Chapter 6

An analysis of the man and dragon combat from the Sigurd legend in the North of England to the first carvings of St Michael and the Dragon

At the start of this dissertation it was proposed that there was some form of connection between the late tenth and early eleventh century carvings of the pagan myth of Sigurd killing the dragon, and the subsequent Norman representations of St Michael fighting the dragon. It was further suggested that this development could take a closely connected form. Here Sigurd would be viewed as having a Christian symbolism, either specifically reflecting the story of St Michael, or simply representing good fighting evil. Alternatively, the connection could be a purely artistic one, with the decorative potential of the dragon motif appealing to sculptors of both periods.

It has been concluded that the Sigurd myth was largely independent of any Christian context, being principally a secular image. If there was any Christian symbolism, this seemed to involve the eucharist, and was not visually connected to the dragon slaying. Furthermore, it was unlikely that the image of St Michael was widely known to the sculptors or patrons of the Sigurd stones, consequently making the first proposal very unlikely. It therefore remains to examine the latter proposal, involving a brief study of some early Romanesque carvings of the St Michael scene, which incorporate Anglo Scandinavian elements.

It has been suggested by T.D. Kendrick that the Normans saw little merit in pre Conquest sculpture, especially in the early years after the Conquest Sculpture at this point was plain, regular, purely ornamental, and using a restraint appropriate to the monumentality of the massive new churches it adorned. He adds that with the impact of the continental sculpture movement on English work, early in the 12th century, Anglo Saxon influence had almost vanished, and contributed nothing to the development of English Romanesque sculpture Note 1. However G. Zarnecki argues that not only did Anglo Saxon motifs influence English sculpture, but that they were also influencing sculpture in Normandy Note 2. For example, the Church of William de Volpiano at Fecamp has capitals copying manuscript foliage derived from the acanthus leaf from the Winchester school. Also Saint Georges de Boscherville in Normandy has an exact copy of a capital in the Steyning Church, Sussex, influenced by the flat relief style of Anglo Saxon carving. While the continental influence provided new subject matter, often the Anglo Saxon designs and methods of decoration were as popular as the Romanesque options. This is probably due to the long standing English love of elaborate ornamentation, seen especially in the English use of Scandinavian interlace styles in the pre Conquest period.

The plain geometric ornamentation of the Normans was also taken over and made increasingly elaborate, involving foliage and grotesques, although often in the early stage retaining the flat techniques of pre Conquest work. Two prime examples of this assimilation are carvings of St Michael and the dragon at Southwell Minster and nearby Hoveringham. The dragon on the Southwell Minster tympanum faces left, with its single tongue flicking St Michael’s shield ( Plate 13). The dragon is heavily influenced by the Urnes style, especially in the tail and lip. The band which twists around the dragon’s body may be a second dragon, in the Urnes manner, with its head appearing above the main one. The forequarters and wings are Romanesque, though the eye is the same elliptical shape seen in Scandinavian based animals. The body is decorated with a lozenge pattern, with the overall effect being one of barbaric power. Although the beast does not reach above St Michael, this is due to the shape of the tympanum, and the dragon does dominate the composition. The figure implies vitality in his pose, facing front, but twisting to meet the dragon with outstretched shield and raised sword. Although St Michael’s ornate wings match the dragon’s, the lines of drapery on the robe are reminiscent of Romanesque free standing figures, falling in ‘hi” shaped pleats. Behind St Michael are David and the lion, depicted almost entirely in a Romanesque manner. The foliate and interlace decoration on the sides and base further reflect this mix of styles, with interlace reminiscent of the Norse Giant’s grave at Penrith, and also Romanesque foliage, and the ever popular acanthus leaf.

The tympanum at Hoveringham contains a very complex dragon, heavily influenced by the Urnes style in the tail, with Ringerlike foliate details ( Plate14). There is at least one subsidiary head, possibly two, appearing above the main head. Therefore there are secondary bodies twisted around the main body in the same way as the Southwell dragon. The forequarters are solid and balanced in the Romanesque manner, although the double tongue has an ambiguous origin, being Scandinavian or continental. Like Southwell, the dragon and St Michael share the same style of wings. St Michael also retains the same pose as at Southwell, with similar vertical “V” shaped drapery that could also be influenced by manuscript illustrations. While the Romanesque style Agnus Dei behind the Saint fulfils the same role as David at Southwell, in emphasising the source of St Michael’s power, Hoveringham also introduces a thoroughly Scandinavian dragon. This is under, and separate from, the narrative, but is a pure Urnes beast, and adds to the barbaric vitality of the whole scene.
The dating of these pieces, and some others included here, depend on the first arrival of tympana. The first tympana on the continent appear in the latter half of the 11th century, with the earliest figural tympanum being at Charlieu in Burgundy in 1094. Narrative scenes appeared around the start of the 12th century in England, possibly through the influence of the Cluniacs, who currently lead in terms of elaborate liturgy, chant and church decoration. It is clear that the Hoveringham and Southwell tympana therefore belong to the 12th century, and through the complex interaction of continental and Anglo Scandinavian influences, an early date is usually accepted, namely around 1120. However the dominant Scandinavian influences have, perhaps mistakenly, led to an alternative dating in the last quarter of the 11th century, despite the first known English tympanum, (which has the earlier purely geometric ornamentation) Note 3, being dated 1071 from Chepstow castle.

Another, much simpler version of St Michael and the dragon appear at Ipswich ( Plate 15), and is also influenced by the Ringerike style, the Scandinavian style prior to, (although also co existing with), Urnes. The Ipswich dragon possesses the “figure of eight” shaped tail of the early Urnes style, as well as the Ringerike spiral hip joint. The beast is generally more Scandinavian based, with the elliptical eye derived originally from the Ringerike style and continued in the Urnes. The dragon is placed in an unstable position, with both its elongated claws raised, and pressing against St Michael’s shield. St Michael also contributes to the impression that this piece is of a slightly earlier date than the Southwell and Hoveringham tympana, for the Ipswich St Michael faces ahead, with no vitality or imagination in portrayal. m e sword, while raised, does not slant towards the dragon in the impression of movement that the Southwell and Hoveringham examples convey. Furthermore, the wings are short and plain, the drapery is stylised herring bone mail, not the Romanesque folds of the early 12th century. The carving is simple, and the composition is basic. Runic inscriptions did cover the entire piece, but all that is left reads, in Old English,

“Here St Michael fought the dragon.”

The simplicity and lack of animation in this piece gives it less of a narrative feel, and more of a devotional aspect, perhaps with the spiritual fight of good against evil in mind. It has already been implied that this piece may be earlier than the Southwell Minster and Hoveringham examples. This is partially due to the Scandinavian influences, although it should be noted that Ipswich was in the Eastern Danelaw, where such influences would be expected to last a little longer. However, further evidence shows that this piece is not a tympanum, which would restrict its dating slightly, but possibly part of a screen Note 4. It is small to go over a door, and when it was removed from a wall in 1966, a cross was found carved on the reverse. Consequently it would have been viewed from both sides, hence making it likely to be a part of a screen. This opens up the dating possibilities, since it is known that the Normans often preferred to decorate smaller scale work. Important works include semi movable church fixtures, such as screens and fonts, although only the latter remain as a whole today. As a screen, it would have been visible to the secular congregation, hence clarifying the devotional aspect. This would also contribute to a date around 1100, when the narrative revival was just starting. Consequently this can be seen as a transitional piece, both in terms of looking to the animated narrative scene of Southwell Minster and Hoveringham, and in terms of figure style, with both angel and dragon lacking the decorative aspect they will shortly gain. The dragon will be embellished by the advanced organic interlace of the Urnes style, while St Michael will benefit from the interest in continental developments.

This argument would also contribute to the idea that the dragon was a popular motif for early Romanesque narrative scenes, since it allowed for elaborate adornment, of which the English were so fond. The Ipswich carving is limited both in terms of narrative excitement and decorative embellishment. The dragon, and especially its tail derived from Anglo Scandinavian influences, is the most confident element of the composition, before the understanding of the combination of drama and decoration are achieved in pieces such as the Hoveringham and Southwell Minster tympana.

In the North of England some evidence of the amalgamation of the two styles exists, although this appears very limited. The Anglo Scandinavian culture had still been particularly strong post Conquest, resulting in William the Conqueror in 1070 devastating the Yorkshire area in order to prevent any further rebellion. This oppression of the Yorkshire people included destruction of the means of life, including loss of homes, crops, livestock and agricultural tools. It also meant the execution of virtually an entire generation. Any form of normal life took decades to re establish, with the previous artistic importance this region held only returning with the Yorkshire school working in the third quarter of the 12th century. It would be possible, considering the above history, to argue that the early motifs involved in the return to the artistic practices may have been partially taken from what remained of the pre Norman heritage, and not from Norman designs.

A geographically isolated example in St Bees, Cumbria, is unusual in the threatening feeling of strength given to the dragon through the nature of the composition on the tympanum (Plates 16 & 17). The dragon is dominant through its size and the fact that it is placed over the warrior figure, leaving only the head and the sword arm visible. The warrior is further faced by the huge teeth in the open mouth of the dragon. The effect is not ornamental, as the Hoveringham and Southwell Minster dragons are, but almost realistic. It could be said that in the simplicity of the St Bees composition, the narrative scene is presented without decoration, the Anglo Scandinavian influenced interlaces being on either side. The dragon is strongly influenced by Norman art, although the figure behind has no attributes appropriate to St Michael. The figure is identified as St Michael because the carving appears to have come from the nearby ruined Chapel of St Michael. The carving is dated to 1120, which would be appropriate to the understanding of the narrative effect, and would still allow for the combination of Romanesque and Anglo Saxon elements. St Bees is unusual in that these elements, while both appearing in the same carving, are not amalgamated in one object. For example, whereas the dragon at Hoveringham has Romanesque forequarters and an Urnes tail, the St Bees dragon is essentially one style Romanesque. The warrior however could have come from any Anglo Scandinavian depiction of a warrior, (such as that at Brompton). The interlaces also, while being very simple, could also be found on many pre Conquest crosses.

Despite this, it can still be argued that it is the interest in the dragon that is the focus of the composition and this could be inspired by the long term popularity of the dragon motif in the North of England. This idea is backed up by the ferocity and primitive force in the dragon, despite its Romanesque style. Despite their clear Scandinavian heritage, the dragons from the Southwell Minster and Hoveringham tympana lack this strength, and consequently it can be argued that in some ways the St Bees dragon is closer to the nature of the Viking dragon than the Urnes influenced beasts from the South of England.

As opposed to this, a 12th century grave slab from Conisbrough shows very few Anglo Scandinavian influences, despite being closer to the carvings of the Sigurd legend, and assigned to the Yorkshire school based in York ( Plate 18). Despite the tongues with the dragon heads, the beast is almost entirely Romanesque. The forequarters have the same solid balanced feet and torso found in the Hoveringham dragon, while the beaded band running along the body is found in Romanesque foliate patterns, as are the leaves which sprout from the tail. Even the knot of the tail is not Scandinavian. Probably the closest tie to Scandinavian styles can be found in the decoration on the left of the slab. This is a modification of the beakhead, where a band flows from the mouth of a grotesque into more bands and acanthus leaves. The beakhead was derived originally from Scandinavian motifs, but developed through use in the South before gaining popularity in Yorkshire around 1160, dying out by 1180. The Conisbrough slab has been attributed to the Yorkshire school Note 5, and through the use of foliage and the beakhead, this slab can be dated to around 1170.

The use of Anglo Scandinavian styles in Norman carving seems to be a matter of regional taste. The areas in or near the Eastern Danelaw, (Ipswich, Southwell Minster and Hoveringham), show an easily recognisable use of Scandinavian motifs. The North of England, on the other hand, had also been heavily populated by Viking settlers and consequently it would be logical to look for similar evidence. This idea is strengthened by the interest in the Sigurd myth in the Yorkshire area. However, perhaps due to the extreme devastation of this area in the late 11th century, this does not seem to happen, even when the artistic culture has been re established. This can be seen in the lack of Anglo Scandinavian motifs or styles in the Conisbrough tomb chest from the Yorkshire school. This example reflects the idea that inspiration was taken from new developments in the South of England and through this, also from the continent. The St Bees stone shows some Northern interest in regional heritage, but any more widespread or uniform interest is very difficult to gauge, and may well have been a totally eclectic use of motifs from local pre Conquest monuments.

The nature of this dissertation was to analyse any apparent connections between the carvings of Sigurd and St Michael. It seemed at first that the carvings of Sigurd slaying the dragon were part of the Viking age combination of pagan myths with Christian counter parts. Yet it has been shown that there is no archaeological or documentary evidence to support any such connection between Sigurd and St Michael. Furthermore there is no reason for a symbolic connection from the Normans’ viewpoint, because this would involve identifying with a heathen myth. There is also no suggestion that the Norman society would have known the myth of Sigurd.

As an alternative it was suggested that it would be possible to trace the decorative styles used in the narrative compositions of the Sigurd story to the use of Scandinavian based ornamentation in early Norman representations of St Michael and the dragon. However it has been concluded that while this can be followed on a purely visual level, other considerations, such as the geographic locations of the e Temples, imply that this idea is not plausible. There is no evidence that in the North of England there is a conscious and logical use of Anglo Scandinavian motifs and subject matter. This includes the use of any remaining Sigurd designs in early Norman scenes of St Michael. This extends to the style of the dragon in these Christian designs. Although Scandinavian styles are used in the Hoveringham and Southwell Minster examples, these are not the same types of dragon as used in the Sigurd carvings. These latter ones are fairly limited in terms of using the full decorative style, for these are plainer beasts for a narrative composition. The Southern Norman dragons are more decorative, and use the Scandinavian based styles for ornamentation, to complicate as opposed to clarify.

To conclude, it may seem from visual and preliminary research that there was at least a connection grounded in the Scandinavian ornamental styles, but it can be argued from more detailed evidence that this was not so. However, the Scandinavian dragon, with its strength and barbaric nature, still left some impression on isolated northern carvings from the early Norman period. Such examples include St Bees in Cumbria, and Langtoft in Yorkshire which has an early Norman font. One scene on this font shows St Margaret bursting out of a dragon ( Plate 19), cruder, but similar to the St Bees carving.

These are isolated examples, but the popularity of the Sigurd myth continued in Scandinavian countries during the Romanesque period. This is reflected in both literature and carvings. Scenes from the Volsung saga are included on five portals of Norwegian State Churches. Hylestad, from the late 12th century, includes the entire story of Sigurd slaying Fafnir in clear narrative sequence. The 13th century portal at Lardal, shows a seated Sigurd, with two swords, plunging one into the dragon. It could be said that this is the Scandinavian equivalent of the English St Michael scene, used during the first centuries of the conversion to Christianity of the Scandinavian countries.


Note l – T D Kendrick, Late Saxon and Viking Art, London, 1949, p.l39 140. 
Note 2 – G Zarnecki, Studies in Romanesque Sculpture, London, 1979, I, p.101. 
Note 3 – T D Kendrick, Late Saxon and Viking Art, p.143. 
Note 4 – G Zarnecki, Studies in Romanesque Sculpture, I, p.100. 1 – 34 
Note 5 – G Zarnecki, Late English Romanesque Sculpture, 1140 1210, London, 1953, p.38. 1 – 37 

Scroll to Top