Printer Friendly

Dendrochronological dating of the Viking Age ship burials at Oseberg, Gokstad and Tune, Norway.

Dendrochronology now provides a date, exact nearly to the year, for three Viking Age

burial mounds of special importance for chronology in Scandinavia and across early medieval northern Europe. Their dating used to depend on the style of the carved wooden artefacts in the grave goods; now the grave-goods are exactly and independently dated by

the tree-rings, those same links will provide dating bridges across the Viking world.

The Norwegian ship-burials and their dating

The dating of the important finds in the burial mounds at Oseberg, Gokstad and Tune in southeast Norway (FIGURE 1), now on display in the Viking Ship Museum in Oslo, is central to our understanding of Viking Age chronology and stylistic development. In particular the Oseberg find, with its content of unique carved wooden items, surpasses all other single finds from the Viking period.


The Gokstad and Oseberg sites are situated in the county of Vestfold on the western side of the Fiord of Oslo. The two sites were excavated in 1880 and 1904 respectively. The Tune site is situated on the eastern side of the same fiord in the county of Ostfold and was excavated as early as 1867. It was a sensation at the time, and although the ships in the Gokstad and Oseberg mounds clearly surpass it, it was the first substantial archaeological evidence of a ship used by the Vikings.

In all three cases we are dealing with burials where the dead lay in a grave chamber constructed of wood and placed in a ship along with the grave goods, the whole covered by a mound.

Until now the dating of the three Norwegian ship burials has been largely based on evaluations of the decorated wooden artefacts recovered during the excavations.

Viking Age art can be divided up into several stylistic periods, which replace and overlap each other throughout the three centuries of the Viking Age. They are all based on the Nordic animal style which was inspired by western European, in particular Irish, ornamentation. The result was an independent and original Nordic style. The traditionally accepted sequence of Viking Age styles begins with the 'Oseberg Style' around AD 800, continuing with the 'Borre Style', named after another important mound find in the county of Vestfold. This was followed by the 'Jellinge Style', named after a silver cup found in a Danish mound in Jutland, dated to the second half of the 10th century. The style of what is regarded as late Viking art has been given the name 'Ringerike Style', after a group of decorated stones in Norway, and ends with the 'Urnes Style' from the late 11th and early 12th centuries, which is named after an early wooden church in western Norway. This is only an extremely brief summary of the chronological sequence of Viking art styles. The definition of the different periods, often with sub-divisions, and their dating are under constant debate, and differ according to the background and nationality of the contributor (Schetelig 1920; Wilson & Klint-Jensen 1966; Karlsson 1983; Roesdahl & Wilson 1992).

The stylistic analyses of items from these finds have been carried out first and foremost by the Norwegian archaeologist Haakon Shetelig. As a young man he took part in the excavation of Oseberg and later played an important role in the publication of the find, as well as contributing extensively to the international discussion about the chronology of Viking art.

In publishing the wooden artefacts from Oseberg, Shetelig also dealt with the Gokstad find as well as the metal artefacts from both Gokstad and Borre. He concluded that what he called the Early Oseberg Style began around AD 800, whilst his Later Oseberg Style was fixed at around the middle of the 9th century. He dated the Borre Style in the Gokstad and Borre finds to after AD 900 (Schetelig 1920). The leader of the excavation of the Gokstad mound, Nicolay Nicolaysen, had earlier dated the burial to c. AD 900 (Nicolaysen 1882) (FIGURE 2). In a lecture to the Nordic Archaeological Meeting in Helsinki in 1925, Shetelig later changed his views with regard to the Oseberg find (FIGURE 3). Based on analysis of the Irish artefacts in the find he revised the date for the start of the Early Oseberg Style to between AD 820 and 830 (Shetelig 1926), which for him consequently meant that the burial at Oseberg had to be dated to AD 850 or later. This date was subsequently, and still is, generally accepted (Roesdahl & Wilson 1992).


In 1917 Shetelig also published the Tune ship-burial. A few fragments of wood carved in the Borre style, possibly from a saddle, were the only datable artefacts. Shetelig concluded that the Tune burial was contemporary with that at Gokstad (Schetelig 1917).

After the reconstruction of the Gokstad mound in 1930, Bjorn Hougen, later Professor of Archaeology in Oslo, again took up the discussion concerning its age. Using as evidence some bronze strap-mounts that were found during the reconstruction work, of which one was decorated in a pure version of Jellinge Style, Hougen concluded that the grave was from 'very early in the 10th century' (Hougen 1934).

Other methods have also been employed in the efforts to date these finds. In 1959 the Radiological Dating Laboratory in Trondheim carried out the radiocarbon dating of a sample of oak wood from the grave chamber in the Oseberg ship. The result in radiocarbon years was 1190[+ or -]60 b.p. (T-37) (Nydal 1959), giving a calibrated age of AD 880, or with [+ or -] one standard deviation AD 780--960 (Stuiver & Pearson 1993).

Results of recent archaeological investigations carried out in Ribe, Denmark (Jensen 1991), indicate that a revision of the chronology of early Viking art styles is necessary; the whole chronology possibly needs to be moved further back in time. Thus the precise dating of these key Norwegian finds is more pressing than ever.

Dendrochronology of the ship-burials

Within recent decades dendrochronology has established itself as one of the most important scientific dating methods in northwest European archaeology (for a general introduction to the method see Baillie 1982). It offers the archaeologist and historian an exact answer to the most important question posed in connection with an archaeological find -- how old is it? The result is often so precise that every recognized theory which conflicts with it is immediately discredited (Bonde & Christensen 1982; Christiansen 1982).

During the course of the 1980s, several major master chronologies have been built up for oak (Quercus sp.) in southern Scandinavia, covering the period from the present day back to the Iron Age. At present there is no chronology for oak in Norway, but work is in progress. Preliminary investigations of the annual rings of living trees in southern Norway show that it is possible to match a regional chronology for oak from the coastal regions around the Kattegat with master chronologies from Denmark and southern Sweden (Kjeld Christensen pers. comm.). This means that under favourable circumstances a 'floating' chronology, based on oak material from the southern Norwegian mixed forest region, can be dated using one or more of the master chronologies which already exist for the areas around the Kattegat.

As the result of co-operation between the University Museum of National Antiquities in Oslo and the National Museum of Denmark, a series of wood samples have been removed from the Oseberg, Gokstad and Tune finds for the purposes of dendrochronological dating. The procedure in the investigation was first to produce a 'floating' regional chronology for the Oslo Fiord area, based on samples from the three sites. Attempts were then made to match this with existing master chronologies for southern Sweden (Bartholin 1985; Brathen 1982) and Denmark.

Selection of samples

The three localities of Oseberg, Gokstad and Tune lie between 20 and 50 km apart. We can assume that climatic conditions were more or less uniform over the whole area around the sites. When choosing items for dating, emphasis was therefore placed on samples from trees which would be expected to have been felled close to the site of end use, i.e. the burial mounds. In all three cases, samples from the ships themselves were avoided as there is always some doubt as to the origin of timber which has gone into the building of a ship. For example, dendrochronological investigations of samples from Wreck 2 from the famous Skuldelev find near Roskilde, Denmark, showed that the ship was built of timber from trees which had grown in the area around Dublin, Ireland (Bonde & Crumlin-Pedersen 1990).

The investigation concentrated on timber which had been used in building the grave chambers at the three sites. The oak trees which produced the timber for the chambers at Oseberg, Gokstad and Tune were surely felled for the purpose of building the burial chambers; the felling dates for these timbers will therefore correspond to the ages of the burials. There would not have been any significant storage or seasoning period for the timber after felling, as it was normal to use and work timber in the fresh, newly felled state. Newly cut trees are much easier to work with hand-tools, axes, wedges and so on. Presumably we only have to take into account the time it took to transport the timber from the felling site to the burial mound, traditionally a matter of months, not years, and in this investigation we might be talking about weeks or maybe even days.

The grave chambers were never meant to be seen. In Oseberg and Gokstad they were primitively built. They were made of roughly cleft and hewn planks and posts and placed like a tent in the centre of the ships. They were more or less intact, if we ignore the damage done when the graves were broken into, probably already in the Viking period.

The Oseberg grave chamber is now stored at the Viking Ship Museum in Oslo (FIGURE 4) and the material is extremely well preserved, although it has never undergone any extensive conservation treatment in any modern sense of the word. Only the surface of the timber has been occasionally treated with linseed oil. The chamber was built completely of oak. Twelve samples were taken, two from the gables, one from a vertical post which bore the ridge pole and nine from planks which probably come from the roof.


The Gokstad grave chamber is on display at the Viking Ship Museum (FIGURE 5) and the timber is also very well preserved. It has probably undergone the same treatment as the timber from Oseberg. The chamber was built of timber from both oak and pine. Four samples were taken, all oak, one of which probably derives from the roof of the grave chamber and three from the gable nearest the ship's mast.


The grave chamber in the Tune ship was only rudimentarily preserved and the material is also stored in the Viking Ship Museum (FIGURE 6). Only four small pieces of oak timber from the chamber itself were saved. It seems to be of a totally different construction from those at Gokstad and Oseberg. It has a square chamber with walls made of radically split planks and a flat roof, showing closer affinities to other Scandinavian chamber graves from the tenth century. Two samples were taken, both of which probably came from the vertical wall of the chamber.


All of the samples were taken as cross-sections by sawing through the timber pieces. These will be restored again after the investigation.


The tree-ring curves from Oseberg can be combined to give a chronology covering 299 years. Similarly the curves from Gokstad and Tune also fit together, producing chronologies of 340 and 184 years respectively. The chronologies from the three localities fit together very well and form the basis for constructing a regional chronology covering a total of 357 years, using results from a total of 18 samples. This regional chronology, given the name 'Oslo Fiord -- Viking Age', can be dated against the master chronologies from southern Sweden and Denmark (TABLE 1). It covers the period from AD 536 to 892.
 Denmark Scania W. Gotland
 (1) (2) (3)
Oslo Fiord
Viking Age 5.36 4.45 3.52

If the results from the three sites, Oseberg, Gokstad and Tune, are treated separately the following is evident (FIGURE 7):

The local chronology from Oseberg covers the period AD 536--834; all 12 samples were dated. Ten samples had intact sapwood, of which five had bark rings in which only early wood had formed. The last annual ring was formed in AD 834, which means that the wood samples came from trees felled in the summer of AD 834.

The local chronology for Gokstad covers the period AD 548--887; all four samples from the grave chamber were dated. None of the samples had intact sapwood. The outermost preserved annual ring was formed in AD 887.

Finally, the local chronology for Tune covers the period AD 709--892; both samples were dated. Neither of the samples had intact sapwood. The outermost preserved annual ring was formed in AD 892.



The result from Oseberg means that the building of the grave chamber can be dated to the summer of AD 834. The identification of the felling season corresponds well with the results of the botanical investigations which were carried out shortly after the excavation of the burial mound in 1904: 'The Oseberg mound was built during the course of the late summer or early autumn, in all possibility August or perhaps September' (Holmboe 1917: 205).

The date for the construction of the grave chambers at Gokstad and Tune cannot be determined with the same accuracy because none of the samples had sapwood preserved. The relative dating of the Gokstad samples shows, however, that the outermost annual ring lies very close to the border between heartwood and sapwood. In three of the samples the last annual ring was formed in AD 883 and in the fourth it was formed in AD 887. Three of the samples show no traces of tool marks (from axes etc.) on the surface where the sapwood joins the heartwood; they have not been worked. The indications are that it is just the sapwood rings which are missing. This observation requires further verification, e.g. from samples with sapwood present.

The felling date can therefore be calculated by adding an estimated number of sapwood rings which have rotted away to the last preserved annual ring in the heartwood.

As there is lack of available information about the number of sapwood rings in Norwegian oak trees, a simple arithmetic average of the number of sapwood rings in the five samples from Oseberg with complete sapwood preserved has been calculated. The average number is 14, the minimum being 10 and the maximum 16. This is clearly an inadequate data set on which to base such a calculation and the figure 14 will need to be confirmed by further work. However, although the number of tree-rings in the sapwood in oak trees in northern Europe is still open to debate, the small number of sapwood rings in the Oseberg material is supported by a number of investigations carried out in other regions which, like the Oslo Fiord area, are close to the limit of the natural distribution of oak in northern Europe (Baillie et al. 1985; Hillam et al. 1987; Wazny 1990; Bonde unpublished).

If we use the average number of sapwood rings from Oseberg as a basis for compensating for the missing rings in the samples from Gokstad we arrive at a felling date of around AD 900, presumably in the range AD 900--905; this is the date for the construction of the chamber.

The same technique can also be applied to the samples from Tune, where the material case is even more scanty. The outermost annual rings in the two samples, formed in AD 891 and AD 892 respectively, are also thought to lie close to the heartwood--sapwood boundary. If we compensate for the missing sapwood rings as above, the felling date for these trees is found to lie in the first decade of the 10th century, i.e. AD 905--910. Here too, the felling date also gives the date of construction of the chamber.


The dating of the three grave chambers provides us with a terminus ante quem for the objects in the burial mounds; it reveals that approximately two generations elapsed between the burial at Oseberg and the burials at Gokstad and Tune. The result also confirms Shetelig's conclusion in 1917 that the Tune and Gokstad burials can be regarded as contemporary. Accordingly, the start of the Early Oseberg Style must be fixed at the end of the second half of the 8th century, in agreement with the results of recent archaeological investigations in Denmark.

The start of the Viking period has traditionally been linked with the plundering of the monastery on the English island of Lindisfarne in AD 793. This was clearly an important event in the local community, but in the light of the results from the Norwegian sites presented here and of the results from the last 10--15 years of archaeological research carried out in Denmark, this date for the opening of the Viking period comes under increasing scrutiny (Hvass 1986; Bonde 1989; Jensen 1991; Christensen 1991; Bonde et al. 1990; 1992).

All the indications are that the period which in Nordic archaeology is referred to as the Viking period has its beginning much earlier than the Lindisfarne raid, presumably early in the 8th century. It is only the absence of written sources from this time which has until now justified the use of the Viking attack on the Northumbrian monastery as the beginning of the Viking Period. The material evidence tells another story!

With the dating of the Oseberg, Gokstad and Tune burials, archaeologists and art historians have been provided with yet another fixed point which can be used to date finds from the Early Viking Age. This result is not only of significance for objects from southern Scandinavia but also for those which are thought to originate in England and Ireland.

Acknowledgements. Dr Thomas S. Bartholin, University of Lund, Sweden, Prof. Dieter Eckstein, University of Hamburg, Germany, and Dr Tomasz Wazny, Academy of fine Arts, Warsaw, Poland, have kindly made their master chronologies available for this study. The dendrochronological dating of the Norwegian finds forms part of a project being carried out by the Natural Science Research Unit of the National Museum of Denmark into the chronology of the Viking Period.

The article was translated into English by David Robinson.


BAILLIE, M.G.L. 1982. Tree-ring dating and archaeology. London: Croom Helm.

BAILLIE, M.G.L. & J.R. PILCHER. 1973. A simple cross-dating program for tree-ring research, Tree-Ring Bulletin 33: 7--14.

BAILLIE, M.G.L., J. HILLAM, K.R. BRIFFA & D.M. BROWN. 1985. Re-dating the English arthistorical tree-ring chronologies, Nature 315: 317--19.

BARTHOLIN, T.S. 1985. Dendrochronology in Sweden, ISKOS 5: 489--97.

BONDE, N. 1989. Dendrokronologiske dateringer pa Nationalmuseet 1988 -- Dendrochronological datings, The National Museum, 1988, Arkaeologiske udgravninger i Danmark 1988: 229--41.

BONDE, N. & K. CHRISTENSEN. 1982. Trelleborgs alder: dendrokronologisk datering -- The age of Trelleborg: dendrochronological dating, Aarboger for nordisk Oldkyndighed og historie 1982: 111--52.

BONDE, N., T. BARTHOLIN, K. CHRISTENSEN, O.H. ERIKSEN & K. HAVEMANN. 1990. Dendrokronologiske dateringsundersogelser pa Nationalmuseet 1989 -- Dendrochronological dating at the National Museum of Denmark 1989, Arkaeologiske udgravninger i Danmark 1989: 245--64.

BONDE, N. & O. CRUMLIN-PEDERSEN. 1990. The dating of Wreck 2, the Longship, from Skuldelev, Denmark, NewsWARP 7: 3--6.

BONDE, N., A. DALY, O.H. ERIKSEN & K. HAVEMANN. 1992. Dendrokronologiske dateringsundersogelser pa Nationalmuseet 1991 -- Dendrochronological dating at the National Museum of Denmark 1991, Arkaeologiske udgravninger i Danmark 1991: 255--71.

BRATHEN, A. 1982. Dendrokronologisk serie fran vasta Sverige 831--1975, Riksantikvarieambedet Rapport 1982: 1. Stockholm: Central Board of the National Antiquities and National Historical Museums.

CHRISTENSEN, K. 1991. Wood-anatomical and dendrochronological studies, in M. Bencard, L.B. Jorgensen & H.B. Madsen (ed.), Ribe excavations 1970--1976 4: 169--81. Esbjerg: Sydjysk Universitetsforlag.

CHRISTIANSEN, T.E. 1982. Trelleborgs alder: arkaeologisk datering -- The age of Trelleborg: archaeological dating, Aarboger for Nordisk Oldkyndighed og Historie 1982: 84--110.

HILLAM, J., R.A. MORGAN & I. TYERS. 1987. Sapwood estimates and the dating of short ring sequences, in R.G.W. Ward (ed.), Applications of tree-ring studies: 165--85. Oxford: British Archaeological Reports. International series S333.

HOLMBOE, J. 1917. Botanisk undersokelse av torven i Oseberghaugen, in A.W. Brogger, H.J. Falk & H. Schetelig (ed.), Osebergfundet 1: 201--5. Kristiania (Oslo): Universitetets Oldsaksamling.

HOUGEN, B. 1934. Studier i Gokstadfunnet, Universitetets Oldsaksamling Arbog 5.arg. (1931--1932): 74--112.

HVASS, S. 1986. Vorbasse -- eine Dorfsiedlung wahrend des 1. Jahrtausends n. Chr. in Mitteljutland, Danemark, Bericht der Romisch-Germanischen Kommission 67: 529--42.

JENSEN, S. 1991. The Vikings of Ribe. Ribe: Den Antikvariske Samling.

KARLSSON, L. 1983. Nordisk Form om djurornamentik (Nordic form on animal ornament, summary). Stockholm: Museum of National Antiquities.

NICOLAYSEN, N. 1882. The Viking ship discovered at Gokstad in Norway. Kristiania (Oslo): Alb. Cammermeyer.

NYDAL, R. 1959. Trondheim natural radiocarbon measurements I, in American Journal of Science Radiocarbon Supplement (Radiocarbon) 1: 76--80.

ROESDAHL, E. & D.M. WILSON (ed.). 1992. From Viking to Crusader. Catalogue of the 22nd Council of Europe Exhibition.

SHETELIG, H. 1917. Tuneskibet, Norske Oldfund II. Kristiania (Oslo): Universitetets Oldsaksamling. 1920. Vestfoldskolen, in A.W. Brogger, H.J. Falk & H. Schetelig (ed.), Osebergfundet III. Kristiania (Oslo): Universitetets Oldsaksamling. 1926. Tidsbestemmelser i Vikingetidens Stilhistorie, Finska Fornminnesforeningens Tidskrift 36(1): 106--12.

STUIVER, M. & G.W. PEARSON. 1993. High-precision calibration of the radiocarbon time scale, AD 1950--500 BC and 2500--6000 BC, Radiocarbon 35: 1--24.

WAZNY, T. 1990. Aufbau und Anwendung der Dendrochronologie fur Eichenholz in Polen. Doctoral dissertation, Universitat Hamburg.

WILSON, D.M. & O. KLINDT-JENSEN. 1966. Viking art. London: Allen & Unwin.
COPYRIGHT 1993 Antiquity Publications, Ltd.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Bonde, Niels; Christensen, Arne Emil
Date:Sep 1, 1993
Previous Article:Identifying and protecting historic landscapes.
Next Article:Chickens in Africa: the importance of Qasr Ibrim.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters