TIMBER BUILDINGS
Over most of lowland England, timber was used for most secular buildings from the end of the Roman period until after the Norman Conquest. Stone was used for the construction of some important churches from the seventh century onwards, and by Ælfred’s time some royal residences were built of stone, but although it might be used for ancillary features, such as porches, it rarely had a structural role in either urban or rural housing. Indeed, the Old English verb used to refer to construction work in early documents is normally timbran.
Timber buildings may be classified into two main categories, according to whether they are ground-level or sunken structures (Rahtz 1976). Ground-level buildings are found in both town and country, but unlike the Early Anglo-Saxon period, when it was common to find sunken workshops in most rural settlements, by the Viking Age the sunken buildings are found almost exclusively in towns, where they are often in equal numbers to the surface buildings. In London, for example, of traces of over 50 buildings of the ninth to the twelfth centuries, 60 per cent were ground level structures, and 40 per cent were sunken.
GROUND-LEVEL BUILDINGS
A basic style of single-storey town house and workshop has now been identified in many Viking Age towns. In London the ground-level buildings are usually found with their gable ends fronting onto the streets, such as at Bow Lane, Botolph Lane and Milk Street. They are generally 4-5m wide, following the width of the tenement plots, with a greater variation in length, ranging from 6.5-10m. Most contained only one room; they had doors in the side and gable walls (Horsman et al. 1988). At Coppergate, York, and Flaxengate, Lincoln, the houses again occupied the street frontage plots. At Lincoln, there was some evidence for internal partitioning with a cross passage joining opposing doors in the middle of the long sides, following the rural fashion.
More substantial timber halls have been excavated at high-status sites such as Cheddar, Faccombe Netherton, Goltho, North Elmham, Portchester and Raunds. The use of internal aisle posts to partition the interior of the larger halls into three aisles has been recognised at many of these sites. A substantial eleventh-century hall, 15 x 7.5m, has been excavated at Waltham Abbey (Essex). The hall has been described as Viking, although there is nothing specifically Scandinavian about it. The position of the aisle posts was marked by clay foundations, but there was no trace of timber wall posts apart from continuous foundation trenches. It has been suggested that the hall had turf walls, but there is little evidence to support this interpretation (Huggins 1976). At Sulgrave, the eleventh-century timber-framed hall was erected on stone footings. It was divided into five bays, with a cobbled porch at one end and a central hearth. The service end, from which the meal was brought, was screened off, whilst at the other end there was a two-storeyed chamber block. A detached timber building near the porch has been interpreted as the kitchen.
BOW-SIDED HALLS
A particular class of rectangular timber halls with bowed walls is often linked with Viking influence. In fact they are part of a long building tradition which has a wide distribution, although they are associated particularly with Viking Age Denmark, and are also found in most of the areas settled by Scandinavians or under their direct influence. They are frequently identified as the houses of the rural aristocracy and the building style may have been popular with a particular class of secular landowners which happened to be dominated by Scandinavians, rather than being specifically Viking.
Various theories have been propounded to explain their bow-sided plan. They are certainly not derived from upturned ships, as was once suggested; nor can they have been designed to provide protection against the wind, although the bowsided structure does give some extra stability.
In England bow-sided halls have been found at Buckden (Huntingdonshire), Catholme, Cheddar, Goltho and Sulgrave, and there are smaller buildings which also have bowed sides at Chester, Durham, Nottingham and Thetford. A bow-sided building at St Neots, with a planked floor over timber joists resting on sill beams has been interpreted as a granary, although it may simply be a medium-sized quality residence.
The ninth-century bow-sided halls at Cheddar and Goltho are both of similar plan and dimensions, 24m long by 6m wide at the centre. In each case the wall posts were set in trenches, with evidence for stave walls at Goltho. At Cheddar the presence of double wall posts with the inner one sloping inwards has led to the suggestion that there was an upper storey whose floor was supported by the inner posts. At Goltho there was evidence that the hall was divided into three rooms, with a raised dais at one end and a cobbled hearth in the centre of the dais. Each hall had three doorways: at Goltho there was one at the east end of the hall, and one either side of the antechamber; at Cheddar there was also a pair of opposed doorways at one end of the hall, but the third appears to have provided access to the upper storey.
SUNKEN BUILDINGS
Sunken buildings are easy to recognise archaeologically as rectangular cuts into the ground surface. In London traces of at least 17 sunken buildings have been excavated dating from the late ninth to the late eleventh centuries, and ranging between 2.8-5.6m wide and 4.2-13.4m long. There is also considerable variation in depth, between 0.41-2.3m, and it is possible to distinguish between two types of sunken buildings.
The first group, such as that excavated at Milk Street, have a floor c.0.5-1m below the ground level, and may be described as sunken-floored buildings. At Milk Street the upcast earth was piled against the outside wall and grassed over, providing a wall up to 1m wide, and much like the stone and turf walls at sites such as Hound Tor and Hutholes in appearance. Another example is known from Lower Bridge Street, Chester, where Building 4 had a sunken floor only 0.84m below the ground surface, and must have been a single-storeyed semi-sunken building. Such sunken-floored buildings are known from the late ninth century onwards.
The second category of sunken building has a floor at c.1-2.5m below the contemporary ground surface, and is more accurately described as a cellared building. These structures are often also distinguished by double linings of horizontal planks affixed to either side of the wall posts, and may have joisted floors. In York this new style of sunken building was introduced across the Coppergate site within a decade of 970, replacing the post-and-wattle structures (Hall 1982; 1984). In London too, their introduction has been dated to the late tenth and early eleventh centuries. These deep cellared buildings are invariably found away from the street frontages in London. No evidence of a hearth has been found in the cellars, which presumably must have been used for storage below a ground-level dwelling and workshop. At Wallingford and Oxford cellared buildings were laid out along the major street frontages by the early eleventh century (Hassall 1986). At Lower Bridge Street, Chester, three almost identical examples have been excavated (Mason 1985). Each had a length-breadth ratio of 5:4 and was some 1.7-1.8m deep, with the lower metre cut into solid bedrock. Post-holes indicated the position of timber walls around the cellars. The floors were planked across at ground level, with access to the cellar down a flight of steps from the outside. At Thetford there were traces of struts sloping inwards that could have supported the upper floor. At York there was no structural evidence for an upper storey but the need to increase the usable space provides the best explanation for the introduction of a cellar, and the comparative evidence from London and Chester suggests that it was likely. A raised first floor could also have contained a clay-lined hearth, conspicuously absent from the basements.
Sunken buildings are almost exclusively an urban phenomenon in the Viking Age, and must have been built in response to particular needs. The sunken-floored type can be seen as developing out of a native tradition, introduced into England by the Anglo-Saxons. They appear to have been mainly used as urban workshops, although there are a few rural examples, such as the ninth-century sunken-floored bread oven at Fladbury (Hereford and Worcester). The cellared buildings appear to be a response to developments within urban communities arising in the second half of the tenth centu
ry, probably the need to store goods in transit, or stock-in-trade. The cellars would have provided cool and secure repositories for foodstuffs and other supplies, and appear to be associated with the tenth-century revival of trade and growth of trading towns (see chapter 6). Cellared buildings also appear in Danish towns such as Århus in the tenth century, but it is not clear if they represent a Scandinavian introduction to England or are simply part of a general north-west European development.
FOUNDATIONS
Timber-framing, by which buildings are constructed around a timber frame held together by carpentered joints, was not widely used in England until after the Norman Conquest. The structural stability of Viking Age buildings therefore generally depended upon their foundation methods. The most common technique was to use earth-fast foundations. In some cases, as for the peasant houses at North Elmham and Barton Blount, the wall posts might still be set directly in the ground. In towns too, such as Lincoln and York, this was common practice, with posts sometimes driven into the ground as stakes, but more usually placed in post-holes packed with stones. Such posts often do not appear to have been set in pairs, but may sometimes have just acted as stiffeners for substantial cob or clay walls.
Increasingly, however, a continuous trench was dug along the wall lines, and the vertical posts or staves set within it, following a technique developed at Middle Saxon sites such as Chalton and Maxey. During the Viking Age this ‘postin-trench’ technique was utilised for major buildings at Cheddar, Goltho, Middle Harling (Norfolk), Portchester and North Elmham. At Catholme side walls were set in trenches, whilst the gable ends were supported by individual posts. Alternatively, individual post-pits could be excavated to hold massive vertical timbers, each up to 0.6m across, for substantial structures such as the West Hall at Cheddar.
During the Viking Age a new foundation technique of using a ‘sill beam’ was introduced at urban sites such as Chester, York and London, and in rural settings such as Buckden, North Elmham, Northolt (Middlesex), Portchester, St Neots, Sulgrave and Waltham Abbey. A sill beam is a horizontal beam which may be set in a foundation trench, or placed directly on the ground surface. The wall posts rest upon it, and may be held in position by a raised timber lip or sill, or they may be set into the beam in rectangular sockets. At Coppergate, York, massive oak beams, up to 7m long, with raised sills, were set in cuts up to 1.5m deep as the foundations of the planked sunken buildings.
Another new technique, first used in the mid-tenth or early eleventh century in London, was to employ a foundation bed, such as a rubble platform or stone pads, to support the structural timbers. In London, the development of mortar or stone sills represents an attempt at prolonging the life of timber buildings, and thereby making more effective use of the woodland resource (Milne 1992). At Flaxengate, Lincoln, one building rested upon a single course of stone footings which may have supported a raised plank floor to lift perishable materials such as grain off the damp ground.
WALLS
The most common walling material at the start of the Viking Age was probably wattlework, following a long Anglo-Saxon tradition. In York in the first half of the tenth century post-and-wattle was the standard method of building construction (plate 6). At neither York nor Lincoln, however, has clay daub been found on the walls, and the small quantities present suggest it may only have been lining clay ovens. Clay and straw may also have been used to make cob walls. It is likely that fur or textile drapes may have been used to cut out draughts. In York, screens of woven willow twigs may have been used to provide wall insulation, or may have fallen from the loft.
Wattlework was inappropriate, however, as a walling material for the new class of sunken buildings, as it would have collapsed under the pressure of the surrounding earth. At both London and York wall cladding for the sunken buildings was provided by post-and-plank construction. At York the wall posts, carefully squared and regularly spaced at short intervals, supported layers of horizontal oak planks which were laid edge on edge (plate 7). There was no evidence for the use of pegs, nails or joints in the lower 1.75m of the walls; apparently the planks were held in place against the wall posts by the sill beam and the weight of surrounding earth. These posts appear to have been paired across the building and so were presumably held upright by tie-beams spanning the width of the building at the top of the walls. There are also instances of an inner skin of horizontal planks being fastened to the wall posts to provide a cavity wall.
Staves, or vertically set planks, first appear in London in the eleventh century, but have now been found earlier at other sites, such as Goltho, where half sections of trees, about 0.45m wide, were set in a trench with their curving face outwards. The method is well known from Norwegian stave churches, and was adopted at St Andrew’s, Greensted-juxta-Ongar (Essex), although there is no evidence that its adoption in England was due to Scandinavian influence. At Goltho the two structural traditions of stave and post construction persisted side by side until after the Norman Conquest.
Any available wood seems to have been used for wall construction, with a clear preponderance of hazel for wattlework and oak for wall posts and planks, with willow, alder and birch in descending order of importance. Where joints needed to be fixed then wooden pegs appear to have been used more commonly than iron nails. The standard mortise and tenon joint was unknown before the Norman Conquest. In contrast to the use of elaborate joinery in late medieval structures it is clear that the builders of Viking Age houses and workshops invested a minimum of labour and skill in these structures (Milne 1992).
ROOFS, FLOORS AND INTERNAL FITTINGS
For the smaller buildings roofs generally rested directly upon the walls, although for larger halls such as those at Cheddar and North Elmham their weight appears to have been borne by internal posts. Straw and hay may have been used for thatching, or in some areas turves may have been used. There is little evidence for wooden shingles although most of the hogback stones apparently have shingled roofs.
Floors were most commonly of beaten earth or clay, but sand, gravel, and mortar are also known. At Lincoln there was evidence of rushes and other grasses, and at Durham sedges, rushes, heather, bracken, meadowsweet and crowfoot were laid over the sand floor as a sweet-smelling covering which could be replaced before it became unpleasant. At Thetford the floor of one of the sunken buildings had been mortared; another was surfaced with closely packed cobbles. In York several buildings had plank floors resting on joints. At Coppergate the floor boards had been carefully cut to fit flush around the internal wall posts so as to lap against the wall.
Ground-level buildings generally had at least two entrances, sometimes protected by a wooden porch, as at Portchester. The basements of sunken buildings could have been entered by an internal ladder, but many were entered directly from outside. In Ipswich, London, and at the Clarendon Hotel site, Oxford, earthen steps have been identified in shallow extensions. At Thetford the cellar of Building J was approached down a 5m long ramp revetted with posts. At Coppergate the sunken buildings were entered from the rear along sunken passageways revetted with stone, but may also have had a second entrance at the street front. At Chester each of the sunken buildings had a ramped extension some 2.5m long cut into the solid rock. These would have been dangerously steep unless provided with a flight of wooden steps, and were probably covered by a porch to keep out the rain.
All buildings must have been provided with wooden doors, although they are rarely found. In London, one was found lying in demolition debris. It comprised four vertical oak boards secured by diagonal battens on inner face and fastened together with iron nails. The remains of iron hinges were also found. Various Viking Age lock mechanisms have been excavated, but these are generally supposed to have originated from chests rather than doors.
There is evidence for window glass from two London sunken buildings, but glass is otherwise rare with no traces even on aristocratic sites such as Goltho, although it may normally have been robbed as a precious commodity. From Copperg
ate the remains of a possible window shutter have been recovered. Internal lighting would have been provided by the hearth fire, and by small oil lamps. In the tenth and early eleventh centuries there are stone and pottery oil lamps with bases splayed to sit on the floor or in a niche on the wall.
All buildings normally had a central hearth, usually consisting of baked clay over a bed of stone. Pottery cooking vessels would be set in the embers. In York a number of industrial hearths, up to 1.2 x 1.8m, were found positioned in the centre of the floor, and built one upon the other as the floor level rose. They had a clay base, of which only a small part was normally burnt, surrounded by a kerb of limestone blocks, reused Roman tile, or wooden beams. Specialist kitchen buildings such as that at Goltho were often provided with ovens built of baked clay and wattle on a stone rubble base; in one London example the clay walls were supported by 33 angled stakes. At Portchester one even had a clay dome set on a base of reused Roman tile and lumps of limestone set in clay.
Many buildings presumably had internal timber fittings such as wall benches and beds, although remains are more common in stone buildings, where they survive incorporated in the stone structures. At Coppergate one building had rows of stakes running parallel to both side walls and 0.6m from the walls for two thirds of the excavated length of the building; these probably represent wattlework revetting for earth-packed wall benches. Substantial pits were dug inside some of the York buildings; one assumes that these were used for storage, and were originally covered by planks. Storage of valuables would have been undertaken in sturdy wooden chests, with perishable goods stored in pottery vessels. At Pudding Lane, London, two abandoned buildings had smashed fragments of large spouted storage jars on the floor.
Viking Age England Page 9
