11 Plan of Viking Age York (after Hall 1988)
Within the walled area, however, there is little evidence for Viking Age structures inside the legionary fortress, and in most areas thirteenth-century levels immediately overlie Late Roman ones, although this could be because conditions here were not right for the preservation of Viking Age organic deposits, or because they were swept away c.1200 prior to the major medieval building programme. Scandinavian chance finds have been found throughout the area, including manufacturing debris. There is no evidence for the imposition of a regular street layout during the Viking Age, although the Roman defences continued to influence the topography. It has been suggested that the Viking royal palace may have been sited near the south-east gate of the legionary fortress, in King’s Square, although no archaeological traces have been excavated. Similarly, no trace was found of the Viking Age Cathedral underneath York Minster although a graveyard with Anglo-Scandinavian grave markers suggests that the church was nearby, possibly to the south-west. The former legionary barrack blocks housed Viking Age activity, including antler and bone working and ferrous and non-ferrous metalworking, with the walls of the former barracks used to demarcate the new tenements. This activity may have been under the control of the church, or may simply have been taking place adjacent to it. In the area between the Roman fortress and Marygate, where there was a defended Roman enclosure, the name Earlsburgh suggests that the later pre-conquest Anglo-Scandinavian earls must have had their residence nearby. Between 1030-55 Earl Siward built or rebuilt a church or private chapel here, which he dedicated to St Olaf.
South-west of the river Ouse the area of the former Roman colonia, or civilian settlement, was also occupied during the Viking Age. There is no archaeological evidence for the defences south-west of the Ouse but it is assumed that they were on the line of the surviving medieval walls. There appears to have been an important Anglo-Saxon ecclesiastical complex in the Bishophill area, and the street names point to continued Viking Age activity over most of the walled area. A new bridge was built across the Ouse, to the east of the Roman one, and Micklegate (literally ‘Great Street’) diverges from its Roman course as it heads for the new crossing. Timber buildings and rubbish pits have been found along Micklegate and on Skeldergate. This new Viking thoroughfare across the Ouse focussed on the Coppergate, Ousegate and Pavement area where the most striking evidence for Viking Age York has been found within the new enclosure. In this area the city has risen on its own refuse, and rich evidence for its people and their lifestyle has been preserved in several metres of Viking Age deposits.
Evidence for intensive occupation, including timber buildings housing leather workers, has been excavated within the basement of Lloyd’s Bank, Pavement, and cellared buildings, once thought to be plank-lined tanning pits, have been excavated at High Ousegate, but the largest sample of Viking York has been excavated at Coppergate, on sharply sloping ground leading down to the river Foss.
Post-Roman activity, including glass-working, commences in the mid-ninth century, but it is not clear whether it starts before the Viking takeover as a result of people seeking the protection of the walled town, or as a consequence of the Viking takeover in 866 or settlement in 876. If the walled town did not include this area until after its Viking refurbishment then the latter explanation may be more likely. The new timber Ouse bridge was presumably erected in the late ninth or early tenth century, and at some time between the late ninth century and c.930-5 the Coppergate street was established, with a series of long narrow tenements defined by wattle fence alignments running back from the street down towards the Foss. These property boundaries remained in force throughout the Viking Age and influenced all subsequent developments over the next thousand years. The plots were of equal width, perhaps implying that they stemmed from a deliberate act of town planning designed to stimulate the development of a Viking Age industrial estate.
On the four tenements within the Coppergate excavations, post-and-wattle buildings were erected on the street frontage with their gable ends facing the street (plate 5). Their backyard areas were riddled with pits; those lined with barrels were probably wells; those with wattle lining may have been used for storage; others served as latrines and cesspits. Within the buildings there is evidence that iron working and other crafts were carried out on a commercial scale. Two of the tenements were used for metalworking, including copper and lead alloys, silver and gold, as well as iron. In fact, the tremendous diversity of craft activities suggests that the buildings were rented out to craftsmen, rather than each being permanently occupied by one individual.
Around 975 the four buildings were simultaneously demolished and replaced by a series of plank-built sunken buildings (plate 7). On three tenements the buildings were arranged in two ranks, suggesting that the reorganisation was perhaps prompted by the increased intensity of occupation and the need to store manufactured and traded goods. The buildings were occupied by jewellers and woodworkers, but industrial metalworking ceased and may have been forced to move to the fringes of the densely settled area. Finally, early in the eleventh century, a large warehouse or boat shed was erected at the rear of the site, closer to the river.
The general picture which has emerged from the Coppergate excavations is of a fairly squalid urban environment, which contrasts with that of Roman York. The town has been described as a large compost heap, composed of rotting wooden buildings with earth floors covered by decaying vegetation, surrounded by streets and yards filled by pits and middens with organic waste. Organic refuse was being dumped at a greater rate than it was being cleared away; during the tenth century the ground level rose by an average of 25mm per year. Nevertheless, whilst no doubt the exterior of the properties was foul and disgusting, their insides may have been tolerably cleanly maintained (Hall et al. 1983). Excavations of three contemporary late tenth-century tenements from Saddler Street, Durham, have suggested a relatively clean environment, with dumps of sand used to seal middens and pits.
The deposits from Durham are very similar to those from York, although there is nothing particularly Scandinavian about them. Indeed, one must also question whether there is anything distinctively Viking about York, apart from a new taste for Scandinavian style ornament. Certainly, there is no evidence to show that the inhabitants of Coppergate were Scandinavian in origin. As has been shown, Scandinavian traders were not responsible for establishing York and other towns as major trading sites. On the contrary, the international contacts of existing sites were disrupted, and only recovered after an interval. Where there were no existing trading sites, such as on the Isle of Man and in Wales and Brittany, Viking activity did not lead to their formation.
MERCIAN BURHS
Although the first Anglo-Saxon towns may have originated as trading sites, a much larger group of towns were established as defended forts, or burhs, probably as a direct response to the Viking threat. The earliest Engish examples were established in Mercia c.780-90 by King Offa, possibly following Carolingian prototypes (Haslam 1987). The Mercian burhs should be seen as a systematic defence against Viking seaborne attack. All were associated with defensive bridges and were placed on main rivers throughout Mercia so that the burh and bridge blocked access upstream to warships. They also functioned as civil and ecclesiastical administrative centres and became important markets, although the markets may have grown up outside the walls without deliberate planning. There appear to have been two classes of site. Some were established on existing fortified Roman sites, where the walls and bridge probably survived; others developed from fortified manorial centres. The Vikings often chose them as military bases in the late ninth century, but the sites were already fortified by this date. A single spinal street normally links the burh, bridge and market areas, although the Mercian burhs may have lacked the regular planned street systems which have been observed in the burhs of Wessex. Initially, the interiors may have remained fairly open, with intensive occupation only dating from the later tenth century. For t
heir defence they would have relied upon a peasant militia derived from the countryside rather than from within the town (Rahtz 1977).
Traces of early defences at Hereford, and perhaps Tamworth, demonstrate the role of Mercia in the evolution of the Anglo-Saxon town. The town of Hereford commands a strategic crossing point on the river Wye; its name means literally ‘the ford of the army’. The town lies on the north bank of the river, where a gravel and clay rampart of the mid-ninth century appears to have enclosed a rectangular area of 13.6ha (33 acres), incorporating the minster church in its south-east quarter. In the early tenth century, possibly at the instigation of Queen Æthelflaed in 914, the walls were extended eastwards to enclose suburban growth across a 21ha (50 acre) area, and improved with timber revetments at front and rear. Finally, c.930, the front timber facing was replaced by a stone wall, 2m thick and 2m high, with a slighter wall at the rear, and a fighting platform 4m wide.
The Anglo-Saxon Chronicle records that in 913 Æthelflaed ‘went with all the Mercians to Tamworth and built the borough there in the early summer’. This has been identified as a 20ha (50 acre) site of similar plan to Hereford, with a V-shaped ditch and a rampart of turf and stone, separated from the ditch by a wide berm. There is also evidence for a metalled intervallum road behind the rampart, and a bridge to carry the rampart walk over the gateway. As at Hereford, there are archaeological traces of an earlier defence of slighter construction, although this might just be the boundary of a Mercian royal palace, as there is nothing to link it to the burh street system (Gould 1967; 1968).
It is claimed that Stafford was also fortified in 913, although excavations have failed to reveal the burh defences. In fact, Stafford has been described as a thinly disguised expansion of a rural manor. In the tenth century there was no regular street grid, and no evidence for planned tenements. The central enclave appears to have contained only the minster church and three centralised crafts: butchery, bread-making and pottery manufacture. The burh may have functioned as a collecting station for the agricultural wealth of the neighbourhood. Stafford ware pottery was exported to other Mercian centres, but there was no evidence of any other commerce.
Gloucester may also have been founded as a Mercian burh, although it is not listed as such. From the sixth to the ninth centuries the shell of the Roman town sheltered a much-reduced population, probably working a number of rural holdings both within and outside the walls. The rapid build up of deposits in the old forum area in the ninth century suggests that animals were stabled here. The people of Gloucester used no pottery, and wooden and leather containers were ubiquitous. They imported little from elsewhere, and the settlement at this stage should probably be seen as a series of small estates, rather than an urban development. In 877 the remnants of the Danish army were able to camp within the town. In the tenth century, however, Gloucester suddenly acquires an administrative and military status. A substantial Saxon timber palace was built on the site of the Roman cemetery at Kingsholm, and the Mercian Council assembled here in 896. At about the same time St Oswald’s Minster was founded, and in 914 the inhabitants of Gloucester fended off a Viking attack. Aspects of the street plan may demonstrate an element of planning, with a possible tenth-century surface in the intra-mural St Aldate Street. The Roman walls were refurbished on the east, south and along part of north, but apparently not to the west, where the city had crossed the Roman boundary and extended down to river (Heighway 1984; 1987).
A similar picture of tenth-century revival is emerging from Chester (plate 28). The Anglo-Saxon Chronicle records that in 894 ‘ . . . [the Danes] marched without a halt by day and night, until they arrived at a deserted Roman site in Wirral, called Chester. The [English] levies were unable to overtake them before they got inside the fort, but they besieged it some two days.’ The Roman defences must have been largely intact if they could withstand a siege, and in 907 the north and east walls were refurbished with timber by Æthelflaed, and probably extended to the river Dee. An intra-mural gravel road was probably laid at the same time. Within the Roman fortress there were substantial upstanding Roman remains, which were re-utilised as the scale of occupation increased in the tenth century. At Princess Street a sunken building was built within the ruined walls of a Roman barrack block. At Abbey Green Roman buildings were stripped of reusable materials in the tenth and eleventh centuries and new buildings erected alongside. Tenth-century Chester became home to a multi-ethnic trading community, including a substantial Hiberno-Norse element, living to the south of the old legionary fortress, by the river Dee. At Lower Bridge Street at least five cellared buildings were erected in the tenth century and the area surfaced with sandstone chippings. Under Æthelstan, Chester became the most prolific mint in England; 24 moneyers worked in the town between 924-39. Ironically, it was probably the Irish Sea Vikings who were the main source of the city’s prosperity at this time. Finds indicate a wide range of trading contacts, with jewellery in Irish, Viking, and Anglo-Saxon styles. Chester’s importance continued until it suffered a Viking raid in 980, and the city itself may have been sacked (Austin 1996-7; Ward 1994).
THE BURHS OF WESSEX
In Wessex the Vikings also provided a major stimulus to the development of towns. It is thought that Ælfred, as a means of defence against Viking raiding parties, established a network of burhs, such that no part of his kingdom was more than 32km (20 miles) from a burh. When Edward the Elder reconquered England from 911-19 he extended the network and fortified a number of new sites (Biddle 1976; Biddle and Hill 1971; Haslam 1984; Hill 1978; Radford 1970; 1978; Williams 1984).
The burghal hidage, a document of c.914-18, lists those burhs defending the coasts and frontiers of Wessex, south of the Thames, in the later years of the reign of Edward the Elder. It catalogues 30 burhs within Wessex, and a further three outside the kingdom. London is omitted as it was technically part of Mercia; Kent is also excluded. The burghal hidage gives a tax assessment for each burh, based upon the extent of its perimeter defences. There are two groups: those generally with a large hidage which were planned as permanent settlements and market centres; their streets still display traces of the original planned layout; and a second class of temporary forts which were comparatively small (less than 16 acres) and were not regularly planned. Only the first category became towns; the second group generally no longer existed by Domesday, and were probably dismantled during the reign of Æthelstan. Most of the burghal forts withered after performing their defensive role although their market function was often transferred to another site, such as from Eashing to Guildford.
Outside Wessex mention should also be made of several Viking Age towns in Kent which were not included in the burghal hidage, but appear to have performed similar functions to the Wessex burhs. These include the Roman walled towns at Canterbury and Rochester, and the wics at Sandwich and Fordwich, as well as the sea ports at Dover, Romney and Hythe (Tatton-Brown 1988).
Where Iron Age or Roman fortifications survived, burhs were often established within the earlier defences. At Pilton and Halwell Iron Age earthworks were probably refortified; at Cadbury the hillfort was reoccupied. In Bath, Chichester, Exeter, Portchester, Southampton and Winchester the burhs made use of surviving Roman stone walls and gates. At Bath and Winchester there is evidence that the outer ditch may have been recut during the Viking Age.
Natural defences were also utilised. A large number of burhs were established on promontory or peninsular sites, frequently controlling access from the sea. These sites include Axbridge, Langport, Lyng and Watchet in Somerset, Bredy (probably Bridport), Shaftesbury, and Twynham (or Christchurch) in Dorset, Burpham and Eashing (Surrey), Malmesbury and Wilton (Wiltshire), Lewes (East Sussex), Lydford (Devon) and Eorpeburnan (probably Newenden) in Kent. At Sashes (Berkshire) a fort was established on an island in the Thames. Many of these burhs were small in extent or low in relief, but made use of natural defences such as marshland, streams or steep slopes, and therefore had an irregular plan. The main access was usually
from one direction only, and this was where the man-made defences were constructed. Lydford was sited on a wedge-shaped promontory; it was flanked by a gorge on two sides, and a narrow valley on the third; the exposed side was defended by a bank c.12m wide and a ditch 200m in length, with a central gate.
Elsewhere, new towns were created on open sites, and provided with rectangular perimeter defences modelled on Roman forts. This group comprises Cricklade (Wiltshire), Oxford, Wallingford (Oxfordshire), and Wareham (Dorset). These burhs were often established on river sites, either controlling a river crossing, as at Wallingford, or where the river ceased to be navigable, as at Cricklade.
Burhs were also sited where there was already a concentration of settlement, such as around an estate centre or minster church (see chapter 9), as at Cricklade, Malmesbury, Shaftesbury, Wareham or Wilton. These were already protourban sites, providing a focus for a non-agricultural population and acting as administrative, fiscal, religious and ceremonial centres, frequently with the association of a royal palace and minster church.
Viking Age England Page 7