The 'Viking Shield' from Archaeology

Peter Beatson - NVG Miklagard


The large circular shield of the Vikings was part of a conservative tradition of manufacture. The best known intact examples from the Viking Age are those that lined the gunwales of the buried warship from Gokstad, Norway (Fig. 1) dated to c.905 AD (Bonde and Christensen 1993). They are similar to shields from Thorsberg bog (Raddatz 1987) and other Danish weapon deposits of the Roman Iron Age. Though archaeological evidence dries up with the adoption of Christian burial rites, art sources (such as the Lewis chessmen) indicate that kite shields were accepted in the Norse lands in the twelfth century, along with a small round buckler (Karlsson 1993). Scandinavian settlers seem to have adopted different (native Insular?) designs in the Irish Sea region, judging by material from burials there. These variants may be discussed in a separate paper.

NEWS - A New Viking Shield Discovery

September 2008 - Wooden parts of a near-complete Viking shield have been found by excavators at the Viking ring-fortress of Trelleborg near Slagelse in Denmark.
The 85 cm broad shield is made of seven fir planks, 8mm thick near the centre but thinning to 5mm at the edges. Patches of blue and white staining are visible on the boards and further testing may determine whether they are man-made, or just natural discoloration from the soil. Any leather facing which may have existed does not survive but near the rim is a row of small holes which could have fastened it, or a binding strip around the circumference. The central opening for the hand (the boss is missing) is somewhat oval and is crossed by a short wooden grip, which spans somewhat less than half the shield. The centre section of the grip is incised with basketwork-like interlaced design.

The shield is currently undergoing conservation and study at the Moesgård Museum near Århus. For images and to keep up with new developments, go here: Kongens Borge - News
NEWS - Viking Valkyrie found in England

In 2002 a metal detectorist uncovered this 40 mm high silver pendant near Wickham Market in Suffolk, England. Of Scandinavian origin and dated to the 9th cent., it shows a figure in a long gown, wearing a (?)crested helmet and carrying a sword and a shield. Unlike the several other known 'valkyrie' pendants the shield is shown from the rear, where a bar-like grip can be seen spanning its full width, the hand grasping it behind the boss. As the usual pinwheel decoration (see below) and dotted rim can also be seen, perhaps the artist has tried to show both faces at once.

More info: Portable Antiquities Scheme Annual Report 2001/2-2002/3 (British Council for Museums Archives and Libraries 2003), p. 58.
Question from a Reader

Q: I plan to build and decorate a viking shield, but historically correctly. But there is one thing I did not find on the internet: In which way are the planks butted together? Do you know anything about it? - Hendrick Brockmann

A: Thanks for the good question, Hendrik. It seems that the planks were not joined directly to each other. Most of the support comes from the leather cover glued to the front and back; but also the parts that were nailed to them, that is the handle which crosses some or all of the boards at the back plus the boss; and the rim, which was usually leather and stitched over the edge, or rarely small metal clamps.
There are dowelled planks in the Roman-age weapon hoards found in the Danish bogs, and some also had thin metal strips nailed along the joins, part decorative and part structural support. But nothing like this survives from the Viking period.
Remember that the boards are only very thin - 6-8mm, so that tongue-and-groove or dowel joins would be very fragile and not much use anyway. In my opinion and experience with reproductions the shield is designed on purpose to be a bit flexible rather than rigid, so instead of smashing the shield it absorbs energy from the blows. The shield is quite light and lively despite its size - you have to be a bit clever how you hold and use it to deflect weapons rather than block them. If you place it in front of you like a wall spears and axes will quickly poke holes in it!
To get this performance it is important to select only timber with straight grain (running from end to end). As you know the Vikings did not saw wood into planks like modern people, but split the logs along the grain for maximum strength and flexibility. The species may be important - linden (lime) wood was preferred in England and conifers such as fir in Scandinavia.
Good luck with your shield project! - cheers, Pete.

Construction and dimensions


Handle or grip

Edge Reinforcement

Other Fixtures

Combat Techniques




Construction and dimensions

Shields were typically 80-90 cm in diameter [1] (Table 1). The board was flat, and made of a single layer of planks butted together. The Gokstad shields were made of seven or eight white pine [2] planks of varying widths [3]. The planks were usually only 6-10mm thick (Table 2), and were bevelled even thinner at the outer edge (Fig. 1; Table 2). There is no archaeological evidence for laminated (ie. cross-ply) construction (Härke 1981) though contemporary poetry and slightly later legislation suggests it (Dickinson and Härke 1992; Nicolaysen 1882).

Figure 1 - Shield from Gokstad ship burial, Westfold Norway c.905AD. Diameter 94 cm (Nicolaysen 1882). a. Front. Boss type is Rygh 564. b. Reverse, note holes for attachment of rim and single wooden crossbar serving as grip - the other reinforcements seen in photographs are modern additions. c. Cross section, note bevelled edges.

The planks were possibly glued together. Extra support could come from the boss, grip and rim bindings (see below), and from a leather covering. At least some shields from Birka had a thin leather facing, and some earlier English shields were covered on both sides (Arwidsson 1986; Dickinson and Härke 1992). However, the planks of the Gokstad shields were painted, indicating that they had no leather facing covering them (Lowe 1990). It is worth noting that their uniform and fragile design suggests that the Gokstad shields may have been ornaments made especially for the burial, and thus not representative of actual combat shields [4].

An interesting parallel to the Gokstad shields comes from a peat bog at Tira, in Latvia. Dated to the ninth century, this near intact shield is constructed of six spruce or fir planks (Yrtan 1961) and covered on front and rear with leather, padded with pressed grass.


At the centre of the shield was a circular hole [5] covered by a more-or-less hemispherical iron boss of ~15 cm diameter (including flange), which enclosed the hand grip. The iron of the dome was fairly thick (3-5 mm), though the flange was somewhat thinner (Lowe 1990; Manx Museum, Douglas Man: pers. obs. 1994; Musee des Antiquites Nationales, St. Germain-en-Laye France: pers. obs. 1994).

Bosses had two main forms - the early style had a high dome and a pronounced neck (Type Rygh (R)564: Fig. 2-a). The later style, low domed without a neck (R562: Fig. 2-b), never completely replaced the former (Graham-Campbell 1980). Less common were a squat style (R563: Fig. 2-c) and a sub-conical style (R565: Fig. 2-d), sometimes with an apical knob (Arwidsson 1986).

Figure 2 - Shield bosses, Rygh classification scheme. a. R564. b. R562. c. R563. d. R565. From Arwidsson (1986).

Single examples of bosses with a toothed flange are known from Telemark, Norway (Fig. 3-a); Birka, Sweden; and Ile de Groix, France (Fig. 3-e). In the latter burial, some unique bosses with elaborate flanges were found (Fig. 3-b,c,d,e). These bosses might have had a Western European origin (Müller-Wille 1978).

The boss was normally attached by broad headed iron nails, the points of which were either clenched (bent over) or flattened on the reverse of the shield (Fig. 3-d,h). In the Birka material four nails was most common (Arbman 1940-3), occasionally six (as for the Gokstad shields). Five nails were sometimes used, as in examples from Cronk Moar, Man and the ship cremation on the Ile de Groix, France (Bersu and Wilson 1966; Müller-Wille 1978).

The flange of some bosses were angled, perhaps to secure the boss to the board by placing tension on the nails (Dickinson and Härke 1992), or possibly because they were attached to convex shield boards. Flanges with decorative edgings of non-ferrous metal strips were found in some Birka graves (Fig. 3-f,g), and nail heads were sometimes inlaid or tinned (Arwidsson 1986).

Figure 3 - Shield bosses. a. Boss with toothed flange, Telemark Norway (Oslo Olsaksamlingen, pers. obs.). b-e. Ile de Groix, France. Nail points were flattened rather than clenched (from Müller-Wille 1978). f. Birka Bj544, showing tin applique on flange; g. Birka Bj850, brass edging on flange; h. Birka Bj581, side view showing nails clenched (bent) for attachment (after Arbmann 1940).

Handle or grip

Wood alone must have been used in the majority of graves where remains are lacking, as in the Gokstad shields where a thin lath of rectangular section is nailed (crossways with respect to the planks) from edge to edge across the back face, it serves as a handle where it crosses the central hole (Fig. l). On more elaborate shields a wooden core was covered by a gutter-shaped sheathing of iron (Arwidsson 1986), usually ornamented with embossed bronze sheet or silver inlay (Fig 4-a).

Figure 4 - Shield grips, 10th cent. a. Two fragments of a silver-embellished iron grip with wooden core from Hedeby boat grave, Schleswig-Holstein Germany (from Müller-Wille 1976). b. Fragment of shield grip with spatulate terminal, Gokstad ship burial (from Nicolaysen 1886). c-d. Three-armed bronze fastenings for shield handle in the form of animal/human masks, Hedeby boat grave and Birka grave Bj944 (from Müller-Wille 1976, and after Arbman 1943).

The handle was long, often crossing the full diameter of the shield, and was tapered towards both ends. The tips could be flattened out into a spatulate terminal which was nailed directly to the board (Fig. 4-b), or be fastened down by separate bronze mounts (Fig. 4-c,d). Occasionally the nails fastening the boss also passed through the handle. The handgrip may have been wrapped with leather (eg. Birka grave Bj504, and as known from early Anglo-Saxon finds: Arwidsson 1986; Härke 1981).

Edge Reinforcement

Continuous gutter-shaped metal edge bindings like those known from Vendel, Välsgarde, and Thorsbjerg were obsolete by the Viking Age. In the vast majority of finds there is no evidence of edge reinforcement, which must therefore have been absent, or of a perishable nature. On the Gokstad shields, small holes are bored about 2 cm in from the edge, at intervals of c.3.5 cm (Fig. 1-a,b), presumably to fasten a rim, all other traces of which have perished. It can be speculated that the edge was bound with a leather strip fastened with stitches or thongs, or possibly very fine iron nails.

Figure 5 - Metal clamps from shield rims. a. Grave Bj944, Birka Sweden. Type A, simple U-shaped clamp. b. Grave Bj369, Birka. Type B with expansion for leather rim binding. c. Lindholm Høye 1112, Denmark. Raised punchmarks surround the rivet heads (a, b. from Arwidsson 1986; c. per. obs. Lindholm Høye Museum, Aalborg Denmark 1994).

Small clamps made of iron or bronze sheet are occasionally found in graves (Table 3; Fig. 5). Clamps were sometimes simply decorated by tinning, punching or engraving (Fig. 5-c). In Birka graves Bj 628 and 736 the clamps were butted to produce a continuous edge (Fig. 6), however, only sections of the rim survive, perhaps indicating deliberate damage before burial.

Figure 6 - Birka grave Bj736, 10th cent. a. Shield remains as found on excavation (after Arbman, 1943). b. Reconstruction of shield, by author.

Sometimes several clamps are distributed evenly around the shield rim (Bj 842, Valsgarde 12), perhaps to fasten a leather edge binding, traces of which sometimes remain. Clamps from grave Bj 850 were fastened over a leather edging (Fig. 7), though their low number and uneven distribution suggests that this was not their primary purpose. Here they might have fastened joins between planks, or shored up a damaged edge.

Figure 7 - Birka grave Bj850, 10th cent. a. Shield remains as found on excavation (after Arbman, 1943) 1 = boss; 2 = rim clamps; 3 = grip terminal, next to remains of grip. b. Reconstruction of shield, by author. c. Cross section of a bronze rim clamp from Bj850, enclosing organic material from the shield, including leather facings and edge strip (after Arbman, 1940).

Other Fixtures

Other metal fittings from shields, including nails (Fig. 8-a) are occasionally recovered. Some Birka graves contained one or two small rings held by eyelets (Fig. 8-b,c) which passed through the boards, and sometimes also the handle, with the ring projecting on the rear side (Arwidsson 1986). They may have served to hang up the shield, or as attachment points for a guige strap.

Figure 8 - Other metal shield fittings. a. Birka grave Bj727, ?10th cent. Clench nail with bifurcated shank used to fasten grip to shield. Most nails were however the normal type. b. Birka grave 407. Ring and staple similar to those from Birka shields. c. Schematic distribution of eyelets with rings on reverse of Birka shields, grave numbers indicated. a, b. from Arbman (1943).

In the 11th C. Valsgarde 11 burial, a shield appears to have been repaired by nailing 13 thin brass strips (15-30 by 6-7 mm) across the break (Museum of Norse Antiquities, Uppsala Sweden: pers. obs. 1994).


Archaeology as well as literary and art sources indicate that the shield was often painted. The faces of the Gokstad shields were painted yellow (?orpiment = As2O3) or black (?charcoal), and arranged alternately along the ship's sides (Lowe 1990; Nicolaysen 1882). Red shields may have been popular [6]. A red shield is mentioned on a Danish runestone (Roesdahl 1992), as well as in several sagas. Distribution of a pigment layer in the Viking Age Välsgarde 9 grave indicated a red painted shield (G. Hedlund, Uppsala Universitet: pers. comm. 1993). Shields from the Roman Iron Age weapon sacrifice at Thorsberg were painted red or blue (Nationalmuseet, Copenhagen: pers. obs. 1994).

Fragments from Ballateare, Man suggest that the leather facing of this shield was painted with black and red patterns on a white background (Fig. 9). It was suggested that a gesso (organic matrix, such as egg yolk) paint was used (Bersu and Wilson 1966). Traces of white paint were found on a wooden fragment from the Manx Cronk Moar shield (Bersu and Wilson 1966).

Figure 9 - Fragment of gesso from shield face, Ballateare, Man 10th cent., after Bersu and Wilson (1966).

A recently discovered 10th C. chamber grave at Grimstrup, Denmark contained a circular wooden board which covered the corpse from head to hip (unpublished: I. Stoumann, Esbjerg Museum, Denmark, pers. comm. 1994). As no other traces (ie. boss) were found in an otherwise fully equipped male burial, it has been suggested that the board is a 'blank' or unfinished shield. The board was elaborately painted with interlace patterns (Fig. 10), though the overall design is no longer discernible. The background colour is dark blue, the interlace is grey-green edged with white lines. Some lines of red paint and white dots are also visible.

Figure 10 - Possible shield from Grimstrup chamber grave A, Denmark 10th cent. Two details of design painted on board. After photos on exhibit at Esbjerg Museum (pers. obs. 1994).

Representations of shields in Viking art (Fig. 11) are frequently marked with 'pinwheel' patterns of radiating curved lines (Fig. 11-a,b,c,d). These might possibly represent metal strengthening bands (unknown from archaeology but required in later law codes for levy equipment: Dickinson and Härke 1992; Nicolaysen 1882); or even seams in the leather facing; or may mark segments originally painted in contrasting colours, as shown in a few contemporary Frankish manuscripts (Fig. 12). Inspiration for decoration of a reconstructed shield might also be sought in surviving painted wooden objects from the Viking Age [7].

Figure 11 - Shields in contemporary art. a, b. Picture stones, Gotland 7-8th cent., after Magnusson (1979). c, d. Silver shield pendants, Birka Sweden 10th cent., after Duczko (1989). e. Bronze 'valkyrie' pendant, Hedeby Germany 10th cent., after Elsner (1985). f. From tapestry fragments, Oseberg Norway ante 834AD, after Hougen (1940).

Figure 12 - Spirally marked and segmented shields in the Golden Psalter of St. Gall, Frankish ms., 10th cent., From Duby (1970).

Sheet metal decorations in the form of beasts or birds fixed to the shield face are confined to the preceding Vendel period, though an applique of wooden strips was suggested for the Cronk Moar shield (Bersu and Wilson 1966). Some examples of decorated metal parts (bosses, grips, clamps) from Viking Age shields have already been mentioned above.

Combat techniques

Analysis of battle damage to weapons from the massive Roman Iron Age deposit of Nydham indicated the primary use for the large round shields was in fending off missiles, while sword duels were conducted blade on blade (Schloß Gottorf: Archäologische Landesmuseum der Christian-Albrechts Universität, Schleswig Germany: pers. obs. 1994). However, the use of shields in hand to hand combat is recorded in customs such as the holmgang duel. The heavy iron construction of the Viking Age boss is unlike the Roman Iron Age examples of thin bronze, perhaps indicating a change to a hand-to-hand fighting style in which parrys with the boss were possible. The thin boards would split easily, and could perhaps have been deliberately made so, in order to snare an attacker's blade.

Table 1

Table 1. Estimated diameters of Viking age shields from archaeological finds. All dated 10th century, except Tira (9th cent.) and Krimylda (11th cent.).

Table 2

Table 2. Thicknesses of Viking age shields from archaeological finds. Most of these graves are probably 10th century, except Tira (9th cent.). For comparison, shields from (much earlier) pagan Anglo-Saxon graves average 0.75 cm thick (103 examples: Dickinson and Härke 1992).

Table 3

Table 3. Metal clamps from shield rims: numbers, dimensions, distribution on shield rim (if known).


ARBMAN, H. (1940). Birka I: Die Gräber. Untersuchungen und Studien. Tafeln. Kungl. Vitterhets Historie och Antikvitets Akademien (KVHAA): Stockholm.

ARBMAN, H. (1943). Birka I: Die Gräber. Untersuchungen und Studien. Text. KVHAA: Stockholm.

ARWIDSSON, G. (1986). 'Schilde'. In: G. Arwidsson (ed.). Birka II: Systematische Analysen der Gräberfunde, vol. 2. KVHAA: Stockholm.

BERSU, G. and WILSON, D.M. (1966). 'Three Viking graves in the Isle of Man'. Society for Medieval Archaeology, monograph 1. Society for Medieval Archaeology: London.

BONDE, N. and CHRISTENSEN, A.E. (1993). 'Dendrochronological dating of the Viking Age ship burials at Oseberg, Gokstad, and Tune, Norway'. Antiquity 67, p. 573-583.

BRØNSTEAD, J. (1936). 'Danish inhumation graves of the Viking Age'. Acta Archaeologica 7, p. 81-228.

CHRISTENSEN, A.E. (1993). 'Kongsgårdens håndverkere'. In: A.E. Christensen, A.S. Ingstad and B. Myhre (eds.). Oseberg Dronningens Grav: Vår arkeologiske nasjonalskatt i nytt lys. Schibsted: Oslo, p. 85-137.

DICKINSON, T. and HÄRKE, H. (1992). 'Early Anglo-Saxon shields'. Archaeologica 110, Society of Antiquaries of London: London.

DUBY, G. (1970). Histoire de la France: naissance d'une nation des origines à 1348, vol. 1. Libraire Larousse: Paris.

DU CHATELLIER, P. and LE PONTOIS, L. (1908-9). 'A ship burial in Brittany'. Saga Book of the Viking Club 6, p. l23-161.

DUCZKO, W. (1989). 'Runde Silberblechanhänger mit punzierten Muster'. In: G. Arwidsson (ed.). Birka II: Systematische Analysen der Gräberfunde, vol. 3. KVHAA: Stockholm.

ELSNER, H. (1985). Wikinger Museum Haithabu: Schaufenster einer frühen Stadt. Karl Wachholz Verlag: Neumünster.

GRAHAM-CAMPBELL, J. (1980). Viking artefacts: a select catalogue. British Museum: London.

HÄRKE, H. (1981). 'Anglo-Saxon laminated shields at Petersfinger - a myth'. Medieval Archaeology 25, p.141-144.

HOUGEN, B. (1940). 'Osebergfunnets billedvev'. Viking 4, p. 85-124. Oslo.

KARLSSON, U. (1993). 'Medieval round shields'. New Hedeby (January 1993), p. 26-27.

LOWE, S. (1990). 'Everything you wanted to know about Viking shields (and one helmet) but were afraid to ask'. Varangian Voice (issue 17), p. 24-25.

MAGNUSSON, M. (1979). Viking: hammer of the north. Orbis: London.

MARXEN, I. and MOLKTE, E. (1981). 'The Jelling man: Denmark's oldest figure-painting'. Saga - Book of the Viking Club 20, p. 267-275.

MULLER-WILLE, M. (1976). 'Das Bootkammergrab von Haithabu'. Berichte über die Ausgrabungen in Haithabu 8. Karl Wachholtz: Neumünster.

MULLER-WILLE, M. (1978). 'Das Schiffsgrab von der Ile de Groix: ein Exkurs zum Bootkammergrab von Haithabu'. Berichte über die Ausgrabungen in Haithabu 12, p.48-84.

NICOLAYSEN, N. (1882). The Viking ship discovered at Gokstad in Norway. Christiana: Oslo (reprinted 1971 Gregg International Publ.: Westmead UK).

OWEN, O. and DALLAND, M. (1994). 'Scar, Sanday: a Viking boat-burial from Orkney'. Birka Studies 3, p. 159-172.

RADDATZ, K. (1987). Der Thorsberger Moorfund Katalog: Teile von Waffen und Pferdegeschirr, sonstige Fundstücke aus Metall und Glas, Ton- und Holzgefuße, Steingeräte. Karl Wachholtz: Neumünster.

RAMSKOU, T. (1976). 'Lindholm Höye gravpladsen'. Nordiske Fortidsminder Ser.B, Bind 2. Lynge & Son: Copenhagen.

ROESDAHL, E. (1982). Viking Age Denmark. British Museum: London.

THORVILDSEN, K. (1957). 'Ladby-skibet'. Nordiske Fortidsminder, Ser.1 v.6. H.J. Lynge & Son: Copenhagen.

YRTAN, V.A. (1961). 'Drevnie shchity na territorii Latviiskoe SSR'. Sovietskaia Arkheologii 1961(1), p. 216-224.


[1] For comparison, dimensions of pre-Viking shields: from pagan Anglo-Saxon graves (23 examples) 42 to 92 cm diam. (Dickinson and Härke 1992); Thorsberg moorfind, Denmark (7 examples, Roman Iron Age) 65 to 104 cm diam. (Raddatz 1987); Välsgarde, Sweden (3 examples, Vendel period) 84 to 110 cm diam. (Arwidsson 1986).

[2] Softwood from conifers seems to have been used in most, but not all cases.

[3] A smaller number of broader planks would seem more practical: eg. the pine central plank of a pre-Viking shield from Välsgarde was 52 cm wide (Dickinson and Härke 1992).

[4] A grip fragment from a more ornate shield was found within the grave chamber (see Fig. 3-b.).

[5] Circular on the Gokstad shields at least. Oval, 'figure-8', and 'D'-shaped openings are known from pre-Viking material (Dickinson and Härke 1992; Härke 1981). The second (fragmentary) shield from Tira, Latvia had a quadrangular opening (Yrtan 1961).

[6] Red pigments in ancient paints seem to derive from mineral sources ie. red ochre (Fe2O3, as on the Jelling figurine: Marxen and Molkte 1981); or cinnabar (HgS, as on the Illerup shield of c.200AD: Forhistoriskmuseet, Moesgard Denmark: pers. obs. 1994). Also on the Jelling figurine were a dark blue paint made by mixing powdered white chalk with burnt organic matter (charcoal?), and a yellow of orpiment (As2O3) in an oil base.

[7] Examples: Jelling figurine (see Footnote 6) and associated fragments; board with snake design from Hørning church, Denmark and painted runestone from St. Pauls churchyard, London (Graham-Campbell 1980); numerous painted objects from Oseberg and Gokstad, Norway (Christensen 1993; Nicolaysen 1882); painted board from Ladby ship burial (Fyn, Denmark: Thorvildsen 1957); casket in Birka grave Bj639 (Arbman 1943), see Fig. 13 below.

Figure 13 - Painting on small casket from Birka grave Bj639 (after Arbman 1943, colours added according to notes). The style resembles the Ballateare shield. a. From 19th c. sketches by Hjalmar Stolpe b. Paint flakes still on wood.


See the LATVIAN SHIELD (in case you missed it).

Peter Beatson, 1995-2010. All rights reserved by the author.

Webbed by Peter Beatson, 2000. All rights to materials herein belong to the original owners.