Anglo-Saxon Assembly Places

It is well established that by the end of the Anglo-Saxon period, and perhaps much earlier, most of England was divided for administrative purposes into districts known as hundreds or wapentakes.Footnote ⁶ The freemen of each hundred or wapentake met every four weeks at a designated site, in order (among other things) to resolve disagreements and discuss legal matters. Although much research has focused on laws and the settlement of disputes, we still have a great deal to learn about the physical attributes of the sites at which these assemblies took place in terms of their landscape context, accessibility, prominence and suitability for large gatherings and public discussion – all the things that determined the choice of meeting location.Footnote ⁷ In part, this is because many meeting places remain unidentified, but a good number can still be located. For one thing, the hundreds or wapentakes were often named after their place of assembly and in many cases the place names still exist. In other instances, contemporary or later medieval documents make reference to the meeting place, and in yet others surviving place names containing elements referring to assembly provide clues as to where meetings may have occurred. Obviously, the various forms of data produce different levels of accuracy and reliability in their identifications and care must be taken, but it is nevertheless possible to locate a sizeable number of assembly sites, and to visit them on the ground.

An important part of the work of the ‘Landscapes of Governance’ project has been to extend the scope and coverage of earlier surveys to visit as many of these sites as possible and to record their various attributes using a standardised pro forma.Footnote ⁸ Systematic field visits to all identifiable meeting places within a large part of England have revealed some interesting but perhaps not unexpected results. Hundred meeting places fall relatively easily into a limited number of types, if defined and categorised by their landscape location. A set of analytical criteria has been set out elsewhere, with example sites,Footnote ⁹ and so only a summary of the categorisation is given here. In essence, meeting places can be classified by their relationship to surrounding infrastructure, since accessibility was crucial; but they also exhibit other important variables, such as monumentality and scale.

Two types of location predominate. Many hundred moots are situated in upland areas, with impressive views over surrounding countryside (although they are not necessarily located on dominating features, nor do they always have commanding views). They tend to be close to major routes, but not directly on them, and seem sometimes to have been distinguished by markers that no longer survive. Bisley hundred in Gloucestershire, for instance, may well have met at a place called Wittantree, recorded as Wittingtree in 1841, perhaps from Old English (OE) witena-trēow, ‘tree of the councillors’.Footnote ¹⁰ The meeting place of Longtree hundred in the same county was presumably also marked by a tree.Footnote ¹¹ In an open expanse of upland, an isolated and prominent tree might serve as a suitable marker for hundred gatherings. Many other meeting places lie directly on or adjacent to major routes, often at fords or road junctions. An example of this type is Holiforde (one of the Domesday hundreds that made up the later medieval hundred of Kiftsgate in Gloucestershire), a site probably marked by the modern hamlet of Ford, in Temple Guiting, and taking its name from OE hol, ‘hollow’, and ford, ‘ford’.Footnote ¹² Sometimes these junctions are indicated by markers, examples of which include Tibblestone (Domesday Tetboldestane), also in Gloucestershire, a hundred which presumably met at the Tibble Stone, the name meaning ‘Þēodbald's stone’ (OE stān),Footnote ¹³ and Normancross in Huntingdonshire, probably named from a standing cross.Footnote ¹⁴

There are variants of these two basic types, depending on relative proximity to major routes and the physical and visual characteristics of the sites themselves. The probable meeting place of Fernecumbe hundred (Warwickshire), for example, lies near Barley Leys Farm in Haslor (national grid reference sp 131557) at the top of a scarp approximately 800m south of the Roman road from Stratford-upon-Avon to Droitwich (Mgy 56b).Footnote ¹⁵ It is neither directly adjacent to a road, nor is it in the middle of an upland expanse. Several other meeting places lie on major routes at the edge of important royal or ecclesiastical centres – Dudstone, for example, on the Roman road (Mgy 180) heading north from the burh of Gloucester to the royal manor at Kingsholm, and Deerhurst, which met at Olanige, a small island formed by the River Severn and Naight BrookFootnote ¹⁶ – and these form a group that might be termed ‘extra potentiary’, their location suggesting a conscious decision to summon the people of the hundred to the potentate's gates. There are clear practical reasons why these various site-types may have been chosen as meeting places and by and large they are characterised most clearly by their spatial relationship with infrastructure and sociopolitically important settlements, and by the need to mark the spot from its surroundings. All the same, there is much more work to be done on classifying meeting places and understanding their nomenclature.

The Characteristics of Hanging Promontories

One group that we might term ‘hanging promontory’ sites can be distinguished from the rest. These meeting places are highly distinctive and share a number of common physical attributes. Most striking is their topography, setting them out from the surroundings not as mundane markers of assembly sites but as dominating platforms. Taking the form of domed hillocks of up to 100m diameter they are located on spurs of land protruding below a crest of higher ground. The high ground is often the location of a junction of several parish boundaries and the site is usually marked by a well-worn holloway descending further downslope beside the promontory.

Two of the most archetypal of these hanging promontories are in Gloucestershire, at Kiftsgate Court, on the edge of Mickleton parish (sp 171431), and Hundred Field, between Botloe's Green, Botloe's Farm and the fields of Little and Great Botloe's Piece on the border of Newent and Pauntley parishes (so 721285) (Fig 1). Kiftsgate Court is much altered by a nineteenth-century manor and Edwardian formal gardens, but Hundred Field exists as an open field of pasture. Both exhibit the aforementioned characteristics. Kiftsgate Court lies on the steep western scarp of the Cotswold Hills at c 160m above sea level. It takes the form of a natural platform c 90m wide projecting from the scarp, from which it is separated by a narrow coombe on the south side, through which runs a clearly marked holloway (Fig 2). Hundred Field, Botloe's Green, has a noticeably domed appearance, occupying a small plateau at the 75m contour (Fig 3). A network of footpaths, tracks and roads converges on Botloe's Green, many of them holloways, including a well-worn route leading south–north down the eastern side of the field.

Fig 1

Archetypal hanging promontory sites: (a) Kiftsgate Court; (b) Hundred Field, Botloe's Green (both in Gloucestershire). Drawing: Stuart Brookes

Fig 2

Kiftsgate Court (high ground to the right) and the views over Gloucestershire. Photograph: Stuart Brookes

Fig 3

Hundred Field, Botloe's Green: the mound-like shape of the field and the holloway on the right are clearly visible. Photograph: Stuart Brookes

Significantly, the second element in Botloe seems to be OE hlāw, ‘mound’, probably compounded with a personal name, Bōta.Footnote ¹⁷ The element hlāw in place names usually refers to entirely artificial mounds, but this is not always the case. Most instances where hlāw refers to a natural mound apparently occur north of the Humber–Mersey line, but Gelling makes the point that a scarped natural hill might in some cases have been thought of as artificial.Footnote ¹⁸ Human attempts to accentuate the shape of the promontory may, therefore, have made the use of the term hlāw appropriate, but it is also possible that another artificial mound formerly existed in the vicinity. Both Kiftsgate and Hundred Field possess commanding 180-degree views over large tracts of the landscape – the promontory of Kiftsgate overlooks much of southern Worcestershire, while Hundred Field is situated on rolling landscape falling away on all sides to give good views, particularly of Herefordshire to the north.

This distinctive combination of features – the physical location comprising mound-like eminence, holloway and extensive views, as well as the place-name evidence discussed – finds parallels elsewhere within the Hwiccan territory, notably at Motslow Hill, Stoneleigh, Warwickshire (sp 329723), the meeting place of Stanleie hundred,Footnote ¹⁹ and Ravenswell Farm, Gloucestershire (sp 038263), the probable meeting place of Wacrescumbe hundred. However, they also exist outside this area, at, for example, Moot Hill Piece, in Gillingham hundred, in Dorset (st 761305),Footnote ²⁰ Winterhead in Winterstoke hundred, Somerset (st 441568), and Silverden in Silverden hundred, Kent (tq 788284).Footnote ²¹

Other hanging promontories in Hwiccan territory, at Ravenswell Farm in Gloucestershire and Motslow Hill in Warwickshire, are less physically impressive but are evidenced earlier. Ravenswell Farm lies on a promontory above the eastern flank of a well-marked coombe known as waclescumbe in the authentic mid-tenth-century bounds of Withington (S610).Footnote ²² Motslow Hill is first mentioned in 1260 (la Mustowe), the name apparently deriving from OE (ge)mōt-stōw, ‘place of assembly’, but perhaps interchanging or confused with (ge)mōt-hlāw. It is a site much disturbed by post-medieval quarrying, but an artificial mound is still visible as an earthwork lying at the neck of a narrow promontory, accessed from Stoneleigh by a holloway. According to the antiquary William Dugdale, this was reputedly the location of the manor court in the mid-seventeenth century, when the tenants of Stoneleigh, called Sokemen, were compelled every three weeks to do their suit at the King's Court held for the manor, upon a place ‘then and yet called Motslow-hill’.Footnote ²³

Moot Hill Piece lies in the heart of West Saxon, rather than Hwiccan, territory, but in all other ways it is identical to the aforementioned sites. Identified on a tithe map of 1837,Footnote ²⁴ Moot Hill Piece lies adjacent to the county boundary between Dorset and Somerset, north west of Bourton. Close to this location is a promontory lying just above the 160m contour which remains well defined on its western flank by a coombe (Fig 4a). The position on the south-facing slope affords expansive views over much of Dorset to the south and may have done so in the early medieval period, provided it was already open ground on the edge of Penselwood (Fig 4b).

Fig 4

(a) Plan of the hanging promontory site by Moot Hill adjacent to the shire boundary of Dorset and Somerset; (b) the extensive views south from the meeting place over northern Dorset. Plan and photograph: Stuart Brookes

A further example is the site of Winterhead Hill in Somerset. According to Anderson this is likely to be the meeting place of Winterstoke hundred,Footnote ²⁵ a large territory stretching from Weston-super-Mare on the Severn coastline and including a detached portion at East Harptree to the east. Winterhead Hill lies on the southern edge of the hundred on its borders with Axbridge (or Cheddar) hundred, and is a large mound-like eminence separated from the village of Shipham to the north east by a deep defile. The promontory lies on the north-facing slope of the Mendip Hills at the point where a natural break in the scarp allows for a descent to Axbridge and Cheddar to the south.Footnote ²⁶ These two important late Anglo-Saxon sites – a named burh and a royal palace respectively – are located within a discrete administrative unit which gives the appearance of having been carved out of the larger territory of Winterstoke. The meeting place of this hundred seems to be remembered in the field name Spelthorne (from OE spell-þorn, ‘speech thorn-tree’),Footnote ²⁷ located on a spur of land projecting into the Somerset Levels between Axbridge and Cheddar.

Hanging Promontories and Administrative Territories

Beyond the physical similarities shared by these sites, they also occupy similar positions within their administrative territories. Two aspects are particularly important. First, these sites often have a complicated association with the hundred in which they are located, and although they may sometimes have served as the hundred meeting place, several of them coexist with sites that provide strong alternatives for the main hundred moot. Second, a number of hanging promontory sites are situated in close proximity to major shire, ecclesiastical and, probably, national boundaries. Both factors suggest that they may have originated as infra- or (perhaps more probably) supra-hundredal sites of assembly.

Kiftsgate and Botloe are again prime examples (Fig 5). The former is situated just within Gloucestershire, about 1km from the boundary with Warwickshire; Botloe is in the north west of Gloucestershire, some 1.5km from the border with Herefordshire. Both sites lay within the diocese of Worcester and probably within the kingdom of the Hwicce, but their liminal locations on shire boundaries might indicate that they were occupying thresholds between different political groups within the territory. Similar liminality is suggested by the location of Silverden in Kent. Lying beside a Roman road (Mgy 13), 1.5km north of where it crosses into Sussex, Silverden also lies near the western edge of the hundred and the lathe of Limen, to the west of which lay lands not yet formally constituted as hundreds by Domesday.Footnote ²⁸

Fig 5

Distribution map of hanging promontories in south-western England, showing the location of sites mentioned in the text: (1) Kiftsgate Court; (2) Hundred Field, Botloe's Green; (3) Mutloes; (4) Ravenswell Farm; (5) Motslow Hill; (6) Moot Hill Piece; (7) Winterhead Hill; (8) Court Farm, Damerham; (9) Silverden. Drawing: Stuart Brookes

In the two Hwiccan cases, circumstantial evidence suggests that the hosting of hundredal meetings was not their only function. For example, Botloe and the neighbouring hundred of Langebrige may once have constituted a single unit,Footnote ²⁹ so that within this earlier hundred district Botloe was perhaps not the only meeting place. Similarly, there are two clear candidates for the hundred meeting place of Kiftsgate. The name itself (Cheftesihat 1086, Keftesgate (hundredum) 1183) seems to derive from OE *c $$\[--><$>\={y}<$><!--$$ ft, cognate with Old High German cumft, ‘coming’, and meaning ‘meeting, conference’, and geat, ‘gate’.Footnote ³⁰ The name survives in Weston Subedge, where an ancient stone known as Kiftsgate Stone stands on a ridgeway on top of the Cotswold escarpment. This is Kyftesgate in the fourteenth century and, assuming Smith's identification to be correct, was the site of an inquisition at that time and of ‘þe Kynges court’ in the sixteenth century.Footnote ³¹ Its credentials as a meeting place would seem to be impeccable, and its location at a roadside marker is reminiscent of hundredal foci such as Tibblestone or Normancross. Yet Kiftsgate Court in Mickleton parish, which is the hanging promontory site described above, also contains the name of the hundred and is in close proximity to Spelsbury Barn (Spellborough 1698). This latter is probably from OE spell-beorg, ‘speech barrow’, referring therefore to another place of assembly.Footnote ³² It is possible that the two Kiftsgates reflect the existence of a meeting place used for the hundred itself, and another used for other assemblies, possibly involving several hundreds. It is important to note that in the fourteenth century Kiftsgate was the name of an amalgamated group of hundreds comprising those of ‘Celfledetorn’, Kiftsgate and ‘Witley’.Footnote ³³ Perhaps significantly, Kiftsgate Court lies just 1.1km south of Upper Clopton, the site of Langeberge, a royal manor to a grouping of three hundreds.Footnote ³⁴ A similar royal association is noted at Botloe's Green, which lies within the large Domesday royal manor of Newent Pauntley.Footnote ³⁵ This might support the idea that these places played some kind of supra-hundredal function.Footnote ³⁶

A similarly complex ascription surrounds the site of Moot Hill Piece, again hinting at a hierarchy of assembly places. Gillingham hundred takes its name from Gillingham (Gelingeham hundret c 1086 Geld Roll, hundr’ de Gellingeham 1130), evidently an important Anglo-Saxon centre, and probably a minster site with several dependent chapels.Footnote ³⁷ Mills notes in Gillingham a place called Hurstyngesmersshe 1501, from hūsting, ‘council held by the king’.Footnote ³⁸ This is likely to be related to the meeting place of the hundred, but the location cannot be more precisely identified. Yet further contenders for meeting places exist in the hundred. A hollow tree called Judge Wyndham's Oak, in a field adjoining Silton churchyard, was traditionally identified as the location of the manorial courtFootnote ³⁹ and the site of Motcombe (Motcumbe 1244), ‘moot-coombe’, lies 2km north of Shaftesbury.Footnote ⁴⁰ These alternative locations exhibit the more common repertoire of meeting place features, and in this way Moot Hill Piece is highly distinctive. Bourton parish, in which Moot Hill Piece is located, is a detached portion of the hundred.Footnote ⁴¹ Moreover, unlike the other sites in Gillingham hundred, Moot Hill Piece lies in very close proximity to the shire boundary, as well as – perhaps crucially – to Penselwood in Somerset, which lies just 1km to the north west. This is named as the location of a battlefield in 1016 and is therefore potentially close to a major military mustering place.Footnote ⁴² It is also close to one of the candidates traditionally put forward for the site of Egbert's Stone, where King Alfred is said to have met his forces en route to Edington in 878.Footnote ⁴³ There is too much uncertainty to place any great weight on this putative and debated identification, but it would raise two pertinent points. First, it would place use of sites of this kind within the second half of the ninth century; second, the name Egbert's Stone might represent a further, ideological monumentalising of the landscape through reference to an earlier West Saxon leader.Footnote ⁴⁴

Fewer complications surround the association of Ravenswell Farm and Motslow Hill with hundredal meeting places; indeed in these cases the relationship with the shire geography is less convincing, though nonetheless interesting. Ravenswell Farm is 2.8km north of the boundary between Winchcombeshire and Gloucestershire proper, while Motslow Hill lies near the exact centre of Warwickshire. The impression that hanging promontories were part of some form of shire organisation pre-dating the creation of hundreds is tantalising, but unfortunately difficult to prove, as few chronological markers allow us to pinpoint their use with any precision. The spatial relationship between Ravenswell Farm and Winchcombeshire, an administrative territory apparently defunct by the mid-eleventh century,Footnote ⁴⁵ is interesting, but not necessarily significant given the site's location near the centre of the Domesday hundred of Wacrescumbe. Similarly, to see Winterhead as contemporary with the hundredal boundary, presumably established at the same time as the burh of Axbridge in the late ninth/early tenth century, requires too much special pleading. Nevertheless, the location of some of the most spectacular examples on the boundaries of the early medieval diocese of Worcester, and perhaps therefore on the edges of the Hwiccan territory, may provide a chronological context, and this should be taken alongside the juxtaposition of Moot Hill Piece, Penselwood and the meeting point of Somerset, Wiltshire and Dorset.

Discussion: A Landscape Template for Assembly?

Hanging promontories represent only a small minority of the types of meeting place analysed by the ‘Landscapes of Governance’ team, but it is worth considering the commonalities of these sites in terms of their formal arrangement. All the promontories have a distinctive mound-like appearance, perhaps in some cases accentuated by artificial modification, and commemorated in place names containing OE hlāw. The identification of mounds as meeting places/cult sites is common to both Scandinavia and Ireland,Footnote ⁴⁶ and has been adopted also – by analogy – in British archaeology. The meeting place of Secklow hundred in Buckinghamshire has been identified with a large mound, shown through excavation to have been purpose-built at some time between the fourth and thirteenth centuries, and many more mounds of both early medieval and prehistoric origins can be correlated with hundred meeting places.Footnote ⁴⁷ The purpose of mounds may have been functional, to serve as ‘soapboxes’ for public pronouncementsFootnote ⁴⁸ or as the seat of high-status representatives at the assembly – an arrangement attested in procedures adopted by the assemblies at Tynewald Hill in the Isle of Man.Footnote ⁴⁹

In all the cases discussed here, the mound is easily accessible only from one clearly defined approach along the neck of the promontory. In the cases where these exist, the neck of land appears to have been accessed from lower ground through holloways along the flank of the site. This form of channelled – perhaps restricted – approach is paralleled by well-known assembly sites elsewhere. The approach to the ‘national’ assembly of Iceland, the althing at Thingvellir, is through an impressive gorge known as the Almannagjá (‘People's Ravine’), forming part of the rift valley in which the meeting place is set. Although assembly-related activities took place over a wide area, legal proceedings were centred on the distinctive Law Rock (Lögberg), located at the head of the Almannagjá. Similar processional routes leading to mounds are known from IrelandFootnote ⁵⁰ and Sweden,Footnote ⁵¹ but a particularly striking analogue to hanging promontories is the meeting place of Tingaholm, the site of Shetland's local parliament until the late sixteenth century. Before water levels were altered in Tingwall Loch in the 1850s, Tingaholm was an islet entirely surrounded by water and accessed by a stone causeway.Footnote ⁵² Officials are thought to have sat around a rough stone table on the holm, while delegates gathered on the slope below the church. Similarly, during the still-practised assemblies on the meeting mound of Tynwald Hill, Isle of Man, access to the hill is restricted to the officials of the court, with public members occupying flat ground beyond. The arrangement of hanging promontories suggests that Anglo-Saxon assemblies at these sites may have taken place in a similar way.

In all cases there is a correlation with specific fields of view. By virtue of their location downslope of higher ground, views are typically channelled to create panoramas in a specific direction, generally across the shire in which they are sited. The location of the promontory below higher ground is reminiscent of Adam of Bremen's description in the late eleventh century of the thing, or assembly mound, of Gamla Uppsala, Sweden: ‘the shrine stands on level ground with mountains all about it like a theatre’.Footnote ⁵³ It could be argued that the extent and character of the views from individual sites may only reflect natural determinants of relief and elevation. To some extent this is true, yet the promontories to which place names indicative of assembly have become attached tend not to occupy summits. In other words, those choosing the venue for assembly may have eschewed immediately adjacent higher ground in preference for raised mounds lower down. In that sense, it is the topography of the sites and the careful defining of their views that is significant, rather than the dominance of their locations in absolute terms.Footnote ⁵⁴ In instances where the shire boundary probably coincided with the higher ground nearby, it may be significant that the location of the assembly site permitted views into one shire only.