Volume IV: South-East England

Select a site alphabetically from the choices shown in the box below. Alternatively, browse sculptural examples using the Forward/Back buttons.

Chapters for this volume, along with copies of original in-text images, are available here.

Current Display: Winchester (Old Minster) 88, Hampshire Forward button Back button

Overview
National Grid Reference of Place of Discovery
Present Location
Winchester City Museum, The Square, Winchester, accessions no. 2943 WS 98
Evidence for Discovery
Found in archaeological excavation north of Winchester cathedral in 1965 in rubble filling of robbed external eastern crypt of Old Minster; Final Phase 59 (Provisional Phase 644), c. 1094
Church Dedication
Old Minster
Present Condition
The block is intact except for the top left-hand corner of face A and a larger section missing from face C. In consequence, the left-hand figure on face A has lost his chest, left hand, and right leg
Description

Only face A is decorated. The figures extend beyond the stone on all sides and must have been continued on adjacent blocks (see Discussion). The remaining faces are undecorated and all except the back (face C) must have been in direct contact with other stones. Face B carries a double rebate (2.5 by 5 cm; 5 by 7.4 cm) and face D a single rebate (5 by 10.2 cm). There is nothing to suggest that these features are necessarily secondary; their tooling remains crisp. Faces E (top) and F (bottom) are more roughly tooled and the former seems worn. Face C is smoothly dressed.

A (broad): A slightly curved rectangular hole, probably a lewis-hole, at the centre of this face is filled with pink plaster. The outline of the seat of the mailed figure is skilfully accommodated to the convex side of the hole, perhaps following an early change in design, before the outline of the figure had been settled, in which the sword was shortened and narrowed, as the scar of its former outline shows.

The figures rise c. 7 cm above the flat background. Since their surface is also generally flat, they have a somewhat slab-like, 'blocky' appearance. This is particularly noticeable along the back of the mailed figure, where the angles between the background, the back, and the surface of the figure are almost right angles. The angle between the background and the figures is emphasized throughout by a slight groove, giving a clean outline (a similar groove is found on the Repton Stone: Biddle and Kjølbye-Biddle 1985, 241).

The left-hand figure appears to be walking to the left. He wears a coat of mail from the shoulders to the middle of the thighs, which are distinctly separate. There is no sign of any covering on the rest of the legs. The mail is shown by alternately recessed squares and is finished on each thigh with a plain band, probably to indicate edge binding. On the left leg the band continues round the inside of the thigh as far as the background of the sculpture, while on the right leg the band continues back to butt the left leg at a higher level than that leg's band: clear indications that separate leggings are being shown, each with its own edge binding. The mail continues up below the arm on to the chest and over the shoulder; the back is plain, but scarcely visible. The mail extends on to the upper part of the arm where it ends at the first of a series of rucked bands. These continue down the arm to the wrist, but at the elbow, where the relief is exceptionally high, they have been almost worn away.

The fingers of the left hand are carved on the left edge of the stone (Face D: Ill. 649) which here forms part of the relief. The knuckles and back of the hand are broken away, but the relative positions show that the hand, although clenched, does not grasp the sword. Instead the wrist passes in front of the hilt. The broad sword is carried in a scabbard on the left hip and hung from a belt. The belt is clearly marked round the back of the figure, although the pattern indicating mail is omitted in this relatively invisible area. A thinner strap runs down from the back of the belt, without any sign of a junction between the two, to join the scabbard at a point where a crossed binding is shown. Both the belt and angled strap pass behind the scabbard. What was previously seen as a quillon (Biddle 1966b, 329) is part of the belt which widens slightly as it approaches the scabbard: the guard and pommel are invisible, except for a small part of the grip or possibly the lower edge of the pommel, which survives just above the wrist. As noted above, the sword was originally roughed out both longer and broader.

The figure on the right is shown in false perspective, lying on his back with his hands, palms upwards, raised to either side of his face. He is bound round his neck with a single broad tether which twists over his right wrist and continues towards the bottom edge of the stone. The lower part of the tether is broken away, but there is the scar of both sides of a wider feature, such as a post, to which the tether might have been fastened. The tether does not continue up around the left arm; rather it appears to return round the neck. The left arm, lying over the animal's paw and under its muzzle, with the fingers just appearing beyond, seems to be free and may be attempting to ward the animal off.

A series of shallow curving lines shows that the man's hair is worn shoulder-length, tucked behind the ear, as if tied back. The ear, jaw-line, and chin are well preserved up to and including the lower lip. The face has been partly removed by two distinct adze cuts, but the upper and lower lips are intact, except for a slight loss where they join, and they show that the mouth is open. The outlines of the nose and forehead can be traced where they have been broken away, but a tiny part of the right-hand edge of the right nostril survives where it projects from the upper lip. There is no sign of a moustache. A deep accidental scar marks the cheek, but the prominent brow ridge and the greater part of the eye are preserved. The upper edge, outer end, and outer part of the lower edge of the eye, and part of the inner end survive, showing that the eye was lentoid. The brow ridge returns sharply down the inner end of the eye before being broken away (cf. Winchester (Old Minster) no. 75 (Ills. 618–19)).

An animal is shown on top of the human figure, its muzzle at the man's face, and one of its forepaws appearing from beneath the man's left hand to rest on, and force or hold open, his lower jaw. The paw has three bulbous toes, each with a claw; a fourth toe, higher up and clearly separated from the other three, is a dew-claw, showing that this is a canid, i.e. a dog, fox, or wolf, and specifically not a bear. Roughening on the neck suggests a mane; the ear(s) are indicated, but, together with the upper part of the face, are broken off, only the clear outline of the forehead and top of the muzzle surviving over the fingers of the man's left hand.

The animal's mouth is open but not biting and reveals the upper and lower tooth rows, where the individual teeth are separated by small (drilled?) indentations. The outline of the lower jaw is intact, but the extremity of the upper jaw is broken away, together with any trace of the canine teeth (fangs) in either jaw. The animal's tongue is shown the length of the open mouth. Although partly broken away, the tongue crosses the gap between the animal's muzzle and the man's face and descends in a curve up to, and apparently into, the man's open mouth (pace Wilson 1985, 206, 'the critical portion of the stone' is not so damaged as to make this unclear; it can be photographed with appropriate lighting (Ills. 646–7)). If the broken ridge passing into the man's mouth is not the animal's tongue, it can only be the man's tongue projecting to touch the animal's. There is no surviving trace of any distinction between the two, and the proposition seems too unlikely to maintain.

Composition The vertical centre line of the stone passes tangentially by the crown of the recumbent man's head and through the lower tip of the sword, effectively splitting the stone into two scenes. The horizontal centre line bisects the vertical axis just below the crown of the man's head, continues along the straight part of the hair and the upper edge of his ear. To the left it runs through the sword, body, and clad thighs of the standing figure at no immediately obvious point. However, a grid with intervals of 173 mm divides the width of the stone into thirds and the height into quarters. A line drawn from the top left corner of the stone (as reconstructed) to a point two-thirds along the bottom from the left corner runs exactly along the left edge of the sword, the most prominent diagonal line in the composition. This line represents the golden proportion, dividing the area into three eighths and five eighths. A line drawn parallel to the left, from the lower right corner to a point one-third from the upper left corner (as reconstructed) passes through the knot of the tether, touches the point of the ear and the eye-brow of the recumbent man, and runs along the slope of the shoulder of the standing man. If the diagonals are drawn in the opposite direction, that running from a point one-third from the lower left corner to the upper right corner runs along the tongue of the wolf and perhaps once along the line of the mouth of the man. This is the strongest line in this direction and lies parallel to the rope around the neck.

Reconstruction If the warrior on the left is completed with a head the same size as that of the recumbent man (who may actually be somewhat larger), one module of 173 mm would be sufficient. To complete the recumbent man at least four modules would be needed, which suggests that more than one stone of the size of the present block is required. Were there to be a border above and below the narrative scene, stones measuring 2 modules (346 mm) in height would be needed, each half as high as the central stone. In the Bayeux Tapestry the narrative frieze occupies two-thirds of the height with a border above and below, each one-sixth of the total height. The proportions of the Winchester frieze may well have been similar, with the figured frieze taking up perhaps six modules, and any border above and below one module each. On this calculation the frieze would be 138.4 cm high, virtually 4.5 English feet or 4 'northern' feet (in length around 34 cm).

Discussion

The date of this piece depends in the first instance on the archaeological context in which it was discovered. The block was found in the rubble filling of the external eastern crypt of Old Minster (Biddle 1966a, 325; idem 1966b, 329, pls. LIXb, LXV), in layers derived from the levelling of the robbed east end. The block was covered by 75 cm of rubble and it is quite certain that it was discarded at the demolition of Old Minster in 1093–4. It cannot therefore have come from New Minster which was not demolished until c. 1110. The demolition materials of New Minster were quite different in character and (in this area) spatially distinct from those of Old Minster. It might be argued that Winchester (Old Minster) no. 88 was discarded during the construction of the present cathedral and thrown away among the rubble from the demolition of Old Minster which was in progress at the same time. Extensive layers of oolite chips and the discovery of a few mouldings of Norman type in the demolition materials of Old Minster show that its site was being used as a quarry where Anglo-Saxon stones were dressed for reuse. Many of the carved fragments from Old Minster in this Corpus are chips from precisely this process. But these chips include fragments of hair (nos. 76–7 (Ills. 622, 625) and limbs (nos. 79–81 and 84 (Ills. 620, 626–7, 630–1, 635–8)) derived from figures similar to those on the present piece. These chips, the general bruising of no. 88, and the absence of any obvious damage which might have led to its abandonment, all indicate that no. 88 once formed part of the decoration of Old Minster and was discarded in 1093–4, along with a few other large blocks (no. 4 (Ill. 503), and no. 62 (Ill. 592) which were lost in the muddle of the robbing.

The carving appears to show parts of two scenes, the subjects divided here, as sometimes on the Bayeux Tapestry, by the first or last figure of a scene turning its back on the scene which precedes or follows (Wilson 1985, pls. 9, 20–1, 24–5, 46, 49–50, 55, 63, 72). Here a mailed warrior walks left into a scene beyond the surviving stone, his back turned to the scene on the right and separated from it by an open field. As shown above, at least two blocks of similar size would be needed to complete the figure of the bound man, and if the scenes are of comparable size, the mailed warrior can only have been one of several figures in a scene extending to the left over at least two further blocks. Five blocks of similar size would thus probably have been needed to complete the scenes partly preserved on the present carving, indicating a minimum length of 2.6 m.

Nor are the figures on this block complete in themselves: portions of the top and bottom of each scene must have been shown on blocks above and below. As indicated above, these blocks need have been no more than about 17.5 cm high, but the disparity in size between this and the main block, 69.5 cm in height, may suggest that the upper and lower blocks were larger and carried a running border above and below. If these borders were also about 17.5 cm high, the top and bottom blocks would each have been about 35 cm high, and the whole scheme on three blocks about 140 cm high, the central block accounting for just half the total.

There are only a few purposes for which a carved panel at least 240 cm long and possibly 140 cm high can have been intended: as decoration for a free-standing monument of royal proportions; as part of a narrative frieze round the interior or exterior of Old Minster; or as the decoration of a free-standing screen. The rebating of the sides of the block is relevant here, for it might seem more suited to interlocking the sides and corners of a free-standing tomb or screen than to a frieze continuous in the plane of a wall. But the double rebate, perhaps suitable for keying a corner, lies behind the bound man, at a point where it would seem impossible to imagine a right-angled turn. The size of the scheme indicated by the surviving block seems also too large even for a royal tomb at this date. The smoothness of the back and the use of rebates to key adjacent blocks could be features of a free-standing screen. The possibilities seem therefore to suggest a narrative frieze, either on a screen or on an interior or exterior wall face.

But was the sculpture carved on a newly quarried stone or on a reused block taken from a Roman structure? If the rebates and smooth back derive from a previous use, they are irrelevant to the possible setting of the stone after it had been carved. The lewis-hole might indicate a Roman date, but it seems to be only an assumption that lewises were not used in the Anglo-Saxon period (see Chap. VIII). The central position of the lewis-hole suggests either that the shape of the Roman block, if that is what it is, has not been much changed, or that the lewis was used to move the block selected for this carving. The flat surface of the carvings indicates that they were cut on a surface already flat and square to the rest of the stone, but this could reflect the shape of the block as it came from the quarry, and need not imply that the block was reused. Very large blocks of oolite for making coffins must have been brought overland in the Anglo-Saxon period from the quarry some 80 miles away, for it seems highly unlikely that these could have been cut on stones recovered from Roman buildings. Roman architectural fragments from Winchester are usually of different geological types, including sandstones and limestones as well as oolitic limestone (see Chap. VIII), while the overwhelming preponderance of carved stones from Old Minster in this Corpus are of Combe Down Oolite, as here. Since no. 88 is not therefore necessarily a reused Roman block, the rebates and smooth back may be significant in discussion of its Anglo-Saxon use.

The distinctive feature of the iconography is the dog or wolf with its tongue at or in the open mouth of the bound man. The scene may be from some lost or unidentified story, such as a saint's life, but an incident in Völsunga Saga provides a close parallel, as was suggested when the stone was first published (Biddle 1966b, 330–1). Sigmund and his nine brothers were clamped by their legs into a large pair of stocks in the forest. For nine successive nights a large and evil-looking old she-wolf appeared, and killed and ate one of the nine brothers until Sigmund alone remained. On the tenth night Signy, Sigmund's twin sister, sent her trusted servant to smear honey on Sigmund's face and to put some of it in his mouth. When the wolf came she sniffed the honey, licked Sigmund's face 'and then thrust her tongue into his mouth. He took heart and bit into the wolf's tongue. At this she gave a violent jerk and strained backwards, pressing hard with her paws against the stocks which as a result split apart. But he held on so firmly that the wolf's tongue was torn out by the roots, and that finished her.' (Finch 1965, 7–8).

The armed figure to the left might be Signy's trusted servant, but it seems unlikely that the story was shown in such detail, and some earlier episode (if this is a narrative frieze) is probably represented. There is one difference between the saga and the sculpture, for in the latter the bound figure is shown lying on (and apparently tethered to) the ground, rather than seated in the stocks, as the only manuscript specifically states (Finch 1965, 7). But such stocks should not be thought of as incorporating a wooden seat: early stocks consisted of two large baulks of timber, cut to fit over the legs, and fastened by iron clamps. The prisoner sat or could lie on the ground (Utrecht Psalter, c. 820 (De Wald 1933, pls. 98, 120, 130); BL MS Harley 603, fol. 54v, c. 1000 (Temple 1976, no. 64, ill. 206)). Whether or not the stocks were shown on his legs, the essential point is that the wolf is at the man's face in precisely the way the saga specifies - licking or thrusting and not biting.

Although the only manuscript of Völsunga saga dates from c. 1400, and 'the saga itself was compiled not later than c. 1260–70, probably in Iceland, though possibly in Norway' (Finch 1965, ix), references to the Voölsung legend in Beowulf, Widsith, and Waldere in Old English (Kennedy 1943, 45–8), and in Eiríksmál in Old Norse, show that the story was well known at the latest by the tenth century and probably well before. Scenes from the Völsungar cycle relating to the deeds of Sigurd have been found carved on wood or stone in Sweden, Norway, England, and possibly the Isle of Man (Blindheim 1972; Ploss 1966), but no representation of the incident of Sigmund and the wolf appears to have been recorded.

Since the royal houses of Wessex and Denmark claimed descent from the same ancestor, Scyld, and thus shared a tradition in which Sigmund had played a part, Biddle suggested in 1966 that no. 88 might be part of a narrative frieze celebrating the shared origins of the two royal houses: 'There could be no more suitable setting for such a frieze than the eastern arm of the Old Minster, in which . . . the royal burials probably lay. Among these was Cnut himself.' (Biddle 1966b, 331). This interpretation has often been repeated, not always with conviction (Davidson 1967, 127; Gatch 1971, 33–5; Cramp 1972, 148; Lang 1976, 94; Jacobs 1977, 40, n. 81; Hinton 1977, 95–6; Brooks 1978, 96; Dodwell 1982, 137–8; Zarnecki 1984, 150–1 (cat. no. 97); Wilson 1984, 198–200; Wilson 1985, 206–8; Zarnecki 1986a; Zarnecki 1986b, 8 and n. 7; Kahn 1992), but, as Wilson notes 'nobody has produced a better explanation' (Wilson 1984, 200).

Two alternatives have since been suggested. The first, biblical, has not been argued in detail. It suggests that the subject is likely to be dogs licking the blood of Naboth, stoned on the order of Jezebel, as recorded in I Kings 21 (Zarnecki 1986b, 25, n. 7 (suggestion by Jolanta Zaluska)). This subject is excessively rare at any date, does not occur in surviving Anglo-Saxon manuscript illumination (Ohlgren 1986; but neither does Sigmund and the wolf), and has no discernible relevance to Old Minster or to Winchester: it celebrates the perfidy of kings.

The second suggestion is that the scene represents the rescue of the king of the Garamantes by his dogs (Alexander 1987; cf. Kahn 1992, 71). The story is represented in art in the section on dogs in a number of Latin bestiaries written and illuminated in England in the later twelfth and thirteenth centuries, the earliest of c. 1170. It is ultimately derived from Classical sources via Isidore of Seville, but the earliest occurrence in an English manuscript quoted by Alexander is in BL MS Stowe 1067, of c. 1120, where the story is not illustrated. In Alexander's opinion the late twelfth- and early thirteenth-century illustrations suggest that the pictorial tradition, already then corrupted, preserves accurately the form and style of considerably earlier models, of Carolingian or even Late Antique date: 'it is possible and even likely that the story was known in visual form at least by the end of the eleventh century' (Alexander 1987, 5).

Alexander's thesis poses a number of difficulties, several of which he underlines. It does not account for the tether around the neck of the recumbent man, nor for the juxtaposition of the animal's open (but not biting) mouth with the man's open mouth, and does not explain why the animal's tongue is at least in contact with the man's open mouth and probably within it, an action made possible by the man's jaw being held down by the animal's paw. Alexander rightly suggests that the scene of the king of the Garamantes rescued by his dogs would be more suited to the decoration of a royal palace, quoting the pertinent parallel of the painting of what must be this subject ordered by Henry III in 1256 for the decoration of the 'wardrobe where the king washes his head' in the palace at Westminster (Alexander 1987, 6). He then goes on to argue that the Winchester stone may have been discarded from some earlier structure when William the Conqueror rebuilt the royal palace at Winchester in 1070, and may have lain around for some years before being incorporated in the destruction rubble of Old Minster. This, he admits, 'is pure speculation'. It must be added that it conflates sites and events in central Winchester and ignores the discovery of fragments of similar figures elsewhere in the rubble of Old Minster (as noted above; see also Biddle 1984, 133). Fig. 27 shows how the carved fragments from Old and New Minster form clear patterns of discard probably reflecting their original positioning (see Introduction, Chap. VIII) and provides additional confirmation that no. 88 belongs to Old Minster. There is also the chronological gap: no literary or pictorial evidence for the Garamantes story survives in England before the twelfth century. Ogilvy suggests that no Anglo-Latin authors used the Physiologus and that it is doubtful that the English had a full copy of the Bestiary (Ogilvy 1967, 100, 222); Ohlgren 1986 provides no evidence of the story in manuscript illumination.

This leads us directly to the question of date and style. The archaeological evidence provides a terminus ante quem of 1093–4. Since the eastern part of Old Minster, in the eastern extremity of which no. 88 was found, was not begun until after c. 980 and was dedicated in 993–4, the stone can be dated to the century between c. 980 and 1093–4. It cannot in any case be later. The condition of the stone, slightly weathered or worn and generally bruised, and the discovery elsewhere in the rubble of chips from similar figures, derived from knocking salient detail off other blocks to prepare them for reuse in building the Norman cathedral, show that no. 88 was not new when discarded and must be related to the decoration of the building then being demolished. If the suggestion that the stone formed part of a narrative frieze celebrating the shared origins of England and Denmark is accepted, a date in the reign of Cnut (1016–35) after his marriage to Emma in 1017 seems indicated. There is, however, no record of any structural work at Old Minster after 993–4 and the possibility cannot be excluded that the stone (whatever its subject) belongs to the great enlargement of the east end completed in 993–4 (cf. Jacobs 1977, 40, n.51).

Against this must be set the view that the piece is Romanesque in style: 'Is it from the reign of king Cnut or is it Romanesque? The archaeological evidence points to the former, but the style to the latter' (Zarnecki 1986b, 8). The discussion resolves into three themes: the style of the figure carving; the comparisons which have been drawn between no. 88 and the Bayeux Tapestry; and the likelihood that Cnut permitted the display of an incident from the pagan past in the cathedral church at Winchester, a city described in the Winchester Annals as regni soli solium (Luard 1865, s.a. 1017; cf. Biddle 1976, 289, notes 3–4).

Zarnecki has stated that 'the solid round forms with well-defined contours [of no. 88] are Romanesque rather than Anglo-Saxon' (Zarnecki 1984, 151). As Kauffmann has argued, Norman book illumination produced 'a harder, drier, more solid version of the Anglo-Saxon style and to this extent was more Romanesque in character' (Kauffmann 1975, 19), but this is in a field where there is ample material for comparison. In stone sculpture there has been less material by which to define the late Saxon figural style in Wessex, but it is doubtful whether no. 88 could be regarded as more solid or well-defined than, for example, the Bradford on Avon angels, the Romsey 1 crucifixion (Ills. 451–2), or the Bristol harrowing of Hell, and it shares with them the flattened rather than rounded surface which seems both characteristic of Wessex sculpture of the period and is probably a reflection of the technique used to work back the flat surfaces of the blocks as received from the quarry. There seems, moreover, little in the corpus of Anglo-Norman Romanesque sculpture to which no. 88 may be directly compared.

As Wilson points out, by comparison with the Bayeux Tapestry, no. 88 'being carved in stone, is of course more solid and rounded' (Wilson 1985, 208). Dodwell suggests 'on stylistic grounds' that its dating 'by archaeologists to the late tenth or early eleventh century is too early' (1982, 138), but gives no details, and Wilson concludes that Zarnecki's claim that no. 88 is Romanesque 'is a happy solution' (1984, 200).

Much of this debate has confused style with date, as Zarnecki's proposed dating '1016–35 or Romanesque' in the 1066 Exhibition Catalogue shows (Zarnecki 1984, 150). This compels him to the complex question: 'could it be that [no. 88] is a Romanesque copy of a subject which was depicted at Winchester on an embroidered hanging presented by Cnut . . .?' (ibid., 151). Wilson's argument for a post-Conquest date adopts Zarnecki's view that the style is Romanesque rather than Anglo-Saxon, and adds two points: the 'remarkable parallel' which no. 88 provides for the Bayeux Tapestry, and the suggestion that since the carving is unpainted, it 'was abandoned unfinished (perhaps due to damage) at some time not long before the building of the present cathedral [which began in 1079]' (Wilson 1985, 206, 208). Since, however, only one of the pieces of sculpture from Old and New Minsters in this Corpus bears any trace of paint (no. 43), although whitewash and plaster survive in the nooks and crannies of twenty-one of the carvings, and there is paint over whitewash on one of the mouldings (Winchester (New Minster) no. 4), this last suggestion has little force.

If there seems to be scant difference between 'the solid round forms with well-defined contours' of no. 88 and the style of the other figural carvings from Old Minster in this Corpus, albeit very fragmentary, one must conclude, accepting these features as Romanesque, that there was a significant Romanesque element in the Old Minster sculpture. It seems hardly likely that this sculpture can all belong to some unrecorded major building campaign between 1066 and the commencement of the new cathedral in 1079. As Biddle wrote in 1984, Zarnecki's view that the piece is 'Romanesque rather than Anglo-Saxon is 'a crucial comment, which, if the archaeological evidence is allowed its proper weight, places the advent of the Romanesque style in England before the Conquest' (Biddle 1984, 134). Romanesque elements are present by the 1050s in architecture (the beginnings of the Confessor's abbey church at Westminster) and in manuscript painting (e.g. BL MS Cotton Tiberius C. VI), and the Winchester carving may suggest, if it does show Romanesque traits, that these were present even earlier in sculpture.

The comparison of no. 88 to the Bayeux Tapestry was made as soon as it was found (Biddle 1966b, 329, 332). Three points are relevant: both are friezes; both show a method of scene division whereby one figure turns its back on another; and they share similarities in figure shape, hair style, dress, and accoutrements. The third point needs some expansion. The left-hand figure on no. 88 has the long, thin legs typical of the Tapestry. The hair of the recumbent figure is worn long and tucked behind the ears, like the hair of the English on the Tapestry (e.g. Harold: Wilson 1985, pl. 2, etc.). The strong jaw-line of the recumbent figure is characteristic of the figures on the Tapestry (e.g. Harold: ibid., pl. 14, etc.). The mail shirt, with its trouser-like thigh protection and edge binding (Brooks 1978, 94–6), short sleeves, and undershirt with sleeves rucked to the wrist, is seen repeatedly on the Tapestry. The long, straight sword worn in a scabbard hung from a belt on the left hip is a commonplace (e.g. Wilson 1985, pls. 7, 9 (unbuckled with belt hanging loose), 10–13, 15, 19, 56, 69), but here there are two discrepancies of uncertain significance. First, the angled strap from belt to scabbard shown on the carving does not appear on the Tapestry, although it would seem essential to be able to adjust the angle at which the sword was worn, at least when walking. Second, when the sword was worn over mail the Tapestry normally shows both Normans and English wearing it without a belt (ibid., pls. 22–3, 51, 55, 64, 70, 72), as if supported by a belt worn underneath and fastened to the scabbard through a slit or slits in the side of the mail. In the two cases where a sword belt is shown over the mail, the figures are both English foot soldiers (ibid., pls. 56, 69). In the scene where Duke William gives Harold arms, Harold's scabbard is shown inside his mail, the hilt of the sword appearing outside in the normal position, and the end of the scabbard poking out from beneath the skirt (ibid., pl. 24). This may be a simple error (ibid., 221–2), but the writer has worn his sword as orderly officer in exactly this way under a greatcoat, with the hilt exposed, and scabbard and belt concealed. It would indeed be awkward to wear a sword belt over either a greatcoat or a mail shirt, especially when riding. It looks as if the Tapestry may show two practices: the English, fighting on foot and wearing mailed trousers (Brooks 1978, 94–6; Brooks and Walker 1979, 19–20), when they carried a sword at all, wore a sword belt over mail, as on no. 88, whereas the Normans, mounted and wearing mailed skirts, perhaps strapped over the legs like a modern riding coat, sometimes wore the sword under the mail.

There seems no doubt that the relief 'provides the closest parallel for the Tapestry' (Zarnecki 1986a), and that it is 'a worthy prelude' to its narrative art (Cramp 1972, 148). The fact, as Wilson has noted, that the sculpture is the more finely detailed of the two (1985, 208) may account for minor variations like the presence on the stone of the angled sword strap. The similarities have even led Wilson to suggest that the Tapestry may have been made at Winchester (ibid., 212; but cf. his earlier comments, 1984, 200). The attribution to Canterbury rests on much firmer ground, however (Brooks and Walker 1979, 17–18).

There remains the question of whether Cnut would have permitted the display in the cathedral of an incident from the pagan past. Wilson rejects this on the grounds that Cnut was 'intent on becoming a Christian king and would probably not want to advertise his pagan ancestry' (1984, 200; 1985, 208). But it is the heroic and not the pagan which matters here (Davidson 1967, 127–8; Gatch 1971, 27–36; Hinton 1977, 95–6). The appearance of Sigurd and Weland on pre-Conquest Christian carvings from northern England has even been seen as a 'pointed juxtaposition of mixed iconography' (Lang 1976, 94), where the Christian connotations of these scenes may be an attempt to redeem pagan ancestors (Ploss 1966, 96ff.). This too is the period when all the surviving manuscripts containing heroic stories of the pagan past, such as Beowulf, or extensive references to that past, as in the poem Deor, were written (Ker 1957, nos. 101, 116, 216, 282; Gneuss 1981, nos. 257, 399, 816). In the case of Beowulf alone, this is sufficient to demonstrate a living acquaintance in the early eleventh century with the saga of Sigmund, episodes of which appear in lines 867–900.

If the Winchester stone was part of a narrative frieze recording the shared ancestry of the English and Danish royal houses in the way suggested, and in the detail indicated by the inclusion of so slight but dramatic an incident as Sigmund and the wolf, the total length must have been very great. On the supposition that it was set on the interior walls of the eastern arm of Old Minster, east of the lateral apses, a length of 80 feet (24.4 m) is possible. There is nothing to show that it was not even longer, for it could have run around an even greater length of the exterior of the building, or have been in more than one register, whether inside or out. The similarities of the Winchester frieze to the Bayeux Tapestry may therefore have included length, for it should be remembered that the Tapestry was over 230 feet (70.4 m) long. The other figural fragments from Old Minster in this Corpus are the merest wreck of what once existed, but they provide a context for no. 88 in Old Minster which was not available when the stone was first published. Another context is provided by the evidence for the sculptural decoration on a very large scale of the late tenth-century tower of New Minster, which Quirk argued thirty years ago consisted of a series of friezes representing the dedication of each of the six stories of the tower (Quirk 1961, 33–5).

The part this particular frieze played in the evolution of such a concept as the Bayeux Tapestry is problematical. Despite the general similarity and the particular parallels, it would perhaps be wiser to see both emerging from a wider and older tradition of narrative art displayed in a variety of media, painting both small and large, carving in wood, stone, and ivory, and embroidery. Völsunga saga itself contains two specific descriptions of narrative hangings: one embroidered by Brynhild which showed the deeds of Sigurd, and one woven by Guďrún which illustrated among other actions the fight of Sigar and Siggeir and the ship of Sigmund sailing along the land (Finch 1965, 42, 62). These are, of course, literary creations. Their appearance in Völsunga saga need only reflect the existence of such pieces in eleventh- or twelfth-century Norway or Iceland at the time Guďrúnarkvi ďa II, which is the source of the Guďrún material noted here, was compiled (Neckel 1927, Guďrúnarkvi ďa II, st. 16). The earliest English reference to an actual narrative hanging belongs to the late tenth century. After the death of ealdorman Byrhtnoth at the battle of Maldon in Essex in 991, his widow Aelfflaed gave to the monastery at Ely a number of lands, a golden collar, 'et cortinam gestis viri sui intextam atque depositam, depictam in memoriam probitatis eius' (Blake 1962, 136).

The Byrhtnoth cortina is alone sufficient to show that a tradition of secular narrative depiction existed in England at or before the proposed date of the Winchester sculpture. There is, in addition, independent Scandinavian evidence for the existence of some kind of narrative embroidery at an even earlier date, for the Oseberg ship-burial of c. 800 contained fragments of a long narrow hanging possibly showing scenes from a saga (Hougen 1940). Had Brynhild and Guðrún embroidered in reality the deeds of Sigurd and Sigmund, this is perhaps how their work would have looked.

To sum up, no. 88 has been tentatively identified as part of a very long frieze depicting the shared traditional history of England and Denmark, a history symbolically united in the marriage of Cnut to Aelfgifu-Emma, widow of Aethelred II, in 1017. The place of their marriage is unknown, but in time both were buried in Old Minster, Cnut in 1035 and Emma in 1052.

Date
Between c. 980/993 - 4 and 1093 - 4, probably 1017 - 35
References
Biddle 1966a, 325, pls. LIXb, LXII, LXV; Biddle 1966b; Biddle 1967c, 661, pl. 7; Davidson 1967, 127; Gatch 1971, 33 - 5, pl. 1; Biddle and Kjølbye-Biddle 1972, no. 16; Cramp 1972, 148; Biddle and Kjølbye-Biddle 1973, no. 18; Kjølbye-Biddle and Page 1975, 390; Lang 1976, 94; Hinton 1977, 95 - 6; Jacobs 1977, 40 and n. 81; Brooks 1978, 94 - 6, pl. 1; Brooks 1979, 19; Biddle 1981, 166, 168, cat. no. J1; Dodwell 1982, 137 - 8, pl. 31; Zarnecki 1984, 150 - 1, cat. no. 97; Wilson 1984, 198 - 200, ill. 258; Biddle 1984, 133 - 5, cat. no. 140; Wilson 1985, 206 - 8, fig. 5; Zarnecki 1986a; Zarnecki 1986b, 8 and n. 7; Alexander 1987; Kahn 1992, 71; Biddle and Kjølbye-Biddle forthcoming a, fig. 153, no. 90
M.B.; B.K.-B.

Forward button Back button
mouseover