Page images
PDF
EPUB

by no means limited to personal names: an enormous number of English place-names contain common nouns which are Danish or Norse. The usual form is not Grimston (which occurs in Leicestershire and Yorkshire), but Grimsby-not the Lancashire Gamelsley, in which the Norse personal name Gamall is compounded with the Saxon leah, a meadow or field, but rather Gamblesholme (also in Lancashire) or the Yorkshire Ganthorpe (in Domesday Gameltorp) in which it occurs with the Scandinavian common nouns holm and thorp. It is just the opposite with the Norman-French stratum. The most frequently occurring French element in English place-names is simply the name of a Norman family tacked on to an old English or Scandinavian village name as a means of distinguishing a manor held by that family from some other manor of the same name. Thus we get such names as Acton Turville, Coatham Mundeville, and Ashby-de-la-Zouch. These only point to the presence of a Norman aristocracy. But the large numbers of names ending with unmistakably Scandinavian common nouns -by, thwaite, holm, garth, and gill (not to mention the less certain thorp)-clearly indicate an immigration sufficient to make the language spoken by ordinary people in the neighbourhood, at least very largely, Scandinavian. An ordinary cycling map of Lincolnshire shows, within a radius of eight miles round Spilsby, over 50 names ending in by, as well as two or three thorpes. Scandinavian suffixes are found combined not only with personal names but with a great variety of other Scandinavian words-for example, we may take the name Rosthwaite which occurs several times in the Lake District and is from the old Norse hross, a horse, and thwaite, a clearing or meadow. In some instances, too, the form of a name shows inflexion, which proves that Scandinavian case endings were used in the speech of the locality.

The invasions of the Northmen involved extensive immigration; and the frontier of their dominion was pushed as far south as the line of Watling Street and the course of the Bedfordshire Ouse, the Lea, and the lower Thames. Those were the boundaries fixed by the peace concluded between Alfred and Guthrum. But a study of place-names shows us that there were some

[graphic]
[graphic]

settlements outside the region marked off by this treaty and also that the Northmen did not colonise the whole of the Danelaw. South-west of Watling Street we notice the name Rugby. There was a settlement in Pembrokeshire there was another little colony in Glamorganshire, as the name of Swansea seems to indicate. Lundy Island is said to take its name from the old Norse word for a puffin; and it is possible that the Helford River in Cornwall is really a fjord. It is more important to notice the limits of the settlements within the area of Danish rule. Suffolk was no doubt part of Guthrum's kingdom of East Anglia, of which the 'Chronicle' tells us that the Danes settled there in 880 and shared out that land.' But there is a striking scarcity of Scandinavian place-names in Suffolk. Only four names ending in by have been noted, and in general Dr Bradley says the evidence 'points to the conclusion, which the phenomena of the modern dialect tend to confirm, that the settlements of Danes and Northmen in East Anglia were mainly confined to Norfolk, and probably to the northwestern corner of that county.'" The county of Huntingdon again was well within the Danish frontier; but Scandinavian names are rare there. In the north the evidence of place-names supplies a corrective to the loose language of the chroniclers. The Alfredian Chronicle tells us that in 876 Halfdene portioned out the lands of Northumbria, and they thenceforth continued ploughing and tilling them'; and Snorri Sturluson, in the 'Heimskringla,' remarks that 'the country of the Northumbrians was mostly inhabited by Northmen since the sons of Lodbrok acquired that country.' These statements can hardly be accepted as regards that part of Northumbria which lay north of the Tees. In the county of Durham only eight by names have been noted by Prof. Mawer and only one toft and one garth. In Northumberland he says there are no examples of by, beck (Wansbeck is deceptive), toft, thwaite, garth, scale'; and he tells us that the vast majority of the names both in Northumberland and in Durham are of English or, more strictly speaking, of Anglian origin.' Nearly 70

6

[ocr errors]
[ocr errors]

English Historical Review,' XXVIII, p. 797. Dr Bradley in this passage uses the term Northmen in its more restricted and more exact sense. I have used it as simply equivalent to Scandinavians.

[ocr errors]

Scandinavian personal names, however, occur in these counties, but they seem to be mostly compounded with English common nouns, and therefore at most indicate the presence of individual settlers.

[ocr errors]

Even in districts of extensive settlement the Northmen did not obliterate the old population to the same extent that their Anglo-Saxon predecessors obliterated the Romano-British peoples. Yorkshire and Lincolnshire,' says Dr Bugge, 'formed the heart of Scandinavian England.' This is no doubt true; but he overstates the case when he adds that these counties' must for centuries mave remained an entirely Scandinavian country.' In the midst of that great concentration of by names in the district round Spilsby several names ending in ton and ham appear upon the map. As regards the East Riding of Yorkshire, Isaac Taylor long ago called attention to the frequent occurrence of parishes containing two townships, one of which has an Anglian and the other a Danish name. Thus the parish of Settrington, of which Taylor was rector, contains the two townships of Settrington and Scagglethorpe. Similarly Langton contains Langton and Kennythorpe: in Catton we have Catton and Kexby-in Cottingham, Cottingham and Willerby. In each of these cases it will be noticed that the larger place which gives its name to the whole parish has an English ending, while the names of subordinate places have endings in by or thorpe. Taylor remarks that it is quite exceptional to find the name of the parish ending in thorpe or by, and the subsidiary Cownships in ton or ham.' He offers the suggestion that the township which gave a name to the parish was the original Anglian settlement, while the later Danish mmigrants settled on outlying waste lands.'*

The settlements in the north-west of England are especially interesting. We read little enough about hem in the chronicles. But Prof. Ekwall's study of place-names has done much to fill in the gaps in our nowledge: he has read the palimpsest of the map with skill amounting to genius.

The Lake District has long been recognised as one of the most Scandinavian parts of Britain. In Cumberland

[graphic]
[ocr errors][merged small][merged small][merged small]
[graphic]

alone there are more than 100 names ending in thwaite, which is a characteristic Scandinavian suffix. In Lanca shire, as Prof. Ekwall remarks, 'place-names wholly or partly Scandinavian abound.' The Wirral peninsula of Cheshire-to quote Dr Bugge-'teems with Norse names.'

In these western settlements the linguistic evidence points to a large West Scandinavian, or Norse, element. Thwaite is usually considered to be Norse rather than Danish or Swedish, though Prof. Ekwall does not include it in his list of West Scandinavian test-words. In his opinion the real tests of Norse influence are bud'booth,' gil 'gill', skáli a 'hut'-which we get, for example, in the Cumberland Seascale, the Westmoreland Holmescales, and the Lancashire Loudscales-and also breck from the old Norse brekka, a 'slope' or 'hill,' slakke from the old Norse slakki, a 'valley,' and perhaps ergh a 'shieling' or shelter in a hill pasture, which is a word borrowed by the Norwegian vikings from the Irish The application of such tests as these has led Prof. Ekwall to the conclusions that in Cumberland West Scandinavian names abound,' that in Westmoreland there are numerous names in gill and scale,' and that 'the Scandinavians in Lancashire must have been predominatingly Norsemen, Norwegians.' His researches have not been extended into the Wirral peninsula, though he approves Prof. Collingwood's suggestion that the test word ergh is embedded in the name Arrow in this district, which occurs in the form Harche in 1312, has a fairly high situation, and can be paralleled by the name Little Arrow near Coniston.

The evidence further indicates that these Norse settlements were largely made by Vikings who had previously lived for some time in Ireland. The presence of the Irish loan-word ergh has already been mentioned. A charter of about the year 1070 mentions a certain Torfynn Mac Thore as a landholder in northwest Cumberland. Here we have two Scandinavian personal names compounded with the Irish patronymic 'Mac.' Besides ergh the place-names of Lancashire and the Lake District incorporate other words, as well as personal names, which are not merely Celtic, but belong to the Goidelic branch of the Celtic language, so that

they cannot be legacies from the Brythonic Celts who inhabited this part of Britain before the Germanic invasions. Again, these Goidelic elements are frequently -Prof. Ekwall says the Goidelic personal names are 'in the majority of examples'-combined in place-names with words of Scandinavian origin. He also points out that 'Goidelic names are almost as common in Lancashire as in Cumberland,' and adds 'this warrants the conclusion that the majority of these names were introduced by Scandinavians, and are not due to Gaelic influence, for in Lancashire Gaelic influence is hardly to be reckoned with.'

Another very interesting fact has been noticed in regard to this stratum of names. In ordinary Germanic names, whether Anglo-Saxon or Scandinavian, the defining word-the adjective or possessive-usually comes before the word to be defined. In English we still speak of 'Smith's farm,' rather than of the 'farm of Smith,' and put our adjectives before our substantives. This is the reason why the endings of place-names in England are much less various than their first elements-why the common nouns ham, ton, by, thwaite, thorpe occur at the ends of names. In the Celtic names of Wales and Ireland, however, the opposite arrangement prevails; and a glance at the map is enough to show that the frequently recurring elements are initial-witness the Llans and Abers and Caers of Wales, the Ballys and Kils of Ireland. Now among the Norse or partly Norse placenames of the north-west of England we find a certain number in which the order of the elements is after the Irish model. Prof. Ekwall calls these names 'inversioncompounds.' Butterilket, the name of a farm in the Valley of the Esk above Boot in Cumberland, is a good example. The name is entirely Scandinavian : it means the booths or dairy farms of Úlfkell. But its constituents are in the Celtic order-it is Booths of Ulfkell, not Ulfkell's Booths, which would be the normal Scandinavian form. In the name of Aspatria (Cumberland) we have the West Scandinavian askr, an ash tree, compounded with the Irish name Patrick, the elements being again in the Irish order. But how is it that we also get names half Norse, half Goidelic, in which the Scandinavian order is preserved-such names, for example, as

« PreviousContinue »