The second problem concerns the age of those settlements that received Scandinavian names. If they are evidence for a massive Scandinavian migration then they should be new sites on virgin ground, but many may simply have been new names for existing places, just old estates under new management. Excavation at Whitby and Osbournby, for example, has revealed evidence for pre-Scandinavian settlement at both these sites. It is now widely accepted that much of England was already intensively farmed at the beginning of the Viking Age. If so, then the high proportion of Scandinavian placenames must represent a renaming of existing settlements.
Many Scandinavian style place-names involve the endings -by, -thorp or -ton compounded with a personal name. In Yorkshire and the East Midlands 40-60 per cent of names ending in -by have a personal name as their first element. This can be explained by the fact that it was during the Scandinavian settlement that much land passed into smallscale private ownership. Previously land had been held in the great or ‘multiple’ estates, frequently under group ownership. In the Viking Age these large estates were broken up and their component elements were handed out to individuals. Scandinavian place-names, therefore, mark not so much an extension of settlement as its reorganisation under new lords. Their density reflects not so much the areas of migration as areas of break-up of older great estates into individual ownership. The Yorkshire Wolds, for example, was a very fertile area which had been farmed intensively from the late Iron Age. The large number of settlements bearing Scandinavian names are not new sites, but dependencies of estates based around the Wolds which were taken over by Scandinavians. We can therefore see which estates were preserved and which were broken up. The area around Bardney Abbey, for example, has no Scandinavian placenames, in contrast to the area around Whitby where there are many.
There is no need to postulate, therefore, a mass folk migration in order to explain the distribution of Scandinavian place-names. Names are given to places when it becomes necessary to refer to them unambiguously. The Anglo-Saxon Chronicle records three partitions of land between the Scandinavians and the Anglo-Saxons, in Northumbria, Mercia, and East Anglia (see chapter 2). In each case the Viking leaders rewarded their followers with allocations of land. Both the new farmers and those expecting tribute from them would have been keen to legitimise their claims to this land, and what better way than by naming it after them? As the habit of buying and selling land developed in the tenth century it remained important to identify the owners, be they Anglo-Saxons or Scandinavians.
LINGUISTIC CHANGES
It is often suggested that other linguistic evidence provides clear proof that the Scandinavian settlements were on the scale of mass folk migrations. Scandinavian pronouns, verbs and other everyday words, such as the words for husband, knife and window, were adopted by the English language. Some have contested that such changes could not have occurred unless the Vikings were in a majority, but other linguists have persuasively argued that it is misleading to draw conclusions about numbers on the basis of linguistic changes (see Ekwall 1930; Hines 1991; Page 1971; Townend 2000). They suggest that the influence of one language upon another depends on their relative status, and by the need to borrow words to describe new things. It is unlikely that ninth-century Northumbrians would have been able to understand Danes and Norwegians easily. Communication would rely on a few individuals who knew both languages. Nevertheless, the similarity between Danish and English would mean that it was easy for the English to adopt Scandinavian words. In particular, the introduction of a large number of Scandinavian words associated with farming indicates that there was an influential Danishspeaking farming population. Most evidence suggests that Scandinavians adopted English fairly rapidly, adding a few of their own words. Vernacular inscriptions from north and east England show a clear continuity in the use of English from the eighth to the eleventh centuries. There is no evidence for Scandinavians continuing to use either their runic script or their own language in England.
PERSONAL NAMES
The number of individuals bearing Scandinavian personal names has also been used as a measure of the density of Scandinavian settlement (Fellows-Jensen 1968; Insley 1994). The number of Scandinavian personal names compounded with settlement names in Yorkshire and Lincolnshire has already been noted. In the Domesday Book 40 per cent of names in Derbyshire are Scandinavian; as are 50 per cent in Nottinghamshire and in Cheshire. Of course, name-giving habits change with fashion, and are particularly prone to influence from the elite, as testified by the number of children named after royalty, media stars or footballers. Scandinavian names may have increased in popularity during the Viking Age, but they do not necessarily denote Scandinavians. Members of the same family mentioned in Domesday have both Scandinavian and Anglo-Saxon names. It also tends to be forgotten that written sources are restricted to the elite group. Most of those individuals mentioned in the Domesday Book are referred to as manor holders. Name-giving may reveal more about class allegiance than ethnic identity (Hadley 1997).
Coins record the names of the moneyers who were responsible for coin production (Smart 1986). The proportion of Scandinavian names increases from zero under Ælfred to 3 per cent under Edward the Elder, 5 per cent under Æthelstan, to 15 per cent under Eadred and his successors. Regionally there is considerable variation. In York c.75 per cent of names are Scandinavian by the reign of Æthelred; in Lincoln c.50 per cent, Chester c.25 per cent, and London c.7 per cent. Yet even if all these individuals were born in Scandinavia they hardly represent a cross-section of tenth-century England. Name-giving habits may well have been profoundly affected by the Scandinavian settlement, but Scandinavian names may still have been confined to the landowning and mercantile classes.
In summary, settlement organisation and language cannot be used as evidence for a substantial Scandinavian migration. What place-names do show is that whatever the size of the invasion, its political implications should not be underestimated. A major consequence of the Viking raids was the fragmentation of the great estates. The Vikings stimulated a new market in the buying and selling of property. Land was taken into private ownership, and often named for the first time.
ARCHAEOLOGICAL EVIDENCE
Few Viking Age rural settlements have been excavated, and even fewer can be positively identified as the homes of Scandinavian settlers (7). The first problem is that most of the successful farmsteads must have grown into medieval villages, and now lie buried under their modern successors. The second is that the typical rural excavation yields few finds, making cultural identification difficult. The Anglo-Saxon migrations can be recognised archaeologically from new settlements and building types which appear over southern and eastern England. Yet we cannot assume that new peoples will inevitably introduce new forms of settlement and, particularly if their way of life was identical to that of the existing population, there is no reason why their farms should be distinguishable from those of the indigenous population. Incoming and native peoples may also interact to create new or hybrid forms. Therefore, all the settlement evidence for the period should be considered, ranging from palaces and manorial centres to isolated upland farmsteads.
7 Map of excavated Viking Age settlements
Our picture of ninth- to eleventh-century royal accom-modation is derived from excavations at Cheddar (Somerset). The site probably functioned as a royal estate centre from the second half of the ninth century, and may well have been used by Ælfred. The first phase included a long hall, possibly of two storeys, erected within a palisaded enclosure with a possible gatehouse. The remaining buildings probably served as private residential buildings and offices. Circa 930 the hall was replaced by a more substantial timber hall and a stone chapel. A massive post-hole outside the entrance may indicate that a flagstaff or decorated pillar stood there. The hall was rebuilt a third time in the late tenth century and a raised dais was added, probably during the reign of Æthelred. The relative cleanliness of the site and the lack of spectacular finds suggests that it may only have
been occupied periodically during royal visitations, and was stripped to a skeleton caretaker staff in the interim. Documentary sources suggest that the Anglo-Saxon court or witan met here three times during the tenth century, in 941 under Edmund, 956 under Eadwig, and in 968 under Edgar (Rahtz 1979 but see Blair 1996).
At North Elmham (Norfolk) excavations have revealed details of life within an episcopal palace complex. A Middle Saxon cathedral community abandoned the site in the early ninth century, possibly as a result of Viking raids, but sometime after 917 the site was cleared and levelled for a new minster and ancillary buildings. A new church was erected over the earlier ruins and a large timber hall, possibly for the use of the bishop, was erected nearby. In the eleventh century a new palace was built elsewhere, and the site was given over to secular use. It was colonised by farmers, with peasant dwellings, sheds and animal pens set in fenced enclosures (Wade-Martins 1980).
Other excavations have illuminated the decline of the great estates and the evolution of the manor during the Viking Age. The extensive landscape project at Raunds has demonstrated the fragmentation of a multiple estate into several component manors over several centuries. At Furnells Manor a Middle Saxon settlement in a ditched enclosure was replaced by a large timber hall, the proto-manor house, in the early tenth century. A small stone church was built adjacent to the enclosure. At about the same time, the first regular tenements of peasant farmers were being laid out at Furnells and West Cotton, marked by ditched enclosures. As none of the ditches could have offered more than minimal protection their main purpose must have been for laying out. Indeed, the evidence at West Cotton shows that equal plots, c.20m wide, were created (Cadman and Foard 1984; Foard and Pearson 1985).
A similar process may have been taking place in another area of Scandinavian settlement, at Wharram Percy (North Yorkshire). The South Manor site at Wharram was occupied from the early eighth century. The date of the laying out of the peasant tofts and crofts is still uncertain, but probably took place during the Viking Age, in the tenth century (Hurst 1984; Beresford and Hurst 1990). The lack of Late Saxon pottery from the fields between the villages, compared with the abundance of Middle Saxon sherds, suggests that the villages in this area were also becoming nucleated at that time. Significantly there was continuity of the elite site from the Anglo-Saxon to the Anglo-Scandinavian phase at Wharram. In the seventh and eighth centuries there was a post-built hall and an associated weapon smithy on the site that was to become the South Manor after the Norman Conquest. This hall was replaced, and the new residence has not been found, but the weapon smithy continued in operation somewhere in the vicinity, and a sword hilt guard has been excavated, similar in appearance to one from Coppergate. There were also honestones of a type of stone peculiar to Norway, and a strap end and belt slide decorated in the Borre style, current in Scandinavia in the ninth and tenth centuries (Stamper and Croft 2000). If not Norwegian himself then the new tenant of Wharram Percy had certainly adopted Scandinavian dress fashions.
In some cases a mother settlement might be replaced by a number of daughter sites. At Chalton (Hampshire), for example, the Church Down site was abandoned by the ninth century, but the nearby villages of Chalton (Manor Farm), Idsworth and probably Blendworth were occupied from about this time. At Cottam (East Yorkshire) parts of an eighth- and tenth-century landscape have been examined as a result of collaboration between archaeologists and metal detectorists (Richards 1999). The first settlement comprised a number of post-built halls set in an enclosure and adjacent to an ancient trackway. This site was rich in metal artefacts and I have suggested that it may have been a local centre for a Northumbrian royal vill at Driffield. In the late ninth-century, however, this site was abandoned. A weathered skull, a possible execution victim, was dumped in a pit adjacent to one of the buildings, and sealed with rubbish, including a coin of ad 858-64. The settlement now shifted to an adjacent site approximately 100m to the north, where an Anglo-Scandinavian farmstead was constructed. The new farmstead comprised a number of regular planned sub-rectangular enclosures (plate 4). It had been badly damaged by later ploughing but there were clusters of post-holes representing a number of buildings, at the head of a ditched trackway. The entrance to the site was a very grand affair, comprising a massive external ditch, an internal rubble bank, itself possibly topped by a timber palisade, and a timber gate superstructure. Given the lack of ditches around the rest of the site this must have been constructed for status rather than defence. Associated tenth-century finds include a Borre-style buckle, a Jellinge-style brooch, and a pair of so-called Norse bells (plate 32) (see chapter 7). This need not necessarily mean that the inhabitants of the tenth-century farmstead were Scandinavian settlers who had dispossessed their Anglian predecessors. Nonetheless the ostentatious new farm suggests nouveau-riche farmers, or old farmers identifying with a new elite. Occupation of the new site was short-lived, however. During the tenth century the settlement shifted again, and new sites laid out which developed into the deserted medieval villages represented by earthworks of tofts and crofts at Cottam and Cowlam.
Many pre-Conquest manorial residences were set within fortified enclosures. In some cases existing fortifications were utilised. At Portchester (Hampshire), several substantial timber houses were erected within the walls of the former Roman fort in the late eighth or ninth centuries. Finds of east Mediterranean glass and coins of Burgred of Mercia testify to the relative wealth of the site. In 904 it was acquired by the king as a royal burh, but it never became a town and appears to have continued to function as a manor site. In the late tenth century a substantial aisled hall and three subsidiary buildings were constructed, served by impressive timber-lined wells. Massive dumps of food waste may represent the remains of great feasts. In the early eleventh century the owner erected a free-standing flint and masonry tower, some 6m square, on a plot of land between the hall and the subsidiary domestic buildings. This may have been a bell-tower or chapel, as it became a focus for burials. It brings to mind that one of the requirements of an eleventh-century thegn was that his property should have a bell-tower (Cunliffe 1976).
At Faccombe Netherton (Hants) an aristocratic residence has been excavated on the north-east edge of Salisbury Plain. It may be that which was mentioned in a will of c.950 as belonging to a noblewoman named Wynflæd, who was probably the mother-in-law of King Edmund. The residence comprised an aisled hall, a building of private apartments with a latrine, and a separate kitchen. In the early eleventh century the site was redesigned with a substantial bank and ditch. By this stage a church had been built adjoining the manorial complex (Fairbrother 1990).
Another late tenth- to early eleventh-century thegnly residence has been excavated at Sulgrave (Northampton-shire). The parish church, which is Anglo-Saxon in origin, is on the same alignment as a tenth-century timber hall and may have originated as a private chapel. At Sulgrave a separate stone building set to one side of the timber hall may have been the base of a separate belltower. The residences at Sulgrave, and at Goltho (Lincolnshire) were each surrounded by a defensive ditch (Davison 1968).
At Goltho an early ninth-century farmstead was superseded by a fortified earthwork enclosure, in which a fine bow-sided hall like that at Cheddar, a bower, a kitchen, and weaving sheds formed three sides of a rectangular enclosure. The kitchen would have provided food for feasts when followers were entertained in the hall. The fortifications comprised a 6m-wide rampart topped by a timber palisade, surrounded by a ditch 5.4m wide and 2-2.4m deep. They are as substantial as those protecting the burhs at Cricklade and Tamworth, and must have been constructed for serious defence, perhaps against the threat of Viking raids. The manorial complex may have been founded by a member of the Saxon aristocracy, although the discovery of a Scandinavian-style bridle bit could be used to argue that it was a late ninth- or tenth-century Viking foundation. During the late tenth and early eleventh centuries the site underwent considerable expansion. The hall was replaced by an aisled version without inter
nal partitions, which would be more suitable for use on official occasions and for estate functions, and the bower was enlarged and partitioned, with a latrine attached to it at one end. After the Norman Conquest it developed into a motteand-bailey castle (Beresford 1987).
Not all rural sites developed into manorial complexes. Catholme (Staffordshire) was already in decline before the Viking Age and was abandoned in the early tenth century. During its latest phase the settlement comprised groups of buildings linked by fences, defining paddocks or small yards. Within the central enclosure a bow-sided hall was the most substantial building (Losco-Bradley and Wheeler 1984). Tenth- and eleventh-century buildings set within ditched or fenced enclosures have also been identified at St Neots and Little Paxton (Cambridgeshire) (Addyman 1969; 1972). These appear to represent individual farm units with ancillary buildings, wells, homefields, and droveways leading to a village centre. Other sites, whilst not large enough to be described as villages, comprise more than single farmsteads. At Springfield (Essex) nearly a dozen rectangular houses of the ninth to eleventh centuries have been excavated. The suggestion of a rectangular tower may mark this as another proto-manorial site; although the 15 rubbish pits contained few rich finds there was evidence for pottery and quernstones imported from Germany (Buckley and Hedges 1987).
Viking Age England Page 5
