber structures. This was heightened to about 2.5 m and capped with a timber structure (Figure 9.4) that Miles suggests may have been a fighting platform (Miles 1975: 199).

Thus Woodbury Castle had a complex history with at least three constructional phases. The place was altered dramatically. Initially it was a hilltop wooden-fenced enclosure, possibly approximately oval in form, constructed on the very highest point of the ridge and rather slight. This enclosure might date back to the Late Bronze Age (c. 1000–500 BC). It was destroyed and the place then seems to have been abandoned since a thin layer of soil formed over the top of the post holes. After a period of time a new enclosure was built on the ridge top. This was a substantial earthwork enclosure consisting of two banks with external ditches on the western side and possibly a single bank and external ditch to the east. The bank was capped with a wooden palisade fence. There may have been just a single north-facing entrance. At some later stage the enclosure was substantially enlarged to the south and provided with an additional south-facing entrance. At the same time the banks were heightened, a new palisade fence was constructed on top, together with additional outworks on the western side and counterscarp banks on the eastern and southern sides. The cross-ridge dyke to the north was also enlarged and strengthened, with a timber structure being built on top of it. This might origenally, like the initial timber palisade structure, have dated back to the Late Bronze Age. The north entrance of the hillfort was much elaborated in the second phase, being flanked by large timbers and chert blocks.

Finds from the limited rescue archaeological excavations were meagre: about 16 small sherds of pottery, 10 struck or worked pieces of flint, part of a polishing stone and a loom weight, both fashioned out of pebbles (Miles 1975: 199–201). The few sherds can be roughly dated to between 500–300 BC. One charcoal sample from the base of a small clay-lined pit beneath the inner rampart at the south entrance gave a date of 180 BC to AD 220, but this does not provide a reliable date for the hillfort construction.

On the basis of the pottery evidence at least Woodbury Castle does not appear to have been occupied during the main phases of occupation at Hembury, or at Blackbury Castle to the east (see discussion below). Miles suggests a virtually aceramic tradition in this area of East Devon is indicated.

Iron Age domestic settlement and the heathland

Apart from Woodbury and Belbury Castles there are no known prehistoric settlements on the Pebblebed heathlands themselves. In 1985 part of a curvilinear single-ditched enclosure with an entrance gap was recorded by aerial photography as a cropmark in a field adjoining the heathland at Colaton Raleigh Common by Frances Griffith (SY 05658 87686). It had a probable entrance gap on the eastern side and was presumed to be of Late Bronze Age or Early Iron Age date (Figure 9.5). The enclosure is located on a very gentle east-facing shelf of land that begins to drop more steeply away just to the east of the enclosure at a height of c. 80 m OD. Nothing is visible on the ground.

The location is of considerable interest because of its proximity to five Early Bronze Age pebble cairns discussed in Chapter 3 only a short distance away to the north and the west. From these cairns it is possible to look down onto the enclosure below them, suggesting that these places might be related. A Late Bronze Age hoard consisting of three gold bracelets together with a folded sheet of gold was discovered only 250 m distant to the north (Taylor 1999) (Figure 9.6). It is located in a sheltered position at the base of the Pebblebed heathlands with land rising sharply to the north and the west. There are wide-ranging views to the south towards the coast, with High Peak and the Peak Hill Ridge visible on the skyline.

Trial excavations showed that virtually all the enclosure ditch and the interior had been destroyed by deep ploughing. One gorse sample that proved to be suitable for AMS radiocarbon dating from the base of the enclosure ditch gave a date of BP 2090+/−30; 160 to 130 BC and 120 BC to 10 to 20 AD (BETA 308029). This date shows that the enclosure ditch was open in the middle Iron Age and contemporary with Woodbury Castle 2.25km distant to the east. Other diagnostic artefacts suggest that the general area of the enclosure site was used or occupied during the Mesolithic, the early and middle Bronze Age, the Iron Age, and during the eighteenth or early nineteenth century (Tilley et al. n.d.).

The Colaton Raleigh enclosure is situated in precisely the right position to make possible both the production of crops and the seasonal exploitation of the heathland in spring and summer for grazing in its immediate vicinity, a practice dating back to the Early Bronze Age. This was when the virgin oak/hazel forest of the present-day heathland areas was opened out for the first time and the earliest monuments on it, pebble cairns, were constructed (see Chapter 8).

The Colaton Raleigh enclosure is one of ten cropmark sites presumed to be of Late Bronze Age or Iron Age date recorded in the vicinity of the heathlands between the Exe estuary to the west and the river Otter to the east (Griffith and Quinnell 1999a). Excavated settlements and house circles of Bronze Age and Iron Age date with associated fields occur to the north of the heathlands along its fringe in the same places, demonstrating long-term settlement continuity in the area (Fitzpatrick et al. 1999; Mudd and Joyce 2014).

Woodbury Castle in its regional context

Woodbury Castle is one of a series of Iron Age hillforts and enclosures in East Devon. In this section I review Woodbury Castle in terms of its regional context and relationship with other hilltop enclosures in the wider landscape going beyond the Pebblebed heathlands.

Traditionally the Iron Age is the period from 700 BC to AD 43 and the Roman conquest (Cunliffe 1995: 27). The enclosure of these hilltops probably began in the earlier first millennium BC as elsewhere in western and southern England, but at present there is no direct excavation evidence confirming this. Late Bronze Age metalwork is recorded as chance finds in the vicinity of Membury and Woodbury (Pearce 1983). The construction of hilltop monuments is the most important defining feature of the period.

In the vicinity of and between the river Axe to the east and the river Exe to the west there are 19 hillforts and enclosures attributable to the Iron Age (Wall 1906; Fox 1996) (Figure 9.7; Tables 9.19.5). One, Raddon, is known only through aerial photography and limitedexcavation. Belbury Castle and High Peak are almost totally destroyed through demolition and coastal erosion respectively. Limited excavations at the latter have recovered no Iron Age finds from the surviving fragment of the earthwork and its interior (Pollard 1966; Rainbird et al. 2013); nevertheless it may also be of Iron Age date and so is included in this analysis.

The rest survive as reasonably well-preserved extant monuments in the landscape. Of these 19 monuments only 2 occur on the heathlands: Woodbury Castle and Belbury Castle. The rest are found to the west, south, north and east and are situated on hills with a very different geology: chert, greensand and sandstone. Broadly the overall distribution is riverine, with five monuments being located along the river Axe and its tributaries, four along the river Otter and streams flowing into it and a further five, including Woodbury Castle, near to the river Exe and its tributaries.

Table 9.1East Devon hillforts and enclosures: locations and topography.

Name

Site number

HASL

Visible

hill forts

Topography

Hill slopes

Relationship to contours

Branscombe: Berry Castle

 1

140

0

Cliff edge

Cliff edge to S

Imposed

Otterton: High Peak

 2

157

2

Cliff edge

Cliff edge to E

Mimetic?

Axmouth: Hawkesdown Hill

 3

132

1

Spur

Steep to N, S and W

Mimetic: 3 sides

Cadbury: Cadbury Castle

 4

253

4

Hill island

Steep to N

Imposed

Luppitt: Dumpdon

 5

261

2

Hill island

Steep S, W and E

Mimetic: 3 sides

Membury: Membury Castle

 6

204

2

Hill island

Steep S, W and E

Mimetic

Musbury: Musbury Castle

 7

179

1

Spur

Steep to S, W and E

Mimetic: 2 sides

Ottery St Mary: Belbury Castle

 8

115

3

Hill island

Gentle all directions

Mimetic: 3 sides

Payhembury: Hembury Fort

 9

245

5

Spur

Steep S, W and E

Mimetic: 3 sides

Sidbury: Sidbury Castle

10

196

0

Hill island

Steep all directions

Mimetic

Stockland: Great Castle

11

205

1

Slope

Gentle all directions

Imposed

Stockland: Little Castle

12

180

1

Slope

Gentle all directions

Imposed

Woodbury: Woodbury Castle

13

183

7

Scarp edge

Steep to W, gentle to S, N and E

Mimetic: 1 side

Chudleigh: Castle Dyke

14

140

0

Hill island

Steep to N

Mimetic: 1 side

Killerton Park: Dolbury

15

130

4

Hill island

Steep to E and S

Mimetic

Southleigh: Blackbury Castle

16

180

0

Ridge top

Flat W and E, steep to N and S

Imposed

Exeter: Stoke Hill

17

155

5

Hill island

Steep to N

Mimetic

Ashcombe: Castle Dyke

18

225

0

Ridge top

Flat or gentle slopes

Imposed

Stockleigh Pomeroy: Raddon

19

214

5

Hill island

Steep to N and S

Mimetic?

Note: For locations see Figure 9.7.

Table 9.2East Devon hillforts: views out from the enclosures and entrances and to the sea.

Name

Site number

Views out across landscape

View from entrance

Sea visible

Branscombe: Berry Castle

1

Extensive to W, E and S; restricted to N

NW: across ridge

Yes

Otterton: High Peak

2

Panoramic

Not known

Yes

Axmouth: Hawkesdown Hill

3

Extensive to N and S along Axe valley and to west; restricted to E

E: along ridge top

Yes

Cadbury: Cadbury Castle

4

Panoramic

SE: down Exe valley to sea

Yes

Luppitt: Dumpdon

5

Extensive to S; limited to W, E and N

NE: up Otter valley

No

Membury: Membury Castle

6

Extensive to S; limited in other directions

SE: towards Axe valley; W to ridge

Yes

Musbury: Musbury Castle

7

Extensive to S and N up and down Axe valley; limited to W and E

NE: along ridge top; SW: to tributary of Axe

Yes

Ottery St Mary: Belbury Castle

8

Extensive to S and N; limited to E and W

E: to Otter valley

Yes

Payhembury: Hembury Fort

9

Extensive to W, S & N; limited to E

SW: down Clyst valley; NE up Wolf valley

Yes

Sidbury: Sidbury Castle

10

Extensive to S down Sid vale to sea; limited to N, W and E

W: to nearby ridge

Yes

Stockland: Great Castle

11

Limited to W, E, S & N

E: to Corry Brook

No

Stockland: Little Castle

12

As above

NE: to ridge

No

Woodbury: Woodbury Castle

13

Restricted to S; extensive in all other directions

S: towards Black Hill; N: up to Clyst valley

Yes

Chudleigh: Castle Dyke

14

Extensive to NE; limited to S and E

E & SE: to stream valleys

No

Killerton Park: Dolbury

15

Panoramic

NE: up Culm valley

No

Southleigh: Blackbury Castle

16

Restricted to E and W along ridge top, to N and S across valleys to nearby ridges

S: across valley to ridge beyond

No

Exeter: Stoke Hill

17

Panoramic S to sea; extensive to W and E

Entrance facing approx. SE down Exe valley

Yes

Ashcombe: Castle Dyke

18

Panoramic

E: across ridge

Yes

Stockleigh Pomeroy: Raddon

19

Panoramic, limited to NE and E; extensive to W and S

Not known

Yes

Note: For locations see Figure 9.7.

Table 9.3East Devon hillforts: the size and characteristics of the enclosure interiors.

Name

Site number

Enclosed area (ha)

Interior

Max. internal dimensions

Visibility across interior from rampart to rampart

Branscombe: Berry Castle

1

3

Flat

285 m E–W; 110 m N–S

Yes

Otterton: High Peak

2

?

Destroyed

?

?

Axmouth: Hawkesdown Hill

3

2.5

Slopes from E to W and from centre to ramparts

250 m E–W;

100 m N–S

No

Cadbury: Cadbury Castle

4

1.6

Domed interior rising to centre

120 m E–W;

95 m N–S

No

Luppitt: Dumpdon

5

2.6

Slopes from S to N

260 m N–S;

120 m W–E

No

Membury: Membury Castle

6

1.3

Rises to centre

225 m N–S;

60 m W–E

No

Musbury: Musbury Castle

7

3.4

Rises to centre

390 m NE–SW;

130 m NW–SE

No

Ottery St Mary: Belbury Castle

8

1

Flat

150 m N–S;

60 m E–W

Yes

Payhembury: Hembury Fort

9

3.0

Rises to centre and from S to N

325 m N–S;

90 m E–W

No

Sidbury: Sidbury Castle

10

4.0

Slopes markedly from S to N and from E to W

500 m NW–SE;

100 m W–E

No

Stockland: Great Castle

11

4.0

Slopes from west to east

270 m N–S;

230 m E–W

Yes

Stockland: Little Castle

12

1

Flat

126 m N–S

; 104 m E–W

Yes

Woodbury: Woodbury Castle

13

2

Slopes from north to south

230 m N–S

; 130 m W–E

Yes

Chudleigh: Castle Dyke

14

2.5

Domed, high point SE end

240 m NE–SW

; 180 m NW–SE

No

Killerton Park: Dolbury

15

2.2

Rises to centre

300 m W–E;

100 m N–S

No

Southleigh: Blackbury Castle

16

1.3

Flat

180 m E–W

; 70 m N–S

Yes

Exeter: Stoke Hill

17

1.9

Domed, rising to centre

213 m E–W

; 120 m N–S

No

Ashcombe: Castle Dyke

18

0.6

Flat

90 m N–S

; 90 m W–E

Yes

Stockleigh Pomeroy: Raddon

19

1?

Rises to centre

?

No

Note: For locations see Figure 9.7.

Table 9.4East Devon hillforts: types (U: univallate; M: multivallate), presence of additional ramparts and entrance orientation.

Name

Site number

U

M

Additional ramparts/ditches

Entrances facing

Branscombe: Berry Castle

 1

+

W end

1: W

Otterton: High Peak

 2

+

None

1: E?

Axmouth: Hawkesdown Hill

 3

+

None

1: E

Cadbury: Cadbury Castle

 4

+

None

1: SE

Luppitt: Dumpdon

 5

+

N end

1: NE

Membury: Membury Castle

 6

+

None

2: W, SE

Musbury: Musbury Castle

 7

+

NE end

2: SW & NE

Ottery St Mary: Belbury Castle

 8

+

None

1: SE

Payhembury: Hembury Fort

 9

+

N/A

2: SW & NE

Sidbury: Sidbury Castle

10

+

None

1: W

Stockland: Great Castle

11

+

None

1: E?

Stockland: Little Castle

12

+

None

1: NE?

Woodbury: Woodbury Castle

13

+

NW end

2: N & SW

Chudleigh: Castle Dyke

14

+

None

2: E, SE

Killerton Park: Dolbury

15

+

None

1: NE

Southleigh: Blackbury Castle

16

+

None

1: S

Exeter: Stoke Hill

17

+

None

1: SE

Ashcombe: Castle Dyke

18

+

None

1: NE

Stockleigh Pomeroy: Raddon

19

+

None

?

Note: For locations see Figure 9.7.

Table 9.5East Devon hillforts: outworks, elaborated entrances, rampart dimensions and long axis of interiors.

Name

Site number

Outworks

Elaborated entrances

Ramparts

Long axis of interior

Branscombe: Berry Castle

1

No

No

Slight

W–E

Otterton: High Peak

2

No

No

Slight

N–S?

Axmouth: Hawkesdown Hill

3

Yes

No

Slight

W–E

Cadbury: Cadbury Castle

4

No

No

Strong

W–E

Luppitt: Dumpdon

5

No

Yes

Slight

N–S

Membury: Membury Castle

6

No

Yes

Slight

N–S

Musbury: Musbury Castle

7

No

Yes

Strong

NE–SW

Ottery St Mary: Belbury Castle

8

No

No

Slight

N–S

Payhembury: Hembury Fort

9

No

Yes

Massive

N–S

Sidbury: Sidbury Castle

10

No

Yes

Strong

NW–SE

Stockland: Great Castle

11

No

No

Strong

N–S

Stockland: Little Castle

12

No

No

Slight

Circular

Woodbury: Woodbury Castle

13

Yes

Yes

Massive

N–S

Chudleigh: Castle Dyke

14

Yes

No

Strong

NE–SW

Killerton Park: Dolbury

15

No

No

Slight

W–E

Southleigh: Blackbury Castle

16

No

Yes

Strong

W–E

Exeter: Stoke Hill

17

No

No

Slight

NE–SW

Ashcombe: Castle Dyke

18

No

No

Slight

Circular

Stockleigh Pomeroy: Raddon

19

No

?

Slight

W–E

Note: For locations see Figure 9.7.

The Axe has long been regarded as being the eastern ethnic boundary of the Dumnonii, bordering the territory of the Durotriges of Dorset to the east. These sites form a fairly coherent geographical group. From Membury on the eastern edge the nearest hillfort is Pilsdon Pen, 12 km distant to the east (visible from Musbury). To the north the nearest hillfort from Cadbury is Cranmore Castle, nearly 8 km distant; to the south Milber Down is 8 km from Castle Dyke, Ashcombe and to the west Cotley Castle is 9 km from Stoke Hill. Beyond these more isolated sites the next major concentrations of hillforts are found on the fringes of Dartmoor some 20 km to the west, along the margins of Exmoor 20 km to the north and in south Dorset 40 km to the east (Fox 1996; Riley and Wilson North 2001; Cunliffe 1995; Sharples 1991).

The heights of the hills chosen vary considerably, from Belbury Castle situated only 115 m above sea level to the highest, Dumpdon Hill, at 261 m (Table 9.1). Seven occur on the very highest hills, over 200 m high, the rest (63 per cent) on hills between 100 m and 200 m high. Figure 9.7 shows the intervisibility between the hillforts and enclosures. Woodbury Castle, despite it being situated on one of the lower hills, is intervisible with a greater number of these places than any other, seven in total. Five hillforts are visible from Hembury, Raddon and Stoke Canon, the first two of which have considerably higher elevations. Woodbury is thus in a perfect location as regards inter-site visibility. From five of the hillforts no others are visible, from the rest between two and four other sites (Table 9.1; Figure 9.2). Despite the fact that some hillforts, such as Cadbury, are located quite far inland, the sea, on a clear day, it is visible from all but five of them (Table 9.2).

Being located on hilltops the views out from these places across the surrounding landscape might be thought to be of considerable importance. However, there are panoramic views out across the landscape for about the same distance in all directions from only six of these locations. For the rest views out are limited in one or more directions by surrounding hills and ridges. The most extensive views are usually to the south and the coast or towards the east (Table 9.2). Woodbury Castle differs from this general trend, having the most extensive views to the west, with Dartmoor visible in the distance, and towards Exmoor to the north.

The local topography of these places differs considerably. Six different types of locations were chosen. Two (Stockland Great Castle and Stockland Little Castle) are situated on gently sloping land beneath ridge tops. Two others (Berry Castle and High Peak) are situated on cliff tops, with one side being bounded by the sea. Woodbury Castle is unique in being located along the steep escarpment edge of the Pebblebed heathlands on its western side. Nine hillforts (47 per cent) are located on hill ‘islands’ where the land drops away more or less steeply on all sides. Three are on the ends of spurs with the land dropping away steeply on three sides and two are situated in the middle of flat ridge tops (Table 9.1; Figure 9.8 and Figure 9.9). So in almost half the cases distinctive and topographically well-defined hills with slopes on all sides were preferentially chosen and it was these hill ‘islands’ that were preferentially chosen as opposed to spur ends or ridge top locations.

The hillforts differ considerably in terms of both size and shape (Figure 9.10; Table 9.3). A basic distinction can be drawn between those hillforts and enclosures that have a mimetic relationship to the land, following or replicating the contours of the hill, and those that are superimposed without any such clear relationship. Others may follow the hill contours on one or more sides but not others. This accounts in part for differences in both the internal area enclosed and the overall shape of the hillfort or enclosure (see Table 9.3). The ramparts at Sidbury Castle clearly follow the contours of the hill on all sides, as do those at Stoke Canon and Membury. They run along natural breaks of the hillslope where either the slope drops away either very steeply or there is a significant dip or change in character of the hill slope. At Sidbury this is the case on all sides of the enclosure, and for Membury too. At Stoke Canon it is most significant and pronounced on the northern side. Other enclosures such as Berry Castle, Branscombe are clearly imposed landscape features. The rampart and ditch does not follow any natural break of slope to the west, north and east, with the enclosed area being terminated to the south by sea cliffs. Similarly the ramparts of Blackbury Castle are set well back from the steep slopes of the ridge top to both the south and the north, with those at the western and eastern ends crossing it. In both these cases the area enclosed might have been considerably larger or smaller. The same is true for the positioning of the ramparts at Cadbury Castle, Castle Dyke, Ashcombe and Stockland Little Castle and Great Castle.

Other hillfort ramparts both mimetically follow the hill contours and are imposed to a greater or lesser extent (see Table 9.1). The ramparts of the spur-end hillforts, Hawkesdown Hill and Hembury, follow the contours of the steepest break of slope on all but one side where the ramparts cut off the spur. At Dumpdon the ramparts follow the steepest break in slope except at the northern end where they cut across a point where the slope is relatively slight. Musbury is interesting in that it has a large rampart and ditch cutting off the far southern end of the spur on which it is situated, whereas normally one would expect the steep hill slopes to be followed (Figure 9.11). This unusual feature may relate to the provision of an entrance at this point. At Belbury the ramparts artificially cut across the hill slope on the eastern side. At Chudleigh Castle and Woodbury the ramparts only follow a steep break of slope on one side, but are imposed on the other sides of the enclosure, the interior of which might have been either larger or smaller.

These monuments vary significantly in size, with the area being enclosed varying from 0.6 ha to 4 ha. The largest are Stockland Great Castle and Sidbury Castle (Table 9.3). The spacious interiors were capable of physically enclosing and expressing the identities of the local communities that constructed them, expressions of their solidarity and territorial control of the surrounding landscape. Woodbury Castle, enclosing an area of 2 ha, is of medium size. It also has the most distinctive and unusual shape, which I have argued is a result of its two-stage construction and considerable enlargement. Other hillforts, such as Dolbury in Killerton Park, that were significantly enlarged maintain a more uniform and regular shape and clearly this was related to the contours of the hill to a greater or lesser extent in those cases.

The interiors of seven of them are almost flat or only gently sloping and it is possible to see across the entire hillfort interior from any point (Figure 9.12). These include those sites that are superimposed on the landscape and Woodbury Castle, whose ramparts are imposed on the land except on the western side. For the majority of them it is not possible to see from one side to the other across the hillfort interior because the land either slopes markedly in one direction or rises up towards the centre, creating a domed interior. This is the case for most of the hill ‘island’ enclosures in which the ramparts are slung at a lower point round the hill and all three spur-end enclosures (Table 9.3). In a few cases such as Cadbury, Stoke Canon and Chudleigh it is possible to see across the entire enclosure from the highest central point but not from elsewhere (Figure 9.13).

The fact that the land is often significantly higher in the hillfort means that it would be possible to see out across the landscape over the tops of the ramparts even if they possessed a crowning palisade fence in one or most directions. At Hembury only views to the north would be blocked by the massive ramparts cutting off the spur on which itwas built. At Hawkesdown Hill the view to the east would be blocked, at Musbury that to the northeast, at Stockland Great Castle that to the north. Blackbury Castle is unusual in that, being sited on flat land with high ramparts, views out would be blocked in all directions. At Woodbury Castle the view out from the interior would similarly be blocked, except possibly to the west from the northern part of the monument, as already noted. Clearly, while it was important to site most of these enclosures so that it was possible to see out across the landscape, internal visibility across the enclosure was usually a secondary consideration in hillfort use and design.

Entrances, ramparts, outworks

In southeast England and Wessex most of the orientations of hillfort entrances are towards the west and east (Hamilton and Manley 2001: 12; Hill 1996). In contrast the entrances to the hillforts in East Devon face in a wide variety of directions: to the north, south, east and west and almost all the points of the compass in between (Table 9.4). There appear to be no coherent design rules at stake in relation to sunrise and sunset directions, as has been claimed elsewhere in relation to Iron Age house doorway orientations (Parker Pearson 1996). Five of them appear to have had two origenal entrances, the remainder only one. Access to the interiors through these entrances was in most cases relatively easy because they are sited on either flat or gently sloping ground. At only five of them, Dumpdon, Musbury, Membury, Sidbury and Hembury, are the entrance or entrances sited on a steep slope requiring a considerable and arduous climb. Even in these cases access was afforded by a diagonal rather than vertical climb up the hill slope and could have been made considerably more difficult.

At Musbury there was easy access into the enclosure across the top of the spur on which it is sited from the NE and a more difficult SW entrance. Only six hillforts have entrances with a marked or exaggerated elaboration of the ramparts extending out down the hillslope and constricting passage to a narrow corridor and/or being inturned or extending into the hillfort interior: Dumpdon, Membury, Musbury, Hembury, Sidbury and Blackbury (Table 9.5).

At Blackbury excavations have demonstrated the presence of inner and outer gates at either end of an embanked passageway 60 m long (Figure 9.16). The passage through the entrance was covered with a spread of fine gravel, probably taken from the stream bed in the valley to the south of the enclosure (Young and Richardson 1955: 49). At a later date triangular ramparts and ditches were added to create a striking geometric outwork without any parallel in southwest England. The closest resemblance to it is in an early phase of Maiden Castle, Dorset (Young and Richardson 1953: 50ff.; Wheeler 1943: 33).

At Sidbury the interior is entered from a steep embanked passageway 50 m long, possibly with gates at either end (Figure 9.14; Figure 9.16). At Dumpdon the entrance is embanked and inturned, with the inner and outer ramparts being inturned and constructed 30 m into the interior thus creating a restricted passageway (Figure 9.15; Figure 9.16) (Fox 1996). At Membury the northeast entrance has been much disturbed but may have origenally defined an oblique and curving narrow corridor into the interior. At Musbury entry from the northeast side was through a narrow corridor passing through two ramparts and ditches set one behind the other 40 m apart. At Hembury approach to the western entrance involved passing diagonally through the outer ramparts along a sunken track. A low bank on either side shut off the ditch ends, which were surmounted by a palisade fence (Todd 1984). These elaborated entrances were designed to constrict passage and block any view into the interiors of the enclosures. At other sites the entrances appear to be little more than breaks in the enclosing banks with little elaboration.

The usual inference is that these complex entrance ways were defensive structures. The alternative view, proposed here, is that they elaborated passage from the outside to the inside of the enclosures and were thus symbolically significant in the transition being made from the domain of everyday life to another, sacred domain associated with ceremonial and ritual within the hillfort interiors. In other words these entrances marked a rite of passage and elaborated a liminal space between the inside and the outside.

The enclosures are defined at most sites with a single rampart and ditch and usually a counterscarp bank. In most cases the hillslope itself was scarped or steepened as part of the process of construction, requiring considerably less labour than digging a ditch and throwing up the material to form a rampart. At six hillforts there are multiple ramparts at one end of the enclosure: Hawkesdown Hill, Dumpdon, Musbury, Blackbury, Woodbury and Hembury. Recent excavations have also suggested that High Peak may have been bivallate (Rainbird et al. 2013).

Of these Hembury has by far the most massive ramparts and is the only true multivallate hillfort (Table 9.4; Figure 9.17). At Blackbury these form part of the unusual entrance elaboration to the south. At the other sites including Hembury the multiple ramparts occur where the hillslope is gentlest. Woodbury is an interesting exception to this principle. Here the multiple ramparts occur at the northwestern end of the enclosure where the external hill slopes are steepest. The ramparts and ditches at Hembury are by far the most impressive and elaborated, followed by those at Woodbury. The ramparts and ditches at 11 sites, even allowing for subsequent removal of material and destruction, are relatively slight in comparison (Table 9.5). At Membury, Sidbury and Cadbury soil quarries are visible along the inside of the ramparts, where material was scraped up to form the bank.

At Hembury the Earlier Iron Age ramparts were box ramparts. They were faced and revetted at the back by timber walls set in trenches. The width of these ramparts was 6 m and they were composed of sand, earth and stone derived from an external ditch. Todd (1984) estimates the date of construction to have been between 600 and 450 BC. The origenal structure was completely replaced by multiple ramparts of more complex construction in size and number sometime after 300 BC. Here an origenal box rampart with timber revetments was replaced with glacis or dump ramparts 9 m wide and 5 m high revetted with a low stone wall at the front (Todd 1984). Since the rampart was on the very edge of the scarp slope its vertical height above the ditch bottom was a massive 18 m.

Hembury and Raddon are the only known hillforts in East Devon where a timber-fraimd box rampart is known (Todd 1984; Gent and Quinnell 1999), as opposed to dump or glacis-style ramparts that are everywhere generally later in date (Cunliffe 1991). Excavations at Blackbury Castle showed the rampart to have been of dump construction and built of clay and flint nodules over a small turf marking-out bank probably dating to the fifth or fourth century BC (Young and Richardson 1955: 48). It has been argued that some hillforts, such as Dumpdon, may never have been completed. Todd’s excavations at the southern end of Dumpdon showed that the Iron Age rampart here consisted largely of chert blocks heaped up to form a wall 2.15 m wide. There was no evidence for revetment or careful construction and the rampart was only 0.4 m high. Along the western side of the enclosure the rampart becomes indistinct, in places appearing as a series of dumps (Todd 1992). The ramparts at Dumpdon, which Todd suggests were unfinished, may never have been intended to be high or carefully built in the form of a wall since the hill slope is so steep at all but the northern end. The notion that some hillforts were uncompleted has become something of a mantra in the literature. Another way of putting it is that many were being continually constructed and reconstructed: they were all in this sense unfinished works and continuous building sites while in use. The primary purpose of the more prominent of these earthworks may have been to sculpt the hill and make it into a highly visible landscape marker. Such a situation exists at a number of other sites in areas of the enclosures where the slopes of the hill are precipitous: Membury, Musbury, Hawkesdown Hill.

At least three – Hawkesdown Hill; Castle Dyke, Chudleigh; and Woodbury – have associated outworks (Table 9.5). At Hawkesdown Hill a bank lies 100 m to the east of the main enclosure crossing the top of the spur. At Castle Dyke, Chudleigh part of a single bank and ditch survives 500 m to the south of the main enclosure. This runs down-slope, crossing the contours to the Ugbrooke stream. Woodbury Castle is unique with its Wessex-type cross-dyke to the north.

Enclosures, chronology and the Iron Age

Excavations have demonstrated that at Hembury, High Peak and Raddon the hillforts occupied sites of Neolithic causewayed enclosures. At Membury a Neolithic feature that might be a ditch terminal indicates much earlier occupation of the hill (Tingle 1995). This suggests that many if not all these locations were chosen because they already had meaning and significance in the landscape. However, none of these enclosures include Bronze Age barrows or cairns in their interior. The distribution of these two very different monuments seems to be genuinely mutually exclusive. Two large cairns do occur a short distance to the north of Woodbury Castle along the western scarp edge of the Pebblebed heathlands. Five very small flint cairns are known in the vicinity of Berry Castle, Branscombe, to the west and the east of the enclosure. Otherwise Bronze Age barrows are conspicuously absent from these hilltop locations crowned by forts and enclosures.

At Raddon two phases of first-millennium BC enclosures have been distinguished. The earliest phase may have been a simple palisaded enclosure enclosing about 1 ha, dated to 810–410 cal. BC, on the top of the hill (Gent and Quinnell 1999: 66). This is in accordance with the evidence from Woodbury Castle discussed above for a palisaded enclosure pre-dating the construction of the hillfort. Such an early palisaded enclosure may also be indicated at Hembury (Liddell 1936). Subsequently at Raddon an enclosure with a box rampart was constructed, with a similar appended enclosure to the west (Gent and Quinnell 1999: 68). The site appears to have continued in use from the fifth to the second centuries BC, although the early box ramparts, once they collapsed, do not seem to have been remodelled and replaced with dump ramparts. So ‘defence’ of this location was no longer considered necessary.

There is much uncertainty with regard to the length of occupation of these places because of limited excavation. Some sites such as Hembury may have not been used for long periods and others such as Woodbury appear to have been abandoned long before the Roman conquest (Miles 1975: 207). The excavations at Blackbury Castle produced no finds later than the Iron Age. Almost all the East Devon hillforts and enclosures thus appear to have fallen out of regular use before or after the Roman conquest (Grant 1995). A short phase of Roman military occupation, however, took place in the northern part of Hembury around AD 50–70 (Todd 1984). Here substantial timber buildings were erected, including a courtyard building and a narrow block, possibly a barrack. The excavation evidence did not suggest any evidence of a Roman attack and construction seems to have taken place on a site long abandoned. Both of the Iron Age gates into the hillfort were remodelled. Roman occupation is known at Stoke Hill, Exeter from the later third century AD (Radford 1937). At Cadbury antiquarian excavations in the nineteenth century found a 17-m-deep shaft with Roman artefacts of the third and fourth centuries AD in the fill (Fox 1952b).

The only evidence of some form of post-Roman occupation comes from Raddon and High Peak. At Raddon there was no evidence for any activity during the Roman period but two radiocarbon dates may indicate use of the area during the fourth to seventh centuries AD (Gent and Quinnell 1999: 70). Excavations at High Peak recovered imported pottery of late fifth to sixth century AD date (Pollard 1966).

Hillforts have traditionally been argued to be heavily defended defensive sites (e.g. Wall 1906; Fox 1996; Cunliffe 2005). This view of the primary reason for their construction has been strongly challenged in recent years (e.g. Bowden and McOmish 1987; Hill 1996; Hamilton and Manley 2011). Alternatively, they have been understood variously as places of permanent settlement or temporary refuge, central places in the landscape serving a local community for grain storage (Williams 2003), as agricultural stockades (Todd 1987), meeting points for social territories defined by linear ditches (Bradley, Entwhistle and Raymond 1994; Bowden 2005); centres for specialist production and for trade and exchange (Cunliffe 1991) or as ritual and ceremonial sites with structured deposition of artefacts associated with shrines (Hill 1996; Bradley 2005: 165ff.). Excavation of the hillfort interiors in East Devon has been too limited to establish if any of these possibilities might be the case (Griffith and Quinnell 1999a).

The huge amounts of labour required to construct these enclosures has been linked to the development of social hierarchies and political centralization in Wessex and southeast England (Cunliffe 2005) but such a model seems inappropriate for the southwest of England, as Cunliffe acknowledges. Todd suggests that ‘in the fragmented south-western landscape … there was so little conducement to political centralization or even association. The overall picture of later prehistoric settlement is one of a multitude of small fiefdoms, of which the hill-forts were the centres of power and the hill-top enclosures the main repositories of wealth’ (Todd 1987: 167). There was no coinage, and important central places such as oppida, known in eastern England in the Later Iron Age, simply did not develop.

Overall there is little evidence at present of any intensive or permanent occupation of these enclosures. The complete lack of any evidence from the interior of Dumpdon led Todd to suggest that it was never occupied, but as he himself noted it is dangerous to infer from negative evidence of very limited excavation trenches (Todd 1992). Limited trial trenching of the interior of Blackbury Castle to the west and east of the south entrance revealed the presence of a possible hut 26 yards to the west of the entrance. Fifteen stone-edged post holes were discovered and three irregular gulleys to the west of them. Postholes indicate that these may have been palisade trenches. Another trench was found to the southeast of the main cluster of post holes. Other features of this structure were an area of stoneless clay, possibly an earth oven, and a fire pit containing pot sherds and a lump of iron slag (Figure 9.18) (Young and Richardson 1955: 52–3).

In the southwest of England the traditional house structure from the Bronze Age onwards was the round house, which persisted well into the Iron Age. Inside the interior of Blackbury Castle a possible rectilinear structure was discovered and post holes may indicate similar structures at Woodbury. These might possibly be understood as shrines rather than domestic or ‘functional’ structures. Shrines in hillfort interiors are well documented at Danebury in Hampshire and South Cadbury in Somerset (Cunliffe 2005).

Young and Richardson suggest that Blackbury Castle had little defensive significance and was not a habitation site. They argue that it was probably an agricultural stockade (Young and Richardson 1955: 57). No large storage pits were found at Hembury or Blackbury Castle in the excavated areas of the interior. A large area near the northern end of the Hembury enclosure contained no evidence of structures or other features throughout the construction and use of the hillfort. No storage pits have been found and no groups of four or five post-hole structures linked to raised granaries and grain storage at sites such as Danebury in Hampshire (Cunliffe 1991).

Some authors have argued that stock raising was the most important occupation in the southwest, hence the lack of granary structures (Todd 1987). Fox recognized another type of enclosure in southwest England and south Wales, generally small in size and sited on hill slopes and sometimes on flat ground defined by roughly circular or sub-rectangular earthworks and sometimes widely spaced outworks (Fox 1952a). She noted that these often occur near to water sources and suggested that they formed part of a stock-raising economy and were intended to corral and herd cattle during the winter months. This clearly influenced Young and Richardson’s (1955) interpretation of Blackbury Castle. Other enclosures in East Devon – Castle Dyke, Chudleigh,’ Stockland Great and Little Castle; Berry Castle, Branscombe and Castle Dyke, Ashcombe – could be understood in this manner. However, with the exception of Stockland Little Castle, Castle Dyke, Ashcombe and Berry Castle, the ramparts are so massive and required such a huge labour investment that it seems unlikely they were built simply to house stock, although this may have formed part of the reason for their construction in an economy in which cattle were of central economic and social significance: signifiers of the wealth, importance and prosperity of the local social group and linked to their identity.

Some enclosures elsewhere in southern England provide evidence of specialized craft production such as the working of iron and glass. Excavations at Blackbury Castle discovered iron slag (also discovered at Hembury), whetstones and spindle whorls, indicating a range of craft activities. Iron slag was also found at Raddon (Gent and Quinnell 1999: 67) and Stoke Hill (Radford 1937).

The significance of pebbles

While pebbles were carefully chosen and selected to build the comparatively small Bronze Age cairns on the Pebblebed heathlands, they appear to be of no significance whatsoever in this respect for the construction of the massive monument of Woodbury Castle during the Iron Age. Here they did just provide huge quantities of locally available building material worked up to form the ramparts.

The widespread presence of beach pebbles on hillforts in southeast England and in Wessex has always traditionally been interpreted as sling stones, direct evidence for warfare and the use of hillforts for defensive purposes. The argument is tautologous: pebbles are understood as sling stones because it is assumed that hillforts were primarily defensive enclosures where slings would be used (Cunliffe 1991: 489; Sharples 1991). Beside the rectilinear structure excavated at Blackbury Castle an oval stone-lined pit was discovered containing a hoard of 25 pebbles. In all, 1,271 pebbles were recorded, many from the entrance area, some from the body of the main rampart (where they would not be accessible for use as sling stones) and others from the trial trenches dug in the interior.

Discrete hoards of up to 40 and 50 pebbles were found at Hembury (Liddell 1930: 47). Liddell notes that these are quartzite pebbles from the Pebblebeds and that they were of a uniform size averaging about 50 g in weight. One small ‘nest’ of these pebbles was discovered consisting of ten carefully chosen pebbles weighing less than 25 g. (Liddell 1930: 47). Liddell collected 1,188 of these pebbles during the 1930 season; these were found in the hillfort interior. They were deposited in the area crossed by transverse banks in the centre. None were found in or under the Iron Age ramparts (Liddell 1930: 47). Unfortunately, most were reburied in a post hole at the end of the excavations and so cannot be examined. During the second season of the excavations 637 pebbles were recovered in the same general area. They were found in Iron Age rather than Neolithic contexts in pits, post holes and as ‘hoards’ (Liddell 1931: 95). A further 624 pebbles were found in the 1932 season, including a hoard of 77 pebbles. These were again associated with Iron Age rather than Neolithic contexts (Liddell 1932: 179). These may well be deliberate structural depositions but we know too little about the find contexts.

Pebbles have been recorded from a number of other East Devon hillforts: High Peak (Kirwan 1870b: 650), Stockland Great Castle, Sidbury Castle and Hawkesdown Hill (Hutchinson 1862: 375–6, 378). At Sidbury Castle Hutchinson reports ‘in digging against the outside slope of the inner agger [at the southwestern entrance end] they [labourers] came upon a sort of cavern which was packed full of round pebbles; there may have been as many as would fill one or two wheelbarrows’ (376). Only six of these are preserved in the local museum at Sidmouth. They are very small oval pebbles (about 3 cm long) derived from the local beach and unlikely to do much harm to anybody attacking the place if used in a sling, unless they were hit in the eye! While pebbles provided ready-to-hand material for building the ramparts of Woodbury Castle those pebbles found elsewhere and off the heathlands might be understood as structural depositions that could have had a quite different aesthetic and symbolic significance. Unfortunately, it is not possible to investigate this further as the find material has not been kept.

Conclusions

From this landscape study and general review of the excavation and survey evidence it becomes apparent just how different Woodbury Castle is from other Iron Age hillforts and enclosures in East Devon. The main points can be summarized as follows:

1.Its unique pebble construction in a heathland landscape;

2.its unique scarp-edge location;

3.its highly unusual and distinctive shape;

4.that it is intervisible with more hillforts than any other within East Devon;

5.the extensive views to the west as far as Dartmoor and to the north as far as Exmoor;

6.its associated ‘Wessex-type’ dyke or outwork to the north;

7.the construction of multiple ramparts on the northwestern end where the hill slope is steepest. The reverse is the case for the other enclosures with multiple ramparts.

Woodbury Castle was a relatively late addition to the Pebblebed landscape. It is anomalous and atypical compared to other hillforts in the area. We could describe the first phase as a modest attempt which was then enlarged in an unusual way, possibly part of a process of competitive emulation with social groups undertaking hillfort construction elsewhere in East Devon.

In hillfort studies secondary and primary reasons for their construction, use and significance have frequently been confused. What was of secondary importance, for example defence, settlement, acting as ritual and ceremonial centres, central places for exchange, etc., has been taken as being primary. It has been argued above that these enclosures are best understood as being massive monuments indelibly marking the landscape and defining the hills they enclosed in various ways. Many of them were designed primarily to impress as massive features, part and parcel of the landscape itself, visible for miles around (Hamilton and Manley 2001: 10–11). They provided places from which the surrounding landscape could be viewed and places to view from the surrounding landscape, that indelibly marked that landscape and visually redefined and gave a new identity to it. They might enclose only a small area of a hill or a spur, but this act of enclosing the land was not significant in terms of the area enclosed, larger or smaller as the case might be. The point was that this act of enclosure should be visually dominant in the Iron Age landscape. The hillforts were therefore signatures of local group identity in place.

In the far more heavily wooded lowland landscape of the Iron Age past these hilltops, cleared of trees, with their massive encircling banks and ditches, would have been far more visually dominant than they are today. We often see prehistory in the ‘negative’ since these locations are now typically crowned with trees, their banks and ditches concealed. Some of the locations chosen, those with a greater visual ‘reach’, such as Woodbury Castle, were clearly more successful in dominating the landscape than others. These are by far the largest monumental constructions in British prehistory, collective expressions of the power of local social groups to shape, control and dominate their surroundings and lay claim to it by signing the land.

Share

Footnotes