<<

Chronology and Typology of the Danish Runic Inscriptions

Marie Stoklund

Since c. 1980 a number of important new archaeological runic finds from the old Danish area have been made. Together with revised datings, based for instance on dendrochronology or 14c-analysis, recent historical as well as archaeological research, these have lead to new results, which have made it evident that the chronology and typology of the Danish rune material needed adjustment. It has been my aim here to sketch the most important changes and consequences of this new chronology compared with the earlier absolute and relative ones. It might look like hubris to try to outline the chronology of the Danish runic inscriptions for a period of nearly 1,500 years, especially since in recent years the lack of a cogent distinction between absolute and relative chronology in runological datings has been criticized so severely that one might ask if it is possible within a sufficiently wide framework to establish a trustworthy chro- nology of runic inscriptions at all. However, in my opinion it is possible to outline a chronology on an interdisciplinary basis, founded on valid non-runo- logical, external datings, combined with reliable linguistic and typological cri- teria deduced from the inscriptions, even though there will always be a risk of arguing in a circle.

Danmarks runeindskrifter

A natural point of departure for such a project consists in the important attempt made in Danmarks runeindskrifter (DR) to set up an outline of an overall chro- nology of the Danish runic inscriptions. The article by Lis Jacobsen, Tidsfæst- else og typologi (DR:1013–1042 cf. Introduction. The in in the English Summary 1947:7–30), gives a clear account of the basis and method

355 used. A sharp distinction was drawn between absolute and relative dating after a radical adjustment had been made of the “historical inscriptions” which con- stituted the framework of the rather narrow datings in the earlier Danish corpus- edition by Ludvig F.A.Wimmer (DRM). On the basis of the radical principles of historical source criticism, Lis Jacobsen in the 1930s had criticized Wimmer’s chronology severely. In DR only three “historical” inscriptions were accepted from the and three from the Middle Ages, apart from Sven Estrid- son’s runic coins. With regard to the earliest inscriptions, with reference to Jo- hannes Brøndsted – the leading Danish archeologist at that time – the archae- ological datings of the bog finds and the were seen as reliable factors. The datings of Medieval inscriptions from churches and on church equipment were based on the relative datings of Poul Nørlund and Mouritz Mackeprang. Thus, according to DR “On the basis of all the datable inscriptions, the runic inscriptions of Denmark may be distributed over 4 periods: the Migration Period (and the time subsequent to the Migrations), The Viking Age, the Pre-Medieval Period, and the Middle Ages.” (Summary:11f.). This system – with further sub- divisions – was used for a relative classification and combined with a descrip- tion of the significant features of the individual periods of archaeological, orna- mental and linguistic character. Runographical and linguistic peculiarities as well as spirit and contents, and the general appearance of runic stones were in DR used to distinguish between the many sub-divisions of the Viking Age in the chronological system, i. e. Period 2.1 Helnæs-Gørlev type (“ascribed to the age from about 800 (or 750) to about 900.”) (Summary 1947:16); 2.2a. The pre- Jelling type (9th century); 2.2b. The Jelling type (10th century); 2.2c. The Post- Jelling type (c. 1000–c. 1050). Subsequently, based on the various diagnostic features and the detailed description of the characteristic features of each period (DR:1018-1034), it was possible when considering the individual inscriptions to place the inscription in question in the proper context simply with a reference such as Per. 2.1, Per. 2.2a. This relative and absolute chronology of DR was much admired and, inci- dentally, considered to be rather cautious. It has had a strong interdisciplinary impact and the datings are still used by many scholars, although it at least since the 1970s has been evident that important adjustments of the absolute datings (and the time limits of the periods) were necessary (for instance Christensen &

356 Marie Stoklund Moltke 1971; Nielsen 1970; cf. Stoklund 1991:289–294 with references). This improved chronology we find for instance in the works of Erik Moltke (Moltke 1976 and 1985 with references). Moltke also rejected the DR-system with four periods and chose a three-period-system: Primitive Norse – Viking-Age – Medi- eval inscriptions (operating with a rather strong overlapping between the peri- ods), while for instance Niels Åge Nielsen stuck to the old system (Nielsen 1983). Moltke’s absolute datings in 1976 were, however, in fairly good accor- dance with the DR-chronology, apart from the bracteates, the Jelling-stones (DR 41–42), Haddeby 1 and 3 (DR 1 and 3) (and Bække-Læborg, DR 29 and 26). In general, the fundamental principles behind the DR-system were upheld by Moltke. Lis Jacobsen stressed that the indskrift-fortegnelse (list of inscriptions) (DR:1035–1042) gave no grounds for a safe dating of the individual inscription. An inscription which typologically belonged to one period might very well be contemporary with another, which belonged to an earlier or later period (DR:1035). There has, however, been a problematic tendency to use the temp- ting possibilities which her chronology offered to take the dated, “annotated” list of inscriptions as absolute datings. It is, however, problematic, on the basis of the DR-chronology, to draw far-reaching conclusions involving a time per- spective as has been done (cf. Stoklund 1991:295f.). Nevertheless, the diag- nostic features, runographically, linguistically etc. on which the DR-chronology was established, could in practice within wider frames be useful tools for a determination of the approximate age and context of an individual inscription. In order to make the material easily available in this survey of the Danish runic chronology, I have chosen to present selected examples of important new finds and crucial altered datings in tables 1–3, listing the inscriptions chrono- logically. During my work with the material I have primarily distinguished be- tween only three groups of inscriptions: Proto-Scandinavian (24-character inscriptions), Viking-Age (16 characters), Medieval (extended runic alphabet). Although only a few inscriptions cannot be classified within one of the three main groups, the transitions between the periods cause problems. I have, however, chosen a delimitation of the material in three sections: before 800, 800–1050, after 1050–, according to the traditional historical Danish datings of the beginning of the Viking and Medieval Ages, regardless of the question when

Chronology and Typology of the Danish Runic Inscriptions 357 the transitions exactly happened and how they should be defined, though I have tried to note the inscriptions with significant changes and innovations, especially in the periods of transition. The general reservations and precautions concerning relative versus absolute chronology, archaeological datings (cf. for instance Steuer 1998) as well as runological ones, which have been the subject of much discussion (for instance Hagland 1998) are presupposed as natural and generally known preconditions which do not need to be repeated in a discussion of the individual cases. In all the tables the first columns give, in addition to the name of the inscription, significant features – no matter whether typological, grapho- phonological or linguistic. In the following columns earlier (runological) datings and in the last column the present (non-runological) datings can be found. As regards the inscriptions with the older runes the dating is mainly based on archaeological considerations.

Table 1, AD 1–c. AD 800 Archaeologically dated inscriptions from Denmark. Changed datings, new finds 1) Earlier datings DR=Danmarks runeindskrifter. KJ=Krause & Jankuhn 1966. 2) Later datings, A=Antonsen, G=Grønvik, M=Moltke. 3) Latest (Danish) datings, 2000 [2005]

1) 2) 3) Inscription (Found or first DR; KJ Later datings Datings 2000 mentioned) [2005] Vimose comb (1865) DR 3rd–4th C.; A 250 B2/C1a; c. 160 harja KJ c. 250 Vimose chape (1901), DR chape 250–300 A chape 250–300; C1b; buckle (1851), buckle 200–400, buckle 200; 210/20–250/60 sheathplate (1853) sheathplate 200– sheathplate sheathplate: runelike inscription, 300; = KJ c. 250, c. 200–300, aw0ings runelike inscription lancehead (1984) wagnijo 200 and c. 400? Vimose woodplane (1865) DR c. 100–300, KJ A 100–300 c. 160–375, most 250–300 likely C1b; 210/20–250/60

358 Marie Stoklund Illerup (1982–92), ™, c. 200 [2005: C1b; mirror-runes òÃ versus 210/20–250/60] W, ìí wagnijo, niþijo tawide, laguþewa, swarta, gauþR Thorsberg (1858, 1860) DR chape 200–250; A 200 C1b; owlþuþewaR, mariR shield boss 2.–5. C. 210/20–250/60 KJ c. 200 Næsbjerg clasp (1949) A 200 warawnis C1b; -ara!fn!is- 210/20–250/60 Nøvling (Lundegårde) A 200 C1b; clasp (1963) bidawarijaR 210/20–250/60 talgidai

Gårdlösa clasp (1949) ek A 200 C1b; unwod"R 210/20–250/60 Værløse clasp (1944) A 200 C1b; alugod 210/20–250/60 Skovgårde/Udby clasp c. 200 [2005: C1b; (1988) lamo : talgida 210/20–250/60] Himlingøje clasp 2 (1949) KJ c. 200 A 200 C1b; ºwiduhudaR 210/20–250/60 Møllegårdsmarken knife C1b–C2; (1992) hth sh"ko 210/20–310/20 Himlingøje clasp 1 (1835) DR start of 4th C.; 300–400 C1b-C2; hariso KJ c. 350 210/30-250/60 [2005: C1b/C2; c. 250] Nydam arrow-shafts DR 250–500. c. 300–350 (1863, 1993, 1994), lua, KJ c. 400 [2005: C1b–C3; la, lua (3th–5th C.) 210/20–375] Nydam axe-shaft (1993) c. 300–350 wagagastiR | : [2005: C1b–C3; -"-hg"usikijaR : aiþalataR 210/20–375]

Chronology and Typology of the Danish Runic Inscriptions 359 Nydam strap suspension c. 300 [2005: C3; (bronze) (1995, 2002) c. 310/20] }harkilaR * ahti | anul"a Nydam sword bead (jet) c. 400 [2005: D1, (1997) "--ala early, c. 380–420] Nydam belt end fitting c. 400 [2005: D1, (silver) (1997) ¬ rawsijo early, c. 380–420] Nydam lance shaft 1 c. 400– (1996) hardly linguistic [2005: D1, late; sense c. 450–475] Nydam lance shafts 2, 3 c. 425–475 [2005: (1999) runelike decoration D1, late; 450–475] Kragehul spear shaft DR c. 350–550, KJ. A 300–550; 525 c. 425–475 © (1877), E, S, , s, ö; c. 500–550 [2005: D1 early; bind-rune variants 400–450] Kragehul knife haft (1865) DR 500–550 A 300 c. 425–475 [2005: D1 early; 400–450] Bracteates DR 400–650 G 450–550; c. 500; D2a; š j, bind-rune variants M c. 500 460/70–530/40 Kalmargården frag- Migration Period; ment (1995), ...uAlis hs (630)650–700 Lousgård bead (1955) 7th C. Migration Period, sHilta 7th C. Ålborg buckle c. 700–750 (Scheelsminde) (1964) s s Skabersjö buckle (1855) DR buckle c. 700, c. 700, 725–800. c. 720 (Birkmann a, n, s Inscription: Jelling M: Inscription 1995:90f. with Period? c. 1025. references) Ribe cranium fragment c. 800, rune 59 c. 725, rune 59 n n (1973) H, M, h, k, s G or U

360 Marie Stoklund The older runes

The increase of inscriptions with the older runes from archaeological exca- vations is most significant. In the survey of 1942 (DR:1042) apart from brac- teates only 19 inscriptions were known from the period 1.1 (the so-called Migration Period in DR from AD c. 200 to c. 650). In accordance with the chro- nology of Johannes Brøndsted, the runic bracteates were dated comparatively late within a rather long supposed -period from 400–650 (DR:791). Now 37–38 objects are known with runes or rune-like inscriptions, which archaeologically have been dated to the period from AD c. 160 to 500 (exclu- ding bracteates and the problematic inscription on the Meldorf fibula from the first century AD from Ditmarschen south of the river Eider). The majority of the new inscriptions belong to the Roman Iron Age, but some of the Nydam finds are later, and from a chronological-typological point of view they are of special interest. The notable finds from Nydam belong to dif- ferent deposits in the former lake and could thus offer some relevant information concerning typological development but it should not be forgotten that they (like the other bog-sacrificial finds) probably originated somewhere outside the later Danish area. The oldest group of inscriptions all have ™ as e-rune and s-runes with several angles. An astonishing tendency towards cutting rounded forms even on wood puts a question mark to the general assumption that runes were “born” with straight lines. The new finds confirm the existence of full-size vari- ants of j, o already in the early Illerup inscriptions, while the assumption that the inscriptions before c. 400 never have word division (for instance Krause & Jankuhn 1966:3) does not seem to be valid. There are division marks in the in- scriptions of the Skovgårde (Udby) fibula, the Nydam axeshaft and strap suspen- sion (Stoklund 1995:322, 342; 1996b:275f., 287f.). When focusing on the Danish evidence from the late “Germanic Iron Age” c. 400 and onwards, significant rune forms could be found on the Nydam belt end fitting c. 400). It is noteworthy that according to Nydam parallels (and den- drodating of Nydam wood) the Kragehul (KJ 27) inscription with important diagnostic runes is now (2005) dated by the archaeologists to the 5th century, differing from the wider margin earlier given. In the course of time somewhat changing opinions have caused notable confusion, the Kragehul inscription for

Chronology and Typology of the Danish Runic Inscriptions 361 instance according to Antonsen dates from c. 300 (Antonsen 1975:35); and to Odenstedt c. 500–550 (Odenstedt 1990:71; cf. Krause 1971:152). The considerable graphic variation through the years which becomes evident from a study of the new inscriptions is, no doubt, an important feature with re- gard to the possibilities for establishing a relative chronology on a typological basis, even though, according to for instance Elmer Antonsen, “for the oldest pe- riod of runic inscriptions we have no guideposts whatever for arranging the vari- ous inscriptions in any kind of fixed chronological progression based on runo- logical or linguistic features” (Antonsen 1998:159). One of the traditional typo- logically founded dating-criteria is the supposed development of the j-rune, which is beautifully confirmed by the Nydam finds (cf. for instance Krause & Jankuhn 1966:2; as opposed to Antonsen 1998:151f.; compare Odenstedt 1990:67–74). It should be noted that no new mirror-runes have appeared since Illerup, ex- cept perhaps among the pseudo-runes of the second of the three – in my opinion non-semantic lance- or spear-shaft inscriptions from Nydam, which on the other hand are also related to the Kragehul and Lindholmen runes, which – partly at least – make linguistic sense. These new Nydam lance-shafts, however, also seem to be related to such problematic inscriptions as Ødemotland, Britsum. Perhaps a tendency around 500 towards a growing use of pseudo-runes and rune-like decorations should be noted, cf. also some inscriptions on bracteates which are not readable in our literate sense, although attempts at interpretation of such inscriptions as Ågedal and Ødemotland have recently appeared (Grønvik 1987:61–91; Stoklund 1994:165f.; Grønvik 1996:255–267). The linguistic position and possible development of the Older Runic lan- guage have been a matter for much discussion. The new inscriptions mean an important increase of relevant linguistic material (see Appendix: Proto-Norse index) concerning inter alia the question of the relations between spoken lan- guage and orthography and the determination of the oldest runic language as be- longing to the North-Germanic or Proto-Scandinavian, “urnordiske” group be- cause of the early examples of men’s names in -o as well as -a. The assumption (for instance Nielsen 1998:545) that the old n-stem ending -o as a nom.sg.masc. marker was ousted some time after 200 AD as the new -a marker gained the upper hand in nom.sg.masc., while -o was restricted to denoting only the

362 Marie Stoklund nom.sg.fem., is contradicted by the – according to the late archaeological date (c. 400) – remarkable rawsijo on the Nydam belt fitting found 1997. It is no doubt a man’s name, but this fitting belonged to the besieged enemy – maybe from the south –, whose gear was sacrificed in the former inland lake of Nydam (Stoklund 1998b:259f., 266f.). A clear case where provenance, regionality ver- sus chronology must be taken into consideration. A remarkable variation in the endings of the weak verbs in the 3 pt.sg.ind. according to the general opinion ought also to be accepted because of the examples of -e : -a : -ai that have turned up. Problematic, however, is a 3 pt.sg.ind. ahti in the inscription: }harkilaR * ahti and anul"a on a bronze strap suspension from a beautiful scabbard of leather and wood (c. 300, Stoklund 2003:176). An interpretation as “H. owned” is doubtful for various reasons (Stoklund 1998a:62f.). Already in the 1950s diverging bracteate datings gave rise to important discussions of the runic chronology (cf. Moltke 1956:6–8; Nielsen 1970:28f.) and led to differing datings among the various runologists and linguists. There are important differences between the chronological systems of for instance Krause, Elmer Antonsen, Ottar Grønvik and Seebold. According to the latest research (computer seriation and correspondence analysis of the large human heads of the A–C bracteates) by Morten Axboe (Axboe 1999), the bracteates date from the period between 450/475–525/560 and no simple typological A–B–C-development can be assumed. The former dating of the late bracteates to the first half of the 7th century is impossible (Axboe 1998:231f., 236).

The language. Late urnordisk

According to DR and Wolfgang Krause the language of the bracteates still represented the “classical” urnordisk, even though the runic bracteates were dated rather late: c. 400–650 (DR:791, 1018) and “Ende des 5. bis zum Ende des 6. Jh.s” (Krause 1971:17). Also Einar Haugen included bracteates in the oldest Runic language before 550–600 (Haugen 1976:123–130). In DR the period of late urnordisk was assumed to begin c. 650 and last until c. 750 (or 800) (DR:1020) and Krause assumed a gradual change from the end of the 6th

Chronology and Typology of the Danish Runic Inscriptions 363 century onwards (Krause 1971:16; cf. the discussion, Nielsen 1970:26–36). Otherwise, Elmer Antonsen will “posit the period of the bracteates (archae- ologically 400–550) as the approximate time of the inception of the first phono- logical changes. Whether these changes occurred already around 400 or not until later in the 5th cent. cannot be determined.” (Antonsen 1975:11f.). According to recent opinions the younger language stage is taken to begin c. 500 (for in- stance Nielsen 1989:9), though the bracteate inscriptions “represent 24-letter ru- nic language material which immediately precedes the inception of the tran- sitional period” (Nielsen 2000:284). Ottar Grønvik assigns the big changes in the language in Scandinavia to a very short period from shortly before 500 to a temporary termination c. 575–600. He thereafter supposes a more gradual course for the linguistic development during the next period (III) lasting until the middle of the 9th century (Grønvik 1998:25f.). It is remarkable that Grønvik is operating only with the possible variations in the spoken language of the diffe- rent generations – with reference to the development in present Danish – not with any dialectal/regional differences or possibly another provenance.

The transition to the 16-character futhark

The development from the older to the younger runes has been a matter of much interest and many investigations – more or less speculative. Lately a purely graphic definition of transitional inscriptions has been suggested, which entails that a system of more than 16 characters could be used for inscriptions in the (Barnes 1998). In my opinion, evidently the transition to the younger runes took place in the period between the bracteate inscriptions, which yield examples of the 24-character futhark, and the 16-character-futharks on the Gørlev and Malt stones and the Hedeby stick. Unfortunately we have no safe datings for these inscriptions but I have concentrated on the now existing datable material before c. 800, i.e. the traditional historical starting point of the Viking Age (cf. for instance Roesdahl 1994:111–113). The Stentoften-Björketorp inscriptions could only be dated relatively al- though there has been an attempt at archaeological dating by Birger Nerman (Birkmann 1995:115f. with references). Around the middle of the 20th century

364 Marie Stoklund it was a rather common view – which one could still meet with in later decades, for instance in Einar Haugen’s history of the Scandinavian languages – that hardly any inscriptions were existing in central Denmark from 600 to 800. After that period, around 800 “a revival of runic writing burst forth in Denmark, per- haps inspired by the Blekinge stones” but now using “the new alphabet, the ‘younger’ futhark.” (Haugen 1976:140). However, the existence of Danish in- scriptions from that period has now been documented. Without determining where and how it happened that the old 24-character system was replaced by the 16-sign-futhark, I have tried to bridge the chronological gap between c. 600 and the inscriptions in the younger fuþark only with established dates and stressed the notable graphic variations and changes in the relationship between sound and written character in these inscriptions. The survey of the then known in- scriptions “From c. AD 650 to the ninth century” which Erik Moltke in 1976 (and 1985) offered was partly typologically based. A recent attempt to elucidate the development by combining with archaeological datings, styles and language history has been made by Thomas Birkmann in his survey of the Scan- dinavian inscriptions from Ågedal to Malt (Birkmann 1995), which has been useful in this connection, though the archaeological datings are not always up- to-date and Birkmann’s own assignations on different premises not unproblem- atic. In these inscriptions new combinations of sound and shape are found in the fragmented clasp from Kalmargården, archaeologically dated to (630) 650–700: ...uAlis, now with h A, s s, and perhaps u instead of W, if the inscription makes sense (Stoklund 1998a:59f.). From the 7th century, too, is the inscription on the Lousgård bead with s, probably for s or j, the only known valid example of s for /a/ seems to occur on the Istaby stone. Lousgård has H, but the reading M is not safe. Probably a should be read as n, not as a, considering for instance Eggja and Setre (Stoklund 2001b:570f.). Significant is especially the dendrodating to c. 725 of the layers from the Ribe excavations with the skull-fragment – an inscription which represents many of the linguistic features of the Viking-Age inscriptions and a rune-system like that on the Helnæs-Høje Taastrup-stones (DR 190, DR 250, Helnæs has h and a as ). On the Ribe skull we find in use only 15 of the 16 runes, which were to make up a complete Viking-Age system, obviously there was no need

Chronology and Typology of the Danish Runic Inscriptions 365 for È, h denotes /a/, and the shapes of the M- and H-runes are those of the older futhark (there is no g-rune G as supposed by Moltke) (cf. Stoklund 1996:202f.; Stoklund 2004:31).

Viking Age

In Table 2 and 3 again selected datable inscriptions are found with the signi- ficant features in the first column, followed by the datings in DR and in the third column the datings which were generally accepted around 1975. In the fourth column the present external datings, established on a different basis, dendro- chronology, reconsiderations of the historical sources, or art-historical chrono- logical studies. In order to avoid the danger of circularity of argument these present datings are never based on relative runological-linguistic criteria.

Table 2 Viking Age c. 800–1050. Selected inscriptions. Inscription DR c. 1975 Present datings Lindholm knife-haft (1953) c. 800 c. 800 È, k, a, s sikasuaiâ þurufiriþili"kaþi Hedeby loose objects c. 800 (9th C.) 9th–10th C. (1962–69) 3 with short-twig runes Elisenhof comb (1965), c. 800–850 c. 850–900 short-twig kÊÍR, kambr End of 9th C. Aarhus comb (1964) (Beginning of) the 10th C. hik'uin 10th C. DR41 Jelling 1 (1586) c. 935 hist. c. 935 Before 958/9(?) DR42 Jelling 2 (1586) c. 985 hist. c. 965 hist. c. 965–74 (or c. 985) DR2 and 4 Haddeby 2 After 934– Between 934 and First half of (1797) and 4 (1887) 950 10th C.

366 Marie Stoklund DR1 and 3 Haddeby 1 c. 1000; 982–1014 982–1014 980s– (1798) and 3 (1857) g, e or c. 1051 DR295 Hällestad 1 g (1668) Hist. uncertain c. 1000? 980s– DR279 Sjörup g (c. 1627) Hist. uncertain c. 1000? 980s– DR66 Århus 3 (DR 4) Hist. uncertain c. 1000? hist. the c. 1000, battle of (1850) battle of Svold Svold? DR345 Simris 2 (1716) Hist. uncertain Hist. uncertain Connection with king Canute (1016–35). Lund Viking-Age Mainly Medieval Lund walking stick, c. 990– inscriptions on loose objects c.1025 before c. 1050 (1881–, mainly after 1975). Bone 19 (1980) knif*melti: Lund comb 5, Maria Minor c. 1000 (1954/1996) (preliminary) reist:runar:þesar; ƒNcT Lund leather scabbard 5, St. Beginning of 11th Clemens 9 (1983/1996) C. (preliminary) Ía, n, cT iar, lifar:kati Roskilde stick with stone Connection with (1997) risti runar, þis"ar wreck 6, after c. instead of þasi, þisi 1025 (dendrodat.) Schleswig inscriptions on 19 12th C. The 11th. C. loose objects walking stick c. (1972–1976/2001). Stick: 1050–1100 (?) runaR iag risti, mogR, o o, g, e; walking stick: krist (nom.) s, c; amber ring: ...fgÍTar

In DR the inscriptions on of the so-called Helnæs-Gørlev-group are treated as Viking-Age, but in this survey new finds from before 800 with very similar features have been treated in connection with table 1. No doubt the definition and (absolute) dating of this group of stones ought to be reconsidered. It seems more likely that the Helnæs-Snoldelev inscriptions should be dated c.

Chronology and Typology of the Danish Runic Inscriptions 367 650–c. 750/800 (for instance Stoklund 1997: 27f.). However, during the archae- ological natural gas pipeline investigations a cemetery with inter alia a rich wo- man’s grave from the early 9th century was uncovered in the area in which the Snoldelev stone was found in the 1770s (Høgsbro 1987:160–163). An important feature, with respect to Viking-Age chronology, is that more inscriptions on objects which could be dated archaeologically have appeared in excavations, such as the Lindholm knife-haft c. 800 and early finds from Hede- by (and Elisenhof), Århus, later from Schleswig and Lund. Of special interest concerning the (time-)relations between the various younger futharks are the rune sticks from Hedeby with short-twig inscriptions but the dating of them seems to be rather loose (9th–10th century), though one could get the general impression that they are earlier: c. 800 (9th C.) (Moltke 1985:29, 371), c. 850 (Birkmann 1995:265; cf. Stoklund 2001a:116). The Jelling monument played a very important role in the chronology of DR and so it still does. Based on the historical written sources, Jelling 1 in DR (as in DRM) was dated to c. 935 and Jelling 2 to c. 985. The “Gnupa-stones” Haddeby 2 and 4 (Asfrid’s stones) were dated some time after 934 (Gnupa’s defeat by the German emperor Heinrich the First). It should especially be no- ticed that the Haddeby stones 1 and 3 were dated rather late: c. 1000 or the first half of the 11th century (DR:8f., 1024) because suin kunukR (on the Skarde- stone) was considered to be either Sven Forkbeard or – according to Lis Jacob- sen – Sven Estridson (1047–1076) i. e. as late as c. 1050. However, soon the late datings in DR of Harald’s Jelling stone and Haddeby 1 and 3 (DR 1, 3) were questioned, and the king Sven from Haddeby 3 identified as Sven Forkbeard (for instance Christensen & Moltke 1949 published 1971). Though it is now gene- rally accepted that all the inscriptions associated with the Jelling-kings belong to the 10th century, there are still discussions with regard to the more detailed chronology. Interdisciplinarily, the Jelling-Hedeby datings from DR have been used quite generally – and they are still rather common (for instance Haugen 1976:140) – though the “1975–status” (cf. table 2) ought to be acknowledged (cf. Nielsen 1970:37–39; Nielsen 1974 with contributions by different authors) and Moltke’s view in his “cautious” sketching of the historical events of the whole period 900–985 (Moltke 1985:199–201). Since the supposed ‘framing’, according to which tanmarkaR but should refer not to Thyra but to Gorm, has

368 Marie Stoklund been rejected (Moltke 1985:207 with references), this epithet cannot – as proposed in DR – refer to Gorm’s supposed victory over Gnupa and be used for the dating of the little Jelling stone to c. 935 (DR: 77). On the basis of excavations 1976–1979 in the present church in Jelling (from no later than c. 1100) Knud Krogh was able to demonstrate the traces of three wooden churches and a large chamber grave under the floor of the first and biggest of the wooden churches. In it were the mortal remains of a man, at least 35–50 years old, probably translated from the emptied chamber in the North mound and perhaps identical with King Gorm. The dendrochronological dating to 958/59 of timber from the grave chamber in the mound (Christensen & Krogh 1987:225f.) was of decisive importance for the chronology of the complex, though there is still some uncertainty concerning the interpretation of the histo- rical sources as well as the results of the archaeological investigations (cf. Anne Pedersen in this volume). According to the predominant opinion, it seems likely that it was Gorm’s death in the winter of 958/59 (not c. 935–940 as earlier as- sumed) which started the building of the grave chamber of the northern mound. If so, the little Jelling stone must have been erected before 958. The great ship-setting, demonstrated 1992 under the north as well as the south mound, cannot be dated, but might have had a connection with Jelling 1 and Thyra’s original kumbl (Krogh 1993: 251–256). The southern mound was started c. 970, according to dendrodatings of wood from the oldest part of the mound (Krogh 1993: 258f.). The stones from the ship-setting under it seem to have been standing in the open air for a rather long period (perhaps 20–30 years) to judge from the lichen found on them (Krogh 1993:259f.). Harald’s stone must be dated to the period after his conversion c. 965, and probably before his defeat in Hedeby (and ) c. 974, contemporaneous with the first church, the ope- ning of the northern mound and the start of the southern mound c. 970. It was built in three phases and after a longer standstill not finished till in the late 970s or 980s (Krogh 1993:257f.) It might be interpreted as a memorial barrow for Thyra, perhaps made by Ravnunga-Tovi and his fellows. The Bække stone 1 (DR 29) tells that Ravnunga-Tovi and Fundin and Gnypli made Thyra’s mound, and the Læborg stone (DR 26) that Ravnunga-Tovi hewed these runes in memo- ry of Thyra, his trutnik, ‘lady’, ‘queen’ (for instance Krogh 1993:266; Stoklund 2000:58 with references). However, according to the very problematic inter-

Chronology and Typology of the Danish Runic Inscriptions 369 pretation of the Bække-Læborg stones, presented by Birgit Sawyer during the conference, Ravnunga-Tovi was commemorating his own wife. Gorm’s Thyra had also been married to Ravnunga-Tovi and was buried by him elsewhere. This should have made it necessary for Harald to sponsor Jelling 1 in his father’s name in order to substantiate his claim to Denmark by a forgery (cf. now Sawyer 2000:159–166). An important argument used by Birgit Sawyer to sup- port her fallacious theory that Jelling 1 should be later than Jelling 2 and raised by King Harald is the fact that the little Jelling stone because of its monographic spellings represents a relatively younger linguistic stage than Harald’s stone, although any possibility of relative dating based on monographic versus digra- phic spelling after the “Helnæs-Gørlev period” is explicitly rejected in DR (DR:853f.). Although too much weight should not be attached to the differences, the relatively younger linguistic stage of the smaller Jelling stone together with the discovery that the ornamental frame on the B-side of Jelling 1 ends up not in a spiral but in a snake’s head, seem to be well in agreement with the possibly shorter time interval between the two stones. The theories of a formation of the inscription on the great Jelling stone in two phases, which have played a certain role in the earlier Jelling debate, could now be considered as rejected (for in- stance Krogh 1993:266). However, the opinion that the stone was erected as the final mid-point of the whole Jelling monument and only after the reconquest of Hedeby 983, before Harald’s death 987 at the latest, is still a matter of discus- sion (for example Krogh 1993:265f.). This late dating seems problematic in relation to the general chronology of the art styles of the period. Problematic too, is the coincidence that King Ha- rald’s stone should probably refer to the same event as Haddeby 3 (and 1), i. e. the reconquest of Hedeby, a dating which is now generally accepted. On Hadde- by 1 and 3 significant innovations such as the dotted runes e, g should be noted, well in accordance with their appearance on the Scanian Sjörup and Hällestad- stones (DR 279 and 295–297). These stones should probably be accepted as historical, too, referring to the battle of Fyrisvall at Uppsala in the 980s – though this was earlier rejected, for instance in DR (cf. Stoklund with references 1991:292).

370 Marie Stoklund It is remarkable that both Karl Martin Nielsen and Erik Moltke inde- pendently advanced an earlier dating of the Post-Jelling-group to c. 970–1025 instead of c. 1000–1050, no doubt rightly. This also meant an earlier stop to the general rune-stone-tradition in Denmark (except Bornholm), quite well in accor- dance with the lack of inscriptions referring to King Canute (except the probably Swedish Simris stone 2 (DR 345)) (Nielsen 1970:41–44; Christensen & Moltke 1971:319; Stoklund 1991:291–294).

The transition from Viking-Age to Medieval runes

The Danish “middelalder” is generally said to start in the first half of the 11th century or c. 1050, but it is noteworthy that the Swedish Viking Age lasts longer according to traditional chronology. Lis Jacobsen in DR operated with a Pre- Medieval Period 3 c. 1050–1150 between the Viking Age and Per. 4 The Middle Ages c. 1100–1350. In that way the Danish Middle Ages in DR would start at the same time as in .

Table 3 Medieval inscriptions c. 1050–. Selected. Inscription DR c. 1975 Present datings Svend Estridson runic coins oÉ c. 1065–1075 c. 1065–75 c. 1065–76 o, Œ‘ d, p p, y y/(w), q ð Lund, different loose Medieval Mainly c.11th–14th C. objects (1881 onwards, mainly Medieval after 1975). Lund wooden bowl (1979). 14th C. Latin (ol9auus:mæ:pos!i ƒ a, a æ, Í o, p p DR Add. 6 Lund Cathedral End of 12th C. c. 1080 12th C. double capital (1938) iƒk͇ i)ak)ob

Chronology and Typology of the Danish Runic Inscriptions 371 Schleswig knife-haft (1980) 11th, 12th C. Í o N n, ƒ a, a æ, ’“ d, Òp p, c s

Schleswig wooden bowls c. 1250–1300 (1971–74) m9aria, ƒ Graffiti and plaster inscriptions c. 1200–, c.1100–1250 in churches Romanesque and later DR413 Gunhild’s cross (1650) c. 1100 c. 1100 or later c. 1150 gun"hild, g“ DR347 Norra Åsum stone c. 1200 c. 1200 c. 1200 (1598) Kalundborg antler (1984) Latin. c.1250–1350 ƒ a, a æ, g g, – d, p p, w v DR215 Stokkemarke reliquary 1286–1300 1286–1300 1286–1300 (1835) æpisk9opus'gisiko DR366 (†)Lösen tombstone 1310 or 1311 1310 or 1311 1310 or 1311 (1746) Carpenter’s runes; Hvidding c. 1200 Dendrodating: after (1953, 1991) fuìÊ 1254/55 Brøns (1953) fuìÍ×khN and c. 1250– Dendrodating: ìi’×Ik þidrik c. 1200–

The runestones from Bornholm and the runic coins made up the main material of this Period 3 in DR and the language of this period was defined as Old Danish, but as “Older Middle Danish” in Per. 4. The rune stones of Bornholm belong to a Viking-Age tradition but the majority have been dated to the epoch from c. 1050 to 1100 or to the period after the conversion of Bornholm by bishop Egino. These limits might be too narrow; there could have been Christians on Bornholm earlier, and some of the stones might have been erected before 1050, some after 1100. In these inscriptions, according to DR, the first examples of the use of the old ansuR-rune “as a rune o” oÉÊ could be noted. The coins belong in a Medieval context. According to the generally accepted, absolute dating, the runic coins form a fixed point in the chronology.

372 Marie Stoklund They were struck in the last decade of Sven Estridson’s reign c. 1065–1076, mainly in Lund, probably with close connections to England. Graphematically, they form a very special group with oÉ for o, more dotted runes: p p and different variants of a dotted long-branch t, Œ‘. In the coin-inscriptions until now only one example of the traditional R-rune: ö, had been found in the Danish name alfkeR AlfgæiRR (Lerche Nielsen 1997:73f.). Now another coin with the same name, spelled alfgeiR, has turned up in Daugmale in Letland (Berga 2001:8). Otherwise in these inscriptions generally r is used also instead of earlier ö. Since 1940 other inscriptions have appeared which contain possibilities of archaeological dating. Owing to urban excavations, especially in Lund and Schleswig, the number of inscriptions on objects from the late Viking- and early Medieval Ages have increased, but many of these finds do not offer relevant linguistic texts. In DR such features “as the use of the rune o and a new diphthong ei” are – “together with sporadic occurrence of the one-sided runes a and n” – considered to be results of a special Swedish impact (especially on Bornholm) (Summary:24, 25; DR:1028, 1029), although o and ei are also found on the runic coins and in that case might be explained as a Danelaw feature. Other recent finds with similar examples of ei from Schleswig and Lund also put a question mark to this assumption of foreign, especially Swedish, influence (cf. Lerche Nielsen 1997:69).

Datable Medieval inscriptions c. 1100–1500

Though the accession of Medieval inscriptions does not affect the overall chro- nology decidedly, some adjustments to take account of new dated material and a few chronological changes should be registered. The important inscription on the Ribe healing stick, however, has been dated on linguistic premises and a considerable increase of lead-amulets with Latin inscriptions could only in a few cases be dated closely. They might in some cases be late-Medieval. It could be noted that the dating of the ‘historical’ Gunhild’s cross (DR 413) probably should be changed from c. 1100 to c. 1150, and a new investigation re- vealed a dotted d, “ instead of t in gunºhild (Langberg 1982:40,73; Moltke

Chronology and Typology of the Danish Runic Inscriptions 373 1985:454). The dating of three of the other historical inscriptions in DR must still be considered safe, they are retained in table 3 as fix-points. According to Lis Jacobsen’s survey, the ecclesiastical inscriptions archae- ologically “may safely be assigned to the 12th and 13th centuries, the greater part belonging to the century between 1150 and 1250” (Summary:28) and, linguistically, the medieval written forms and flexional cases chiefly agree with those of the most ancient Danish codices, and the ornamentation must be dated to the Romanesque epoch (Summary:26).

The expansion of the younger futhark in the 11th century and later

No doubt the development of runic writing systems from the Late Viking Age and in the Medieval period ought to be seen in a Common-Scandinavian light. A sort of interaction between different futhark types took effect in the use of originally allographic variations to denote linguistic implications. The survey of Norwegian rune writing by Magnus Olsen (NIyR 5:238–245) not only covers the West-Nordic cases but also holds good for the East-Scandinavian or Danish development, perhaps with certain reservations concerning the datings. It is difficult to get a similar survey of the Danish development, however, because the information in DR mainly has to be gathered from the articles on the “indi- vidual runes and their sound-value” in Runerne combined with Tidsfæstelse og typologi. It is also important to stress that DR was written before the structural linguistic principles for description had had their breakthrough. Thus of course there is no clear distinction between phonemic and phonetic or graphemic (allo- graphic) aspects. Evidently in the period late Viking Age/early Medieval time, changes in runic writing began to manifest themselves more frequently: increasing use of stung runes, not only the dotted e, g, and u, which we have already seen, but a dotted t and b (on the coins), too, supplemented also by grapheme-variants, de- rived from various futhark types to designate (new) phonemes (and allophones). The short-twig-futhark, obviously, seems to play an important part for instance in the distinction made possible by employing a and ƒ respectively to designate

374 Marie Stoklund /æ/ and /a/. In Medieval inscriptions proper we could find ö, ýr, used as , since had been ousted. A significant feature is the use of the old ansuR-rune as an o-rune, according to the sound-change in its name. Moltke has suggested that the oldest dated Danish example of the left-sided Í (the most common in Medieval inscriptions) might be found in the Jakob-inscription DR Add. 6 Lund Cathedral double capital (Moltke 1985:426). His proposed dating of the capital to around 1080 (against DR: the end of the 12th century) could be ques- tioned and the use of bindrunes in the inscription might rather indicate a later date. The use of a for /æ/ has been defined as diagnostic for Medieval inscriptions but it can be problematic to distinguish between the inscriptions with or without æ, in the cases where the text for instance is too short for a safe distinction to be drawn. This is the case for instance in the Lund-inscriptions, which seem to have been rooted in a traditional use with preference for long-branch runes. It is inte- resting that the secondary value as /æ/ was designated by the old long-branch a- rune, while the short-twig a-rune, which is rather unusual in the Danish Viking- Age inscriptions, was used for /a/.

Further extension of the inventory of signs

According to Lis Jacobsen, all true medieval inscriptions were characterized runographically by a futhark that was adapted to the , though non- dotted runes instead of expected dotted ones could appear occasionally. In my opinion, the results now attained show a two-level increase in the use of the new graphemic (and/or allographic) variations. Parallel to the Norse (and Swedish) development, the number of dotted runes rose from around 1200, cf. the Norra Åsum stone (DR 347), the Kalundborg antler, the runic manuscripts and the Schleswig-inscriptions for instance. The rather early archaeological da- ting of the Schleswig knife-haft, however, is problematic compared with the inventory of signs (Stoklund & Düwel 2001:229). It is obvious moreover that in the Danish Medieval material, too, it was a common feature that the dotted variants were not used obligatorily. The conventions of the Viking-Age runic script survived – even in Latin inscriptions. The strong position which the fu-

Chronology and Typology of the Danish Runic Inscriptions 375 thark with only sixteen runes kept is clearly demonstrated by several medieval futhark-inscriptions from Old Danish territory, found after 1940. Especially im- portant are the cases where an archaeological dating might be possible as for instance in Lund and Schleswig or based on dendrochronology, as in the case of the dating of roof-timber from Hvidding Church (Madsen 1994:42–65, 67f.; Madsen forthcoming).

Linguistic development 1050 onwards

After a probably rather stable period during the 9th and 10th centuries in regard to linguistic development, the inscriptions from the 11th–12th centuries obvi- ously reveal some important language changes, for instance changes in the inflectional system: the lack of the nominative marker -R/-r, examples of wrong vowels or a swarabhakti vowel in the unstressed syllables, as shown in my survey. These features are probably not without connection with the so-called vowel weakening of the unstressed vowels, which could be traced in various ways in the law-manuscripts from the different Danish regions. The three vowel- system a–i–u of the Viking-Age futhark reflected the basic phonemic distinc- tions in unstressed syllables and allowed adequate and necessary distinctions within the inflectional system between the three commutable vowels. However, changes within this system seem to have started in the period between the Vi- king Age proper and the manuscript period. Anders Bjerrum has made an inte- resting attempt to analyse the Danish runic inscriptions from the period c. 1000–c. 1250 based on DR with regard to the distinctions made possible by the new graphemes, e, æ and o (Bjerrum 1952, revised in English 1973). Even though this is an example of a rather problematic use of the relative DR-datings in order to establish an absolute chronology of the dialectal split and regional merging of the three vowels in unstressed syllables, a new investigation along these lines ought to be done in consideration of the increase of material and better datings of part of the new inscriptions recovered since 1952.

376 Marie Stoklund Closing remarks

The many new finds in Denmark since c. 1940 form an important supplement to the stock of inscriptions which were the basis for the chronology and typo- logy of DR. Though nearly all of these inscriptions have been published – but not easily comprehensible and not always satisfactorily – it is evident that the knowledge of this new material ought to be made more easily accessible than it is now. It is rather shocking to read a brand new contribution concerning the development of the Scandinavian language c. 200–900 based on runic evidence but obviously with an outdated, very sporadic knowledge of Danish inscriptions and relevant literature on this topic (Isakson 2000). Especially the increase of inscriptions with the older runes makes it evident that a new corpus-edition is badly needed. Now Wolfgang Krause’s as well as Elmer Antonsen’s works (Krause & Jankuhn 1966, 1971; Antonsen 1975) can no longer be accepted as a sound basis for interdisciplinary work because of the increase in the number of finds and since neither their detailed chronology nor their interpretations are reliable. They have, however, often been used for larger investigations i. e. of historical linguistic development, magic or “an investi- gation of the chronological and geographical distribution of the different allographs” (Odenstedt 1990). In some ways Makaev’s survey would be a better choice but the Russian original from 1965 has only been translated, not up-dated 1996 (Makaev 1996). In general the basis for datings of runic material used interdisciplinarily ought to be checked. In a linguistic context for example, it is worth noting that the chronology in the three first volumes of Gammeldansk Grammatik is based on Ludvig Wimmer’s work, but that in the later volumes DR is the basis. Ottar Grønvik’s important and ambitious attempt to create a new historical linguistic chronology is mainly based on Norwegian inscriptions from Tune to Ødemot- land and Eggja and combined with archaeological datings by Egil Bakka (latest survey: Grønvik 1998:25f.). During the last decades of the 20th century new attempts have been made to outline especially the relative and absolute chrono- logy of the inscriptions in older runes. It is, however, remarkable that when runology is used in connection with, and as a basis for, language history the au- thors often create their own chronology, frequently based on presupposed lin-

Chronology and Typology of the Danish Runic Inscriptions 377 guistic premises. Another problem is the general tendency to use a problematic foundation of questionable interpretations as basis for a chronology. When it comes to further interdisciplinary utilization a certain runological tradition of speculative imagination without restraint becomes a problem.

Acknowledgments

The Danish Humanities Research Council from 1 August 1999 to 31 January 2000 supported my project: Chronology and Typology of the Danish Runic Inscriptions by paying half my salary. This project was based on my preliminary work at The National Museum on a planned new corpus edition of the Danish Runic inscriptions for the Society of Danish Language and Literature (DSL), which during 1998 was also supported by The Danish Humanities Research Council. I would like to take this opportunity to thank the Research Council as well as DSL and the National Museum for support to pay student assistants.

Postscript 2005

In the main I have kept my old presentation from 2000 with only a slight updating. New finds have been included and minor changes concerning the original (preliminary) datings.

378 Marie Stoklund Appendix: Proto-Norse index. Increase after 1941/1942–2004 ahti Nydam strap suspension anulºa Nydam strap suspension aiþalataR Nydam axe-shaft alu Nydam axe-shaft (la, lua Nydam arrow-shafts) alugod Værløse brooch balika Sorte Muld antler fragment bidawarijaR Nøvling brooch ek Gårdlösa brooch fulaR Sorte Muld antler fragment gauþR Illerup fire-steel handle hagiradaR Stenmagle box }harkilaR Nydam strap suspension laguþewa Illerup shield handle mount (silver) 1 lamo Skovgårde/Udby brooch niþijo Illerup shield handle mount (silver) 2 rawsijo Nydam belt end fitting swarta Illerup shield handle mount (bronze) sikijaR Nydam axe-shaft talgida Skovgårde/Udby brooch talgidai Nøvling brooch tawide Stenmagle/Garbølle box, Illerup shield handle mount (silver) 2 þewaR Sorte Muld antler fragment unwodºR Gårdlösa brooch wagagastiR Nydam axe-shaft wagnijo Illerup lance heads 1, 2, Vimose lance head $widuhudaR Himlingøje brooch 2 witrÄ Slemminge/Fælleseje antler tool

Chronology and Typology of the Danish Runic Inscriptions 379 Bibliography

Antonsen, Elmer H. 1975. A Concise Grammar of the Older Runic Inscriptions. Sprachstruk- turen. Reihe A. Historische Sprachstrukturen 3. Tübingen: Niemeyer. — 1998. “On Runological and Linguistic Evidence for Dating Runic Inscriptions.” In: Runen- inschriften als Quellen interdisziplinärer Forschung, 150–196. Axboe, Morten 1998. “Die innere Chronologie der A-C-Brakteaten und ihrer Inschriften.” In: Runeninschriften als Quellen interdisziplinärer Forschung, 231–252. — 1999. “The chronology of the Scandinavian gold bracteates.” In: The Pace of Change. Studies in Early-Medieval Chronology. Ed. by John Hines, Karen Højlund Nielsen and Frank Sigmund. Cardiff Studies in Archaeology. Oxford: Oxbow Books, 126–147. Barnes, Michael P.1998. “The Transitional Inscriptions”. In: Runeninschriften als Quellen inter- disziplinärer Forschung, 448–461. Berga, Tatjana 2001. “Danske mønter fra fund i Letland i 1000- og 1100-tallet.” In: Nordisk Numismatisk Unions Medlemsblad 2001:1, 3–8. Birkmann, Thomas 1995. Von Ågedal bis Malt. Die skandinavischen Runeninschriften vom Ende des 5. bis Ende des 9. Jahrhunderts, RGA–E 12. Berlin–New York: de Gruyter. Bjerrum, Anders 1973. “The unstressed vowels in Danish runic inscriptions from the period c. 1000–c. 1250.” In: Linguistic Papers. Published on the occasion of Anders Bjerrum’s 70th birthday, 12th March 1973 by Selskab for Nordisk Filologi. Copenhagen: Akademisk Forlag, 116–129. Christensen, Kjeld & Krogh, Knud J. 1987. “Jelling-højene dateret.” In: Nationalmuseets Ar- bejdsmark 1987, 223–231. Christensen, Aksel E. & Moltke, Erik 1971. “Hvilken (kong) Svend belejrede Hedeby?” In: Historisk Tidsskrift (København) 12. række, 10, 297–326. DR = Jacobsen, Lis & Moltke, Erik et al.: Danmarks runeindskrifter, København: Munksgaard 1941–1942. DRM = Wimmer, Ludvig F.A.: De danske runemindesmærker 1–4, København: Gyldendalske Boghandel 1893–1908. GG = Brøndum-Nielsen, Johannes: Gammeldansk Grammatik i sproghistorisk Fremstilling 1–8, København: Schultz. 1928–1973. Grønvik, Ottar 1987. Fra Ågedal til Setre. Sentrale runeinnskrifter fra det 6. århundre. Oslo Bergen Stavanger Tromsø: Universitetsforlaget. — 1996. Fra Vimose til Ødemotland. Nye studier over runeinnskrifter fra førkristen tid i Norden. Oslo: Universitetsforlaget. — 1998. Untersuchungen zur älteren nordischen und germanischen Sprachgeschichte. Osloer Beiträge zur Germanistik 18. Frankfurt am Main: Peter Lang. Hagland, Jan Ragnar 1998. “Innleiing. Litt om datering på runologisk og språkleg grunnlag.” In: Innskrifter og datering. Dating Inscriptions. Ed. Audun Dybdahl, Jan Ragnar Hagland. Senter for middelalderstudier. Skrifter nr. 8. Trondheim: Tapir, 9–15.

380 Marie Stoklund Haugen, Einar 1976. The Scandinavian Languages. An Introduction to their History. London: Faber and Faber. Høgsbro, Kirsten-Elizabeth 1987. “Gravpladsen ved Gunvalds sten.” In: Danmarks længste ud- gravning, Arkæologi på naturgassens vej 1979–86. Ed. Rigsantikvarens arkæologiske Sekre- tariat, Nationalmuseet og de danske naturgasselskaber, Herning: Poul Kristensen: 160–163. Isakson, Bo 2000. “The Emergence of a Scandinavian Language: Development until around 900.” NOWELE 36, 3–43. KJ = Krause, Wolfgang & Jankuhn, Herbert 1966. Krause, Wolfgang 1971. Die Sprache der urnordischen Runeninschriften. Heidelberg: Carl Winter Universitätsverlag. Krause, Wolfgang & Jankuhn, Herbert 1966. Die Runeninschriften im älteren Futhark. 1–2. Göttingen: Vandenhoeck & Ruprecht. Krogh, Knud J. 1993. Gåden om Kong Gorms Grav. Historien om Nordhøjen i Jelling. Vikinge- kongernes Monumenter i Jelling 1. Udgivet af Carlsbergfondet og Nationalmuseet. Herning: Poul Kristensen. Langberg, Harald 1982. Gunhildkorset. Gunhild’s Cross and Medieval Court Art in Denmark. København: Selskabet til udgivelse af danske Mindesmærker. Lerche Nielsen, Michael 1997. Vikingetidens personnavne i Danmark belyst gennem runeind- skrifter og stednavne på -torp sammensat med personnavneforled. Unpublished Ph.d.-thesis. Institute of Name Research, University of Copenhagen. Madsen, Per Kristian 1994: “Dendrokronologiske undersøgelser af Hvidding kirkes tagværk.” I : By, marsk og geest 6. Årsberetning 1993. Den antikvariske Samling i Ribe. 1994, 42–65, 67f. Madsen, Per Kristian forthcoming: “Middelalderlige kirketagværker i Sydvest- og Sønderjyl- land”. To be published in Aarbøger for nordisk Oldkyndighed og Historie. Makaev, È.A 1996 [1965]. The Language of the Oldest Runic Inscriptions. A Linguistic and His- torical-Philological Analysis. Translated from Russian by J. Meredig in consultation with E.H. Antonsen. Stockholm: Kungl. Vitterhets Historie och Antikvitets Akademien. [Russian edition originally published in Moscow, 1965]. Moltke, Erik 1976. Runerne i Danmark og deres oprindelse. København: Forum. — 1985. Runes and Their Origin. Denmark and Elsewhere. Copenhagen: The National Museum of Denmark. Nielsen, Hans Frede 1998. “The Linguistic Status of the Early Runic Language of Scandinavia.” In: Runeninschriften als Quellen interdisziplinärer Forschung, 540–555. Nielsen, Karl Martin 1970. Om dateringen af de senurnordiske runeindskrifter. Synkopen og 16 tegns futharken. København Akademisk Forlag, Institut for nordisk Filologi, Studier nr. 2. [Also publ. in: Aarbøger for nordisk Oldkyndighed og Historie 1969]. — 1974. “Jelling problems: A discussion.” By Karl Martin Nielsen, with contributions by Niels Åge Nielsen, Erik Moltke, Else Roesdahl, Harald Andersen, Olaf Olsen. In: Mediaeval Scandinavia 7, 1974, 156–234. Nielsen, Niels Åge 1983. Danske Runeindskrifter. København: Hernov.

Chronology and Typology of the Danish Runic Inscriptions 381 NIyR 5. 1960. Norges Innskrifter med de yngre Runer 5. By Magnus Olsen et al. Oslo: Kjelde- skriftfondet. Odenstedt, Bengt 1990. Odenstedt: On the Origin and Early History of the Runic Script. Ty- pology and Graphic Varition in the Older Futhark. Acta Academiae Gustavi Adolphi 59. Uppsala. Roesdahl, Else 1994. “Dendrochronology and Viking Studies in Denmark, with a Note on the Beginning of the Viking Age.” In: Developments Around the Baltic and the North Sea in the Viking Age. Ed. Björn Ambrosiani and Helen Clarke. Stockholm: Birka Studies [3], 106–116. Runeninschriften als Quellen interdisziplinärer Forschung. 1998. Abhandlungen des Vierten Internationalen Symposiums über Runen und Runeninschriften in Göttingen vom 4.–9. August 1995. Hrsg. Klaus Düwel in Zusammenarbeit mit Sean Nowak, RGA–E 15. Berlin- New York: de Gruyter Sawyer, Birgit 2000. The Viking-Age Rune-Stones. Custom and Commemoration in Early Medi- eval Scandinavia. Oxford: University Press. Steuer, Heiko 1998. “Datierungsprobleme in der Archäologie.” In: Runeninschriften als Quellen interdisziplinärer Forschung, 129–149. Stoklund, Marie 1991. “Runesten, kronologi og samfundsrekonstruktion.” In: Mammen. Grav, kunst og samfund i vikingetid. Ed. Mette Iversen et alii. Jysk Arkæologisk Selskabs skrifter 28. Århus: Aarhus Universitetsforlag, 285–297. — 1994. “Myter, runer og tolkning.” In: Myte og Ritual i det førkristne Norden. Et Symposium. Ed. Jens Peter Schjødt et alii. Odense Universitetsforlag , 159–174. — 1995. “Die Runen der römischen Kaiserzeit.” In: Himlingøje - Seeland - Europa. Ed. Ulla Lund Hansen et al. Det Kongelige Nordiske Oldskriftselskab. København, 317–346. — 1996a. “The Ribe Cranium Inscription and the Scandinavian Transition to the Younger Re- duced Futhark.” In: Amsterdamer Beiträge zur älteren Germanistik 45, 199–209. — 1996b. Stoklund, Marie: “Runer 1995.” In: Arkæologiske udgravninger i Danmark 1995. Ed. Rigsantikvarens Arkæologiske Sekretariat, Det arkæologiske Nævn. København, 275–286, (287–294). — 1997. Stoklund, Marie: “Runefundene fra yngre jernalder — tidlig vikingetid og den danske runetradition før Jelling. Eller myten om svenskevældet.” I: Beretning fra sekstende tværfaglige vikingesymposium. Højbjerg: Hikuin, 25–36. — 1998a. Stoklund Marie: “Neue Runenfunde aus Skandinavien. Bemerkungen zur metho- dologischen Praxis, Deutung und Einordnung.” In: Runeninschriften als Quellen interdiszi- plinärer Forschung, 55–65. — 1998b. Stoklund, Marie: “Runer 1997.” In: Arkæologiske udgravninger i Danmark 1997. Ed. Rigsantikvarens Arkæologiske Sekretariat, Det arkæologiske Nævn. København, 259- 265, (266–269). — 2000. Stoklund, Marie: Jelling. §2. Runologisches. In: Reallexikon der Germanischen Alter- tumskunde. Begründet von Johannes Hoops. Zweite Auflage 16, 56–58.

382 Marie Stoklund — 2001a. Stoklund, Marie: “Die Inschriften von Ribe, Hedeby und Schleswig und die Bedeu- tung der Schwedenherrschaft.” In: Von Thorsberg nach Schleswig, 201–208. — 2001b. Loftsgård. Runologisches. In: Reallexikon der Germanischen Altertumskunde. Be- gründet von Johannes Hoops. Zweite Auflage 18, 570f. — 2003. “The first runes – the literary language of the Germani”. In: Lars Jørgensen, Birger Storgaard & Lone Gebauer Thomsen (eds.): The Spoils of Victory. The North in the shadow of the Roman Empire [English edition]. Copenhagen: Nationalmuseet, 172–179. — 2004. “The Runic Inscription on the Ribe Skull Fragment”. In: Mogens Bencard, Aino Kann Rasmussen & Helge Brinch Madsen (eds): Ribe Excavations 1970–76 vol. 5, Højbjerg: Jutland Archaeological Society, 27–42. Stoklund, Marie & Düwel, Klaus 2001. Die Runeninschriften aus Schleswig. In: Lerche Nielsen, Stoklund & Düwel: Neue Runenfunde aus Schleswig und Starigard/Oldenburg. Mit archäo- logischen Beiträgen von Ingo Gabriel und Volker Vogel. In: Von Thorsberg nach Schleswig, 201–237. Summary = The Runic Inscriptions of Denmark. By Jacobsen, Lis & Moltke, Erik et al. Sum- mary, 1947. Copenhagen: Munksgaard. Von Thorsberg nach Schleswig 2001. Sprache und Schriftlichkeit eines Grenzgebietes im Wandel eines Jahrtausends. Internationales Kolloqium im Wikinger-Museum Haithabu vom 29. September–3. Oktober 1994, unter Mitarb. von Lars E. Worgull hrsg. von Klaus Düwel, Edith Marold, Christiane Zimmermann. RGA–E 25.

Chronology and Typology of the Danish Runic Inscriptions 383