Geography, toponymy and political organisation Geography, toponymy and political organisation in early Scandinavia by Stefan Brink

Th e accounts of the voyages of Ohthere and Later, the name occurs in a runic inscription Wulfstan are, together with Rimbert’s Vita from Jelling (ca 980) in the form Nuruiak. Anskarii, the oldest written descriptions of Th e form of this name on one of the two Jell- Scandinavia that we have, although Jorda- ing stones has caused problems to philolo- nes’ listing of several ethnic groups in Ultima gists, who would have expected the dental Th ule around the year 500 is more than three (ð) in the name. Adolf Noreen therefore as- centuries earlier. We are here for the fi rst time sumed that the fi rst element was nór ‘short in writing introduced to a number of place river, narrow bay’ and that the name was to names and historical matters. Although these be interpreted as ‘the narrow way’.4 Th is is travel accounts contain few descriptions and not necessary, since we have the dental in the provide little information about Scandinavia, oldest surviving forms of the name, Nortua- the sources are of extreme importance for the gia and also Norðweg in the Old English Oro- understanding, or rather interpretation, of sius. Most certainly Norway goes back to a the geography and political structure of the Proto-Nordic *Norð(r)vegr, where the frica- early Viking Age. tive dental must have been lost early, reduced To begin with the geography: in the ac- between two other consonants, in the same count of Ohthere’s report to King Alfred, way as in the Old Norse adjective norrœnn it is stated that it took more than a month ‘northerly’ (< norðrœnn). We may compare for Ohthere to travel by ship from his home Old Norse vestrvegr ‘land to the west’, aus- in Hálogaland (mod. form Hålogaland) in trvegr ‘land to the east’ and suðrvegr ‘land to northern Norway down to Kaupang (Sciringes the south’ (used in particular of Germany, healh) in Vestfold, following the coast with a Italy etc.).5 favourable wind and camping at night.1 Th is Initially Norway must have been the account seems fair enough if one compares it name of the sailing route along the Nor- with the information Rimbert gives for Ans- wegian coast, from Viken and Agder up to gar’s second journey to Birka around 850;2 Hálogaland, most certainly a very important here it says that this trip from Schleswig to trade route with the kaupangr at Skiringssalr Birka took twenty days. as the - or, at least, a - hub, where luxury com- Th is often used sailing route along the modities from the arctic regions were stored Norwegian coast, from Hálogaland in the and traded with Danish, German, English north down to Viken, which Ohthere de- and Frisian merchants. It is generally ac- scribes from his home down to Kaupang, cepted that the name of this well-known sail- became so identifi ed with the land along the ing route must, through metonymic transfer, route that it gave its name to the country, subsequently have been extended to the ac- namely Norway, and the second oldest ref- tual land mass along the northerly route. We 1. Cf. Sawyer 2002: 87. erence to this is to be found in the account may compare the transfer of the name Red- 2. Vita Anskarii 26. of Ohthere’s report (Norðweg, ‘North way’); väg ‘the riding path or route’, used of a land 3. Seip 1923: 9–19; Stem- shaug 1976: 129. in a Latin source from ca 840 we have the route from the sea to Falbygden in the prov- 4. Noreen 1897: 22. 3 th Latinized form Nortuagia. Hence, by the 9 ince of Västergötland, to a ‘hundred’ district 5. Cf. Stemshaug 1976: 130; century the name Norway was already in use. (härad). I would suggest that the transfer of Andersson 2000c: 558.

66 II. Geography the name Norway very probably originated seems to be no doubt that the older name with Danish travellers and merchants sailing for Huseby in Tjølling was Skíringssalr (or in these waters and visiting Kaupang. maybe Skírings-al, though, if so, to be seen in A name that is related both geographi- the same semantic fi eld as a name in -salr).10 cally and toponymically to Norway is Jæren, Th e fi rst element in this name is unclear, it the name of the large settlement district has been variously suggested that it is to be south of Stavanger. It is derived from ON interpreted as a by-name for a god, Freyr or jaðarr ‘edge’, an appropriate description for Ullr, as a name *Skíringr ‘the clear one’ for the sharp coast line between Eigersund and the inlet to Skíringssalr and Kaupang, and as Stavanger,6 and its name, the ‘edgy’ part, was a name of the actual salr ‘hall’, ‘the glorious, most certainly given to it by people sailing on “shining” hall building’.11 the ‘North way’, hence given from the per- Janet Bately, taking as her starting point spective of the sea. the Old English term healh, notes amongst Central to Ohthere’s description and the wide range of meanings found for this most certainly very central for early Viking place-name generic that of ‘slightly raised or Age Norway is Sciringes healh. Th is place low-lying land in close association with wa- name is in my opinion a corrupt form of ter’ (possibly ‘land beside an inlet or bay’),12 Skíringssalr, the second element -salr having and suggests that ‘Sciring’s haugh’ seems an been wrongly identifi ed with OE healh (cor- appropriate description of the site of the responding to modern Scottish and North- trading place and harbour now under exca- ern English ‘haugh’, southern ‘hale’) by an vation at Kaupang. However, she emphasises Anglo-Saxon note-taker or interpreter. Since that we can only conjecture as to what Norse the archaeological discoveries at Kaupang, word or phrase may have lain behind the 6. Sandnes & Stemshaug the Sciringes healh of the report of Ohthere’s choice of the topographical term Sciringes (eds) 1980: 177; Andersson voyages has been identifi ed with that place. healh, which the Old English text gives as the 2000a: 1. However, this is only partly correct. To un- name of the port (‘port’, ‘harbour’, ‘trading 7. For the occurrence of 13 Skíringssalr in early docu- derstand the names ON Skíringssalr and ON centre’) on Ohthere’s route to Hedeby. ments, cf. NG 6: 304 and Kaupangr one has to look at the actual settle- Th is is a possibility to be considered, but Skre this volume. ment historical situation here and the topo- it rests upon the assumption that this area 8. Brink 2000a. nymical milieu. had a name such as *Skíringr, *Skíringsangr or 9. Brink 1999; Brink 2000b. Th e site now known as Kaupang is situ- something like that. Unfortunately no such 10. Hoel 1986: 132 – For names ending in al in Scan- ated in a small settlement district (bygd), name is recorded. What we do know is that dinavian place names, see which in some early documents is identifi ed in medieval times a place called Skíringssalr Brink 1992. with the name Skíringssalr (an areal denota- was located in this area. Hence I fi nd as the 11. For interpretations of the tion which is probably secondary).7 Th e bygd most plausible interpretation that Old Norse element Skíring- in the name, see NG 6: 304-306, Fries contains only a handful of settlements. At its Skíringssalr and Old English Sciringes healh 1980, Sandnes & Stemshaug centre is a farm by the name of Huseby. Th is are to be seen in context, and I assume that 1980: 281, Hoel 1986, name goes back to an administrative term, the latter is a misunderstanding and an Old Nordberg 2003; cf. Brink the appellative husaby(r), which, at least in English adaptation of a misheard or misinter- 1996: 271-273; Andersson central Sweden, referred to a royal farm or preted Old Norse name, so that the Anglo- 2000b: 340. hamlet, probably part of the bona regalia, Saxon writer used the, to him well known, 12. See above, page 000 and 8 the bibliography there cited. the Uppsala ødh. As can be demonstrated, word OE healh, ‘bend (perhaps also bay)’, 13. See above, pages 000 and these administrative terms have replaced an hence Sciringes healh. Th is is the common 000. older name for the settlement.9 Very often interpretation of toponymic scholars who 14. NG 4: 304–6; Fries 14 1980: 94; Sandnes & Stem- these older names were of a special kind, have looked into the matter. Presumably shaug 1980: 281; Andersson theophoric or denoting a centre of power, Ohthere used this place name because it was 2000b: 339. such as Óðinssalr, Tésalr, Ærnavi etc. Th ere the ‘place of domination’, which the harbour

Geography, toponymy and political organisation 67 and trading place (the kaupangr) were under king or chieftain’. Maybe it is possible to re- the control of and to which they belonged. late the two sites to each other, by interpret- It is important to understand the fi rst el- ing them as belonging to the same historical ement, but for a cultural historical interpre- period, but to two diff erent socio-political tation of the settlement the second element ‘layers’ in society, where þjóða(r)lyng was the is vital. It is the word salr (or al, with prob- ‘stable’ geographical focus over time, the site ably similar meaning), an element found in where to meet, and Husaby(r)/Skíringssalr the some other prominent place names, such as site of a chieftain or a king, whose political (Gamla) Uppsala (< Upsalir), Óðinssalr, Tésalr power could grow or diminish, so that the etc., a word that can be demonstrated to have site could lose or gain ‘importance’, perhaps, denoted a major banqueting hall, a king’s or in some cases, to the degree that it could a chieftain’s hall;15 the old Scandinavian word ‘vanish’ from the socio-political scene, and for a ‘hall’ was salr, cf. OE sele in the poem sink down to become an ordinary farm.18 Beowulf. In a paper from the mid 1990s I Circa 2 km south of Huseby/Skíringssalr tried to ‘provoke’ Norwegian archaeologists we have the famous settlement Kaupang if not to stop excavating in Kaupang, at least (ON Kaupangr). Th is name contains an ap- to ‘put the shovel in’ in Huseby, which is the pellative ON kaupangr, OSw køpunger, ODa most ‘interesting’ site here in Tjølling. Th is køping denoting a trading and market place, has now been done under the new leadership obviously for an organised and controlled of Dagfi nn Skre, Oslo, and here at Huseby, market; several of these early markets today on an elevated plateau, they have found have the name Köping(e), Kaupang, Købing, the site of a large hall building, the actual X-köping(e) etc.19 In Ohthere’s description we Skíringssalr or a successor or predecessor of read of “an port …þone man hæt Sciringes that hall. We thus today have evidence of a heal”, i.e. a harbour, trading place which is large hall both in (Gamla) Uppsala as well as called Sciringes healh. In the light of what has in Huseby/Skíringssalr. It would be fascinat- been discussed above, this is in my opinion ing also to start looking for a hall at sites such to be understood as “the harbour and trad- as Óðinssalr and Tésalr in Østfold. ing place ‘belonging to’ or ‘under the control A couple of kilometres to the north is of’ Skíringssalr”. When shortly afterwards the parish church of this bygd, in the parish Ohthere refers again to Sciringes healh as the called Tjølling. Th is name goes back to an place he visited on his voyages, he does not ON þjóða(r)lyng, which is to be translated as of course use the name Kaupangr, because ‘the peoples’ heath’, and most certainly to be there were several ‘kaupangs’ along his travel- understood as ‘the heath where people gather ling route in Scandinavia. It is hence, in my for Th ings and other assemblies’, hence the opinion, wrong to assume that Kaupang had name of an assembly site to be seen in the the older name Skíringssalr. Th at it was an same semantic fi eld as Th ingvellir etc.16 Many important place (“port”) has been confi rmed of the northern Swedish and Norwegian par- by the archaeological excavations here during ish names have this background, as the name the last four or fi ve decades.20 Th ese excava- 15. Brink 1996: 255-58; for an assembly site for a bygd, a district or a tions reveal the most important harbour and Meulengracht Sørensen 17 2003. congregation. It is not so easy to understand trading place in Norway during the Viking 16. Cf. Brink 2004: 64-66. the links between Tjølling and Huseby. Do Age, and this site is to be seen in context with 17. Cf. Brink 1990. they belong to the same chronological phase, the nearby centre of power, which probably 18. Cf. Brink 1999. or to diff erent historical periods? Th e seman- owned the harbour and controlled the trade, 19. See Ljunggren 1937, Ljunggren 1965, Pelijeff tic content of the former is ‘site where peo- namely Skíringssalr/Huseby. 1988, Ersgård 1988. ple meet for assemblies etc.’, while the latter All this discussion of Kaupang, Huseby, 20. See Skre this volume: breathes power and authority, ‘the hall of the Skíringssalr etc. leads to the conclusion that 000.

68 II. Geography one has to look at the settlements in their Sweon). We have absolutely no knowledge of landscape context, not just at individual the political situation in Sweden (the early sites. Any attempt to understand them in Sviþjóð) around 900.23 We believe that at isolation creates a problematic stance for in- that time Sviþjóð was a polity around Lake terpretations. Instead settlements have to be Mälaren in central Sweden, with the prov- seen in relation to each other, in a settlement ince of Uppland at its core, and the provinces complex. Typically for early Scandinavia it of Södermanland (“the land belonging to the is the settlement district (bygd) that was the people living to the south [of Uppland]”), important corner stone in society, something Västmanland (“the land belonging to the that is very well illustrated here in Tjølling people living to the west [of Uppland]”) and parish. perhaps also including the province of Närke Th e homeland of Ohthere was Háloga- (in the west). When the larger Sweden was land (mod. form Hålogaland). Th e fi rst writ- formed, including in particular the provinces ten evidence of this name we fi nd in his of Öster- and Västergötland (“the land of the account as Halgoland, the fi rst element of gautar living in the east” and “the land of the which has been assumed to have been wrong- gautar living in the west” respectively), is a ly identifi ed with the word for ‘holy’,21 prob- very much debated problem, but maybe this ably by the Anglo-Saxon writer, in very much is to be placed as late as in the late Viking Age the same way as I believe Skíringssalr to have or early Middle Ages, hence after the time of been misinterpreted as Sciringes healh and Wulfstan. An interesting reminiscence of this Bleking as Blecinga eg. In this province lived integration of the provinces into the “new” the people known as the háleygir (cf. Háley- Sweden is probably the so called Eriksgata, gjatal), and the name means ’the land (prov- the journey a newly appointed king among ince) of the háleygir’, the fi rst element here the Swedes had to undertake to each of the being an older form, genitive plural háloga-. above mentioned provinces, to attend their No evident convincing etymology of háleygir provincial assemblies and there to be anoint- has yet been proposed;22 in my opinion the ed and accepted as their new king. fi rst part is probably to be understood as the In the light of this it is most peculiar to ON adj. hár ’high’, denoting the remote situ- fi nd Wulfstan in the late 9th century declar- ation of Hálogaland in the north, but -leygi- ing that Blekinge, Möre, Öland and Gotland /-loga- is more problematic. “hyrað” to the Svear. Does this mean that Th e account of Wulfstan’s voyage con- these provinces were part of a kingdom of tains a description of the route that he took Svear (Sviþjóð)? Probably not. I assume that between the two trading places Hedeby and given the wide range of reference of the col- Truso. Th e land he passes on his port side location hyran to,24 the provinces are being 21. Andersson 1999: 455; Th is assumption may be belongs mostly to the Danes and that on his described as part of the Svear’s sphere of in- questioned on linguistic starboard side to the Wends. It is notable fl uence - and not that of the Danes; that this grounds, see Translation that Bornholm (Burgenda land) is said to be area was part of the cultural region where notes. a “land” of its own, with its own king. Th e the Svear dominated, including in respect of 22. Sandnes & Stemshaug intriguing part of this geographical descrip- trade, or even that these provinces were sub- 1980: 169; Andersson 1999: 454-5. See also Storli this tion is when he enumerates the “lands” on ject to the Svear. volume: 000. the port side after Bornholm, namely Blec- We know that toponymically and on the 23. For a plausible political inga eg, Meore, Eowland and Gotland, easily evidence of the dialects, the eastern part of interpretation see Sawyer recognisable as the provinces Blekinge and present-day Sweden, down to and includ- this volume: 000. 24. For the ambiguous Möre, and the large islands of Öland and ing Möre (but not Blekinge), is and has been expression hyran to see Bately Gotland, and says that they belong to or were very much infl uenced by the Mälar region. this volume: 000. the subjects to the Svear (þas land hyrað to Th e dialects are so closely related that the

Geography, toponymy and political organisation 69 ‘languages’ down to Möre are placed in the trading place Truso.30 Utlängan, the eastern- large group of “Sveamål”, i.e. dialects related and southernmost island, must have been to Swedish (spoken around Lake Mälaren),25 well-known to all sailors along the coast, and and the toponyms down to Möre have a no- this is evidenced by a rune stone, standing in table affi liation to the place-name structure Ny Larsker on Bornholm, with the inscrip- around Lake Mälaren. We do not know how tion: kobu suain raisti stain þina a[f]tir old this infl uence is, but judging from the bausa sun sin tr[i]n þan is tribin ua[r]þ i evidence of several of the place names, this [ur]ostu at ut la[nk]iu.., i.e. “-- Sven erected infl uence from central Sweden seems to have this stone after Bøsi, his son --, the one who been active during the late Iron Age. was killed in battle at Utlängan... ”. In oth- Gotland has always stood somewhat er words, at the time of the erection of the apart, but dialect and place-name evidence rune stone a young man from Bornholm was indicate an infl uence from east central Swe- killed in a battle (at sea) off Utlängan. Th e den.26 According to the Guta Saga, Gotland cultural and trading activities are obviously agreed during the late Iron Age to become a to be tied in with the eastern part of Ble- tributary of the Svear, paying each year a trib- kinge, around the islands mentioned above ute to the Svear king.27 Öland shows greater and on the mainland. It is probably to this infl uence from the north, with the place area that we can locate the name Blekinge, name Svibo (from Svea-boþ) on the southern which seems to go back to an old bay name part of the island as an intriguing piece of *Bleking or *Blek, ‘the glittering, the calm’, evidence. Möre, on the mainland, was an old on, as has been suggested, the bay Ham- ‘land’ of its own with a central place at Hoss- marbyviken.31 It is unique that a name for a mo (*Husar), which is some kind of predeces- bay has been given to a whole province, and sor to the town of Kalmar.28 Another ancient this is probably to be explained in the light ‘land’ to the north, mentioned by Jordanes, of the importance of the ancient sailing route is Tjust (Th eustes). Th e dialects and topony- here. Th e name form given to Blekinge in the my of both these ‘lands’ reveal the infl uence account of Wulfstan’s voyage is Blecinga eg, of, and contacts with, central Sweden. like Sconeg (< Skán-ey) a name with second Th e greatest problem is provided by Ble- element Old English íeg, êg, îg ‘island’, Old kinge. It seems obvious that this province has Norse ey, Old Swedish ø (< øy), that is to say played an important role in seafaring and Blekinga ey “the island of the people of Ble- trade, both from the topographical situa- kinge”, denoting one (Utlängan) or several tion on the ‘corner’ between the north- and of the islands mentioned above, rather than east-going sailing routes and the one south- a misinterpretation by the writer, which has bound to Denmark and Hedeby, where the been a common explanation. 25. Hesselman 1905; Fries strategically placed islands of Senoren, Has- We have already touched upon the po- 1962. th slö, Aspö, Tjurkö, Sturkö and Utlängan must litical situation around the late 9 century. 26. Hellberg 1988. 29 have been of particular importance. Th is is Th e anonymous translation of Orosius’ His- 27. For a discussion indicated by e.g. the famous gold bracteate toriae begins with a geographical overview, regarding Gotland and its from Tjurkö, by the fact that several place which the author has expanded with a de- connection to the Svear, see Blomkvist, N. 1995: 222-4 names along the coastline bear witness of scription of northern Europe, and it is in this and Blomkvist, T. 2002. the presence at some time of Frisians, Danes, expanded part that we fi nd the insertion of 28. Hellberg 1979. Jutes (from Jutland), Gutar (from Gotland) the accounts of the reports of Ohthere and 29. For the importance of and Est(onian)s, and also by the form taken Wulfstan, almost as “eyewitnesses” for the ge- the Utlängan area and these south-eastern islands, see by two eponyms in the form of Heaby(holm) ography in the periphery. Ohthere describes Stenholm 1995. from Hedeby and Trusö, which actually seems how several ethnic groups interact in the far 30. Ohlsson 1939: 27. to be derived from the much more famous north, and how the háleygir’s main income 31. Ohlsson 1939.

70 II. Geography was the tribute the Sami (Finnas) paid them there by the Danish king, or perhaps he had in fur, feather, whale bone and skin from accepted the Danish king as his overlord. But whale and seal etc. Why the Sami paid this Ohthere does not say anything about who tribute, we do not know, but an assumption ‘ruled’ in Skíringssalr, though he does so for may be that the háleygir stood on a higher Hedeby. May we take it from this that the societal level, with some kind of organised ‘lord’ at Skíringssalr was not under Danish power, with chieftains and retinues etc., and overlordship? Peter and Birgit Sawyer have therefore could demand these kind of trib- produced the plausible theory that Norwe- utes.32 gian rulers and chieftains during this period Ohthere describes the “land” of the probably had accepted Danish overlordship northmen as long and narrow, which seems “in order to have safe passage through Dan- accurate and a good description of Norway. ish waters ... and access to Danish markets”.34 From this, we may assume that the area form- It seems plausible to suppose that the Danes ing what is present day Norway was looked had some political, economic and cultural upon, certainly not as a united kingdom, but dominance over Viken in particular in this in any case as the “land” where the northmen period.35 lived, so there must have been a feeling of When Ohthere goes on to describe the some kind of unity among the diff erent re- voyage from Sciringes healh, he states that the gions, Hálogaland, Þrændalög, Møre, Hörða- land on his port side was Denmark (on þæt land, Jaðarr, Agðir etc., already in the late bæcbord Denamearc). Such a statement fi ts in 9th century, at least according to Ohthere. If well with the early medieval political situa- Opland and Hedmark could also be included tion with regard to Skåne and Halland, both is uncertain. It is also interesting to note that old Danish lands. But what about Bohuslän Ohthere states that to the east of the moun- (the old Ranaríki)? Was this part of the coast tains (moras) we have Sweden or the land of also in the hands of the Danes? Perhaps. It the Svear (Sweoland). Th is statement is some- would be no surprise if all the coastal land up what bewildering. It has been assumed that to Viken was under Danish control or infl u- Ohthere simply means that Sweden is to be ence in the last quarter of the 9th century.36 found to the east (hence quite a long distant Wulfstan states that Blekinge, Möre, away, probably around Lake Mälaren),33 but Öland and Gotland belonged to or were sub- it may also suggest that “Sweden” by this ject to the Svear (þas land hyrað to Sweon).37 time had a larger extension. Is this then to As we have seen above, it seems possible to be interpreted as indicating that the Svear al- accept this statement with reference to Möre, ready had a fi rm infl uential grip on provinces Öland and Gotland at this time, on the like Värmland, Dalarna, Jämtland and Häls- grounds of toponymy, dialects and the tra- 32. Sawyer, B. & Sawyer, P. ingland, where the fi rst three today all border ditions of Gotland as being a tributary land 1993: 145. on the mountains and on Norway (and so to the Svear, but Blekinge is more diffi cult 33. Th is is the position of e.g. did Hälsingland before 1645)? to understand in such a context. However, Gahrn 1988: 43. It would have been extremely interesting bearing in mind what Ohthere said above 34. Sawyer, B. & Sawyer, if Ohthere had said something of the political about Sweoland being on the other side of the P. 1993: 88; cf. Roesdahl 2001a: 129. situation down at Skíringssalr, but nothing is mountains on the Norwegian border, and 35. On this matter cf. Stanley mentioned apart from the “port” there. If he Wulfstan’s statement that all “land” along this volume: 000. had done so, Ohthere might well have told the eastern coast of present-day Sweden be- 36. Cf. Sawyer 2002: 22 and us of an important chieftain or ‘king’ resid- longed to the Svear, it is perhaps possible to Sawyer this volume. ing in Skíringssalr (but certainly not perma- look upon the Svear in the same way as, for 37. See Text and translation. 38 38. Hyenstrand 1989: 11-30; nently living there), controlling the trade at instance, Åke Hyenstrand has done, as a cf. Widmark 2004: 740-46. the kaupangr. Maybe the person was placed kind of trade organisation, but with the ex-

Geography, toponymy and political organisation 71 tra ingredient of Svear probably forcing dif- sumably later Heðaby (> Hedeby), seems to ferent “lands” to pay tribute to them, hence be the name of this port and market used by giving the Svear a more political off ensive di- the Scandinavians, hence the people of the mension, in the way that, for instance, Lars north, whereas it has been assumed that the Hellberg has advocated.39 From Ohthere’s Saxons and the people to the south named and Wulfstan’s statements it seems possible the place Sleswig (< Sliaswic). Th is assump- to assume that already around the late 9th tion is based on a statement by the Anglo- century, provinces like Värmland, Dalarna Saxon writer, Æthelweard, who says of Hede- and Jämtland were under the infl uential by, that “in the Saxon language [it] is called rule of the Svear, perhaps paying tribute, as Slesuuic, but in Danish Haithaby”.41 Peter the Sami did to the Háleygir, and the Gutar Sawyer has, however, suggested that we are to the Svear. Th is could be the background dealing with two diff erent places, and that to Ohthere’s statement that the land of the Æthelweard has misinterpreted his source, a Svear came after the bordering mountains. It passage in Rimbert’s Vita Anskari.42 From a is, however, noticeable that in, for instance, Scandinavian toponymic perspective we can the Historia Norwegie (ca 1160-1175) it is see that the latter form, Hedeby, must be fairly stated that to the east of Norway are Sweden, young. Names in -by with a fi rst element in Götaland, Ångermanland and Jämtland (sed the genitive - often another place name - are de sole Swethiam, Gautoniam, Angariam, Iam- normally medieval or at the earliest Viking toniam),40 which implies that in the 12th cen- Age, such as Visby on Gotland (‘the port and tury diff erent peripheral provinces still had market town beside the vi/Vi’) and Hedeby a somewhat independent status vis à vis the (‘the port and market town on the heaths’). Svear and Svíþjóð. In a similar way the Svear In these last two cases we must also assume dominated all the land on the eastern coast, that the second element -by is not the ‘nor- together with the Baltic islands, down to Ble- mal’ -by element found in hundreds of farm kinge. Perhaps the eastern part of present-day and hamlet names, but the Low-German Blekinge was some kind of bordering area or loanword by ‘town, trade and market place’, a res nulla between the Danes and the Svear. found nowadays in Danish and Norwegian Th e political status of the small but old “land” by ‘town’. of Lister is diffi cult to understand. Perhaps it Ohthere explicitly states that Hedeby, al- was an “independent” ”land” of its own, as it though situated between the Wends, Saxons had clearly been a couple of centuries earlier, and Angles, belonged to the Danes (se stent according to some important rune stones, in betuh Winedum 7 Seaxum 7 Angle 7 hyrð in on the same way as Bornholm obviously was at Dene). Th e politico-historical interpretations this time. by historians, onomasiologists, runologists Th e biggest problem, as regards po- etc. regarding Hedeby has - as is well known - litical situation and affi liations, is however been changing, with the early assumption of provided by Hedeby. Th e name of the port the existence of a Swedish overlordship over and market place in the 9th century is said in the town, and later a rejection of this Swed- Ohthere’s report to be æt Hæðum. i.e. ‘at’ or ish dominance of Hedeby. We have evidence ‘on the heaths.’ Th is place name makes good that the languages spoken in Hedeby during 39. See e.g. Hellberg 1979. sense; a heath is a suitable site for a meet- the Viking Age must have been a mixture, 40. Historia Norwegie, Ekrem ing place or a market, and such a name is to with several multilingual persons. Th e Nor- & Boje Mortensen 2003: be seen in the same toponymical light as for dic spoken here was obviously infl uenced by 52–3; for the dating of this text, see Boje Mortensen example the previously mentioned Tjølling, a other languages, acquiring new items of vo- 2002: 11–24. compound also denoting a heath, where peo- cabulary and perhaps also undergoing chang- 41. Campbell 1962: 9. ple assembled. Th e name æt Hæðum, and pre- es in syntax, linguistic traces that later spread 42. Sawyer 2002: 72.

72 II. Geography from Hedeby to the north, into the diff erent along the Swedish east coast, in parts of Den- dialects of Scandinavia. Th is language change mark as well as in the place-names Haddebyer and this Nordic language in Hedeby were fi rst Noor and Selker Noor.46 Here we have a clas- identifi ed by Bengt Hesselman as Birkasven- sical example of the hen-and-egg problem. ska (Birka Swedish), later by Gun Widmark Th e latter meaning is - as etymology shows as Hedebynordiska (Hedeby Nordic).43 us - secondary to the former. But where was In the case of Hedeby, archaeological the new meaning developed, and which area evidence, written records and annals have infl uenced which? It is new evidence of this all been consulted, as also the famous rune kind, which may fuel a new debate on the stones in the vicinity, in order to provide the discussion of the historical implications of foundation for such interpretations. But one Hedeby. Furthermore, in Schleswig, not must also take the place-names in Schleswig far from Hedeby, we have the place-name into account. Th e task of interpreting does Schwesing. Th is name is obviously to be un- not become easier with this material. On derstood as an older Svea-husar, dative Svea- the contrary the picture becomes wider and husum. We can identify the fi rst element as perhaps also more complex. In the vicinity Svear and the second element as the word of Hedeby, for instance, we have two place- husa(r), which we fi nd in particular along the names ending in nor, normally to be trans- eastern Swedish coast, in Möre and Tjust and lated with ‘narrow inlet, short river’,44 but around Lake Mälaren. How this name and found in eastern Sweden and parts of Den- other important place names here are to be mark with a secondary meaning ‘small lake interpreted and historically understood, we or lake-like inlet by the coast, suitable for a do not know - sorry to say. Here we have an harbour’.45 Th is secondary meaning is found interesting task for the future.47

43. Hesselman 1936, Wid- mark 1994. 44. Strid 1981: 63; Kousgård Sørensen 1985: 44. 45. Hellberg 1986. 46. See Strid 1981: 62. John Kousgård Sørensen (1985) states that the older meaning ‘narrow, inlet, watercourse’ is to be found in some ancient Danish place-names like Nors and Norring, but that the normal meaning in place names and in dialects in Denmark is ‘lake, which has a connection with the sea or a (larger) fj ord through a nar- row inlet’, a meaning, which, according to Kousgård Sørensen, is not be found in e.g. Sweden. Th is is not entirely correct, since there are several nor-place-names with that secondary meaning in Sweden. 47. I am most grateful to Janet Bately for useful com- ments on this text.

Geography, toponymy and political organisation 73