Place-names and historical geography:
some issues arising from the Scandinavian settlement in England
In England, onomastics is most often studied as a linguistic discipline. This is because place-names are recorded in early documents in languages (Anglo-Saxon, Icelandic, French or Latin) which are not used today, and in spellings which bear little relation to the modern spelling of the name. For example, in the twelfth century CE the ancient city of Lincoln was recorded in some documents as Nicol, probably because the French-speaking scribes of the city at that time found the echo of /lin/ and /con/ difficult to pronounce (Cameron 1985:2), and so wrote a form that was used in official circles for a short while. This is a linguistic matter, and the strangeness of the spelling has been resolved by linguists. However, linguists (and I speak as a historical linguist) are not always the best people to interpret historical geography. My paper will relate to the interpretation of the Scandinavian settlement in England, a subject that has been debated for many years, but where linguists, historians and geographers have only recently taken each other’s approaches seriously (see Darby 1976 for a standard work which raises some doubts about the solely linguistic approach, and more recently, Abrams and Parsons 2004).
The Scandinavian Vikings attacked England over a period of nearly three centuries. The early attacks, beginning at the end of the eighth century, were for plunder, and declined in number and intensity before the end of the ninth century. Attacks were renewed at the end of the tenth century, before the Scandinavian Cnut became king in 1017. Shortly after the middle of the ninth century, Viking armies were wintering in England, and in 865, there came what the Anglo-Saxon Chronicle calls ‘the great army’, which systematically reduced the Anglo-Saxon kingdoms (Garmonsway1972:68). The only effective resistance was that of King Alfred and he reached a position of relative equality with the great Viking army under Guthrum. The treaty between the two kings, probably to be dated around 886, divided England between Wessex and the Danelaw, roughly along a line running from London to Chester, with the ancient Roman road, Watling Street, as the major boundary (Whitelock 1955:380–1). Half of England was now under Danish rule which before had been ruled by English kings. The boundary is rather neatly reflected in the place-names, many of which are Scandinavian in language, or show Scandinavian influence, north of Watling Street (Smith 1956: map 1).
This map is clear-cut, giving us the pattern of linguistically Scandinavian names in the Danelaw. In terms of numbers, in the Domesday Book 1086, there are 303 names with the Scandinavian element by ‘farm, village’ (Cameron 1987:118), 109 with thorp ‘hamlet, dependent settlement’ (Cameron 1987:141), and 57 hybrid Scandinavian-English names in the East Midlands (Cameron 1987:158). Around Leicester, there is a cluster of by-names: Groby, Ratby, Kirby Muxloe, Enderby, Blaby, Oadby, Thurnby, Barkby and so on, and there are many more along the Wreake valley towards Melton Mowbray; and along the Trent north-east of Nottingham there are names like Gunthorpe, Caythorpe, Thorpe and Staythorpe, but also some hybrid names like Gonalston, Thurgarton and Rolleston, where a Scandinavian type of personal name, Thorgeirr or Hroald, is combined with an English element, -ton. Smith’s map is clear-cut: insofar as we can be confident that names with by and thorp (and thwaite and toft and many others) are Scandinavian, and the personal names are of Scandinavian type, we can plot the linguistic influence of the Vikings on place-names.
What the map does not show, though, is what it claims to show, namely ‘Scandinavian settlement’ (Abrams and Parsons 2004). I want to illustrate why we have to be cautious in interpreting maps like this in relation to the distribution of, and numbers involved in, the Viking settlement; and in the process, I hope, I will show some of the intricacies of place-name study. I have four brief points: I will look at the historical geographic context, the name context, the onomastic interpretative context, and return to the historical geographic context to show some activities that place-names might indicate.
First the historical geographic context. The decline of the Viking attacks in the ninth century is mentioned in the Anglo-Saxon Chronicle. In 875, ‘Halfdan shared out the land of the Northumbrians, and they were engaged in ploughing and in making a living for themselves’ (Garmonsway1972:74); in 876, the other part of the Viking army went ‘into Mercia, and some of it they shared out and some they gave to Ceolwulf’ (Garmonsway1972:74); in 879 the Viking army from Cirencester went ‘to East Anglia, and occupied that land, and shared it out’ (Garmonsway1972:76); in 896 the Viking army ‘dispersed, some to East Anglia, some to Northumbria, and those without stock [or money] got themselves ships there, and sailed south oversea to the Seine’ (Garmonsway1972:89). We can press this evidence too far, in the sense that it is likely that the chroniclers did not have precise information about the events they were describing, or even very much interest; but these accounts hint at different types of settlement. In 875 we are told the Vikings began to work the land; in 876 and 879 they occupied and divided it, and since they shared it with the English under-king Ceolwulf, it may be that this was taking overlordship in Mercia rather than actually working the land themselves; by 896, retiring Vikings appear to have been able to return to lands in Northumbria or East Anglia, but those without money or stock had to go off to France to raid further, and thus the writer gives the impression that they needed money to settle. We should not press this evidence, but there must have been different types of settlement which in turn would have had varying degrees of impact on the local names.
If we go back to the distribution map of Scandinavian names in the East Midlands, we might conclude that the density of by-names around Leicester means that there were more Vikings settling here than along the Trent valley around Nottingham (map 2). But we should ask what kind of settlement and occupation was going on. The area on the outskirts of Leicester and along the Wreake valley has a whole string of tiny settlements with Scandinavian names, with relatively few English names among them. The predominance of -by names makes it likely that these small settlements were farms cut out of unused land by the settlers from the Scandinavian armies (Cameron 1987:126). -by names are mostly names of the earlier stages of settlement, formed in a Scandinavian-speaking environment, showing little English linguistic influence (Parsons and Styles 2000:104–5). The quality of the land here is poor and these settlements would not have supported a large population in the first millennium. There are thorp names in the area too, dependent settlements, smaller generally that the bys and in situations even more marginal.
Compared with the Trent valley, the Wreake valley is dense with Scandinavian names. In the Trent valley the -by names are few and the distribution is away from the river for the most part. The Trent valley has a series of moderately important, if rather small, settlements with English -ton, but with Scandinavian personal names, like Thorgeirr in Thurgarton, or showing Scandinavian influence like Fiskerton for a presumed Old English *Fisherton. Bleasby, right in the middle of the Trent flood plain, is first recorded in 958 as Blisetune (Gover, Mawer and Stenton 1940:155). The hybrid names suggest that originally these villages were English, but they changed hands when the Scandinavians arrived. The likelihood is, then, that the Trent valley was already populated by the time the Vikings came to settle, and the very modest impact of that arrival on the major place-names may bear no relation to the number of Vikings involved and the density of the settlement. In fact, in the Wreake valley there could have been relatively few Vikings who gave a relatively large number of names, and in the Trent valley a relatively large number of Viking settlers who gave relatively few settlement names. Clearly historical and geographical factors, not just names, have to be taken into consideration when assessing the Viking settlement.
Secondly, the name context. Scandinavian personal names are often different linguistically from English ones. So once again, we can point to Scandinavian influence in English place-names with some confidence. But we have to ask how far personal names relate to ethnicity, and specifically here to Viking origins. That is, were these people in any sense Vikings? Earl Godwine’s family, the leading family of mid-eleventh-century England under the king, is case in point. We know from literary sources that Tosti thought of himself as a bit of a Viking and fought on the Norwegian side at the battle of Stamford Bridge in 1066 against his brother, Harold Godwinesson (Garmonsway 1972:196); but that might be as much due to the fact that he was in the shade of his brother as to any sense of Scandinavian ethnicity. In the battle of Maldon in 991 (Scragg 1991), the Old English poem records a Wistan (English name) son of Thurstan (anglicised Scandinavian name) fighting against the Vikings in Essex. Archbishop Oda of Canterbury, who died in 961, was of Scandinavian name and origin, but how much of a Viking he thought himself to be must be left to conjecture.
In a charter from 1046 (Whitelock 1925:84–7), a woman called Wulfgyth, widow of Alfwine, and sister of Eadwine and Wulfric leaves property in East Anglia. All these names are English. But Wulfgyth mentions her sons and daughters: the sons are Elfketil, Kytel, and Wulkitel, all anglicised versions of Scandinavian names; the daughters are Bote, Gode and Ealdgyth, all English names (Fellows-Jensen 1994:19). Clearly these male names were fashionable in East Anglia, and if these men should have named their farms Kettlethorpe, as is perfectly possible, it would be natural for us to think that they were named by Scandinavian settlers when in fact they might have been nothing of the sort. My second note of caution, then, is that we need to be careful not to assume that Scandinavian names have any necessary reference to Vikings.
My third point relates to interpretation of names and spellings. Here are early spellings for two Lincolnshire place-names, Roxby and Scawby.
Roxby, Lincolnshire (Cameron 2001:93)
Roscebi 1086
Roxebi, -by 1086, 1201
Rochesberia 1090–1100
Scandinavian Hrók + bý
Scawby, Lincolnshire (Cameron 2001:101)
Scaleby 1086 DB, 1194
Scaleberia 1090–1100
Scandinavian Skalli + by
These spellings appear to represent a confusion of Old English byrig dative of burh ‘fortification’ and Old Danish by ‘farm, estate’. It is difficult to be sure which spellings to trust: do we go for the earliest spelling, the majority spelling, or the one which survives? The first element seems to be the deciding factor in Roxby and Scawby: the first element is a personal name of Scandinavian origin and the idea is that Scandinavian elements go best together. We disregard the -beria spellings and attribute them to an eleventh and twelfth century fashion for spelling some bys as if they were byrigs — perhaps a local pronunciation or scribal habit.
Now compare this with these names from Cheshire.
Greasby, Cheshire (Dodgson 1972:291)
Gravesberie 1086
Grauisby 1096–1101
Greuesby 1249–1323
Greseby 1271
Whitby, Cheshire (Dodgson 1972:198)
Wieberia 1096–1101
Witebi(a) 1096–1101
Witeby 1260
Whiteby 1241
In both of these cases, we only have one spelling which might suggest that the names are English. There is no immediately obvious Scandinavian element for the first part of Greasby, but this Whitby shares spellings with the North Yorkshire Whitby; this we know to be Scandinavian because it replaced the Old English Streonaeshalch, and it contains either the personal name Hviti or the adjective ‘white’. Yet these Cheshire names are interpreted as English, namely ‘stronghold at a wood’ with OE græfe and burh, and ‘white or stone-built stronghold’ with Old English hwit and burh. Here, then, we disregard the majority spellings and present-day forms and decide that these are originally English names modified by Scandinavian naming-patterns in the locality. Compare the interpretation of the two pairs of names, and it can be seen that we have decided on the basis of an impression that two names are Scandinavian and two English. So a third note of caution is that even in assessing the linguistic composition of names there is an element of doubt.
Fourth, the historical sources give much more attention to military matters than they do to more peaceful interchange between people. The first recorded Viking ‘raid’ on England in the Anglo-Saxon Chronicle was probably a trading expedition that went wrong (Garmonsway 1972:55):
789. [In the days of King Beorhtric] came three ships for the first time... the reeve rode thither and tried to compel them to go to the royal manor, for he did not know what they were: and then they slew him. These were the first ships of Danes to come to England.
The writer interprets this attack with hindsight as a Viking raid; if the visitors had traded peacefully, it would have gone unremarked. We know from archaeology that the Vikings were great traders: Dublin (Ireland) and York (England), Hedeby and Ribe (Denmark) were markets for goods from all over the world, and they were developed by the Scandinavians. Scandinavians also made rather a lot of coins. If we want more information about the Vikings, it may be that place-names can give us more evidence of trading than is to be found in the documentary record.
English names for trading-places use the word ceap ‘trade, merchandise, market’ quite frequently, as in Chipstead (Kent and Surrey), Chepstow (Monmouthshire) and the Chipping names (Chipping Barnet, London; Chipping Norton, Oxfordshire etc.). The English names Chipstead Surrey and Chippindale Lancashire are recorded in Domesday Book 1086 or earlier. There are also references to merchants in Chapman names: Chapman Sands Essex, Chapmanslade Wiltshire and so on, but none of these is very early recorded, or a major place.
The Scandinavian equivalents are kaup and kaupmaðr, and these are the ones I have found:
Capernwray Lancashire, Coupmanwra c.1200 ‘nook of the merchant’, with ON vrá (Ekwall 1922:187)
Warcop Westmorland, Coupland, Coupemanebec, Coupelandbek 13th cent., ‘stream on the merchant’s [or merchants’] land or bought land’, with ON bekkr (Smith 1967:83)
Copeland Barony Cumberland, Caupalandia 1125 ‘bought land’, foreste de Coupland 1282 ‘forest of Copeland’, with OFr forest (Armstrong, Mawer, Stenton and Dickins 1950:2, 37)
Copmanthorpe West Riding of Yorkshire, Copeman Torp 1086 DB, ‘merchant farm’, with ON þorp (Smith 1961:227)
Copeland Durham, Copland c.1040, ‘bought land’ (Watts 2002:29)
Coupland Castle Northumberland, Coupland 1250 ‘bought land’ (Mawer 1920:52)
Cockpit Hill Derbyshire, Copecastel 1085 ‘market castle’, with OFr castel (Cameron 1959:451)
Chatsworth Derbyshire, Coupmonlondes 1296 ‘merchant lands’ (Cameron 1959:73–4)
Cowmans Essex, Comanneshey 1289, Copmaneshey 1291, ‘enclosure of Copeman, the merchant’, with OE gehæg (Reaney 1935:390–1)
Coppingford Huntingdonshire, Copemaneford 1086 ‘ford of the merchants’, with OE ford (Mawer and Stenton 1926:237–8)
Buscot Berkshire, Copemaneford 1199 ‘ford of the merchants’ (Gelling 1974:355)
I have many later medieval names, and of course much later names with the surnames Copeland or Copeman deriving from these are too many to list.
The interesting things about these names are that they are quite early (before 1300), there are major settlements or land units among them, the geographical spread is wide (including the Berkshire name outside the Danelaw) so they are unlikely to be transferred names, and they combine freely with French and English elements as well as Scandinavian. All these factors, I suggest, point directly or indirectly to Scandinavian mercantile activity, activity which had an impact on the communities in which it took place, and which was consensual and co-operative. By comparison with the English names of the same type, these names are generally earlier and the places generally more important, and we could suggest that these names indicate a new and positive contribution from Scandinavians to the economy of Anglo-Saxon England.
If we cannot always pin the Vikings down to particular places, we can nevertheless see the huge and widely diffused impact that they actually had on the language and naming fashions of Anglo-Saxon England and later. But we can also see that onomastics needs the insights of historical geography in order to interpret its data properly. The place-name survey on which I am working aims to incorporate as much of the available historical and geographical data as possible in the hope of getting a clearer picture of the complex factors that led to the naming patterns we find in England. But on the other hand, too, onomastics can contribute important insights to historical geography.
References
Abrams L. and D. N. Parsons (2004), ‘Place-names and the history of Scandinavian settlement in England’, in Hines, J., A. Lane and M. Redknap, ed., Land, sea and home, Leeds, pp. 379–432.
Armstrong, A., A. Mawer, F. M. Stenton and B. Dickins (1950), The place-names of Cumbria, English Place-Name Society, Cambridge.
Cameron, K. (1959), The place-names of Derbyshire, English Place-Name Society, Cambridge.
— (1985), The place-names of Lincolnshire, part 1, English Place-Name Society, Nottingham.
—, ed. (1987), Place-name evidence for the Anglo-Saxon invasion and Scandinavian settlements, Nottingham (reprinted articles from various sources).
— (2001), The place-names of Lincolnshire, part 6, English Place-Name Society, Nottingham.
Darby, H. C. (1976), A new historical geography of England before 1600, London.
Dodgson, J. McN. (1972), The place-names of Cheshire, part 4, English Place-Name Society, Cambridge.
Ekwall, E. (1922), The place-names of Lancashire, Manchester.
Garmonsway, G. N. (1972), The Anglo-Saxon chronicle, London.
Gelling, M. (1974), The place-names of Berkshire, English Place-Name Society, Cambridge.
Gover, J. E. B., A. Mawer and F. M. Stenton (1940), The place-names of Nottinghamshire, Cambridge.
Fellows-Jensen, G. (1994), The Vikings and their victims: the verdict of the names, Dorothea Coke memorial lecture, London.
Mawer, A. (1920), the place-names of Northumberland and Durham, Cambridge.
Mawer, A. and F. M. Stenton (1926), The place-names of Bedfordshire and Huntingdonshire, English Place-Name Society, Cambridge.
Parsons D. and T. Styles (2000), The vocabulary of English place-names: Brace–Cæster, Centre for English Name-Studies, Nottingham.
Reaney, P. H. (1935), The place-names of Essex, English Place-Name Society, Cambridge.
Scragg, D. G., ed. (1991), The battle of Maldon, AD 991, Oxford.
Smith, A. H. (1956), English place-name elements, 2 vols, English Place-Name Society, Cambridge.
— (1961), The place-names of the West Riding of Yorkshire, part 4, English Place-Name Society, Cambridge.
— (1967), The place-names of Westmorland, part 2, English Place-Name Society, Cambridge.
Watts, V. (2002), A dictionary of County Durham place-names, English Place-Name Society, Nottingham.
Whitelock, D. (1925), Anglo-Saxon wills, Cambridge.
Whitelock, D., ed. (1955), English Historical Documents Vol. I, c. 500-1042, London.
