Not a cloak after all?

Conflicting interpretations of Bj 824 A

By Hilde Thunem (
(Last updated December 17 2022) (PDF)

Grave 824 A at Birka in Sweden is sometimes cited as evidence that cloaks or back-cloths could be pinned directly by the oval brooches (Geijer 1938 p. 139-140, Ewing 2006 p. 35). However, the evidence is far from conclusive, and there are other interpretations (Hägg 1974 p. 77-78). In this article, I take a closer look at the textiles in the grave and the reasoning behind the different interpretations of the garments worn by the deceased.


The archaeological remains

The main excavation of Birka was conducted in the 1870s by Hjalmar Stolpe, yielding a large amount of textile material from the 9th and 10th century. The textile fragments were stored, and later analysed by Agnes Geijer (1938) and Inga Hägg (1974, 1986).

Grave 824 A is dated to the 10th century (Geijer 1938 p. 167). The grave contained remains of a skeleton and several different artifacts; two oval brooches, made of bronze and with traces of gilding, an iron knife in a leather sheath with bronze fittings, remains of a pair of iron scissors, and a comb fragment (Arbman 1943 p. 297).

Some fragments of textiles also survived. Fragments of two types of tablet-woven bands had been preserved due to their metal weft, and there were small fragments of silk stitched to them (Geijer 1938 p. 85, 167). In addition, there were some fragments of woollen diamond twill in the grave, and a few linen loops preserved inside the oval brooches (Hägg 1974 p. 45, 65).

Grave plan drawn by Stolpe
Illustration: Stolpe 1872-1879 (section)

Tablet-woven band(s) with silver weft (B 2 a and b)

The grave contained several fragments of tablet-woven bands with silk thread in the warp and a silvered thread as weft.

The fragments are 12-13 mm wide, with 17-19 silk threads in total in the warp, whereof 9 threads are in the pattern section, and 4-5 threads at each edge. Several of the silk threads along the edges have deteriorated, leaving the silver weft sticking out as a row of tightly packed eyelets along the edges of some of the fragments. The weft density is 14-15 threads per cm (Geijer 1938, p. 85).

Geijer (1938, p. 85) notes that although some of the larger fragments look different due to deterioration, they have similar patterns, and may originally have come from a single tablet-woven band.

Which fragments belong together?

The grave plan indicate that the band fragments were found as two larger pieces of band (B 2 a and B 2 b), one at each oval brooch. Unfortunately, while the grave plan has a schematic drawing indicating the placement of the two band pieces, it cannot be used to identify which of the different band fragments belong to each of the pieces. To complicate matters further, Geijer (1938) and Hägg (1974) arrives at slightly different conclusions regarding which specific fragments belong together.

According to Geijer (1938 p. 85) band B 2 a consists of two separate fragments, respectively 19 and 12 cm long, and B 2 b consists of one 9 cm long fragment, and several smaller fragments.

Hägg (1974, p 77-78) describes B 2 a as a single fragment, 18 cm long, and places the 12 cm long fragment (that she describes as two separate fragments) together with the 9 cm long fragment as part of band B 2 b. The Swedish history museum have placed the bands together in a manner consistent with Hägg’s grouping.

Tablet woven bands B 2 a and b
Photo of fragment groups B2a and B2b, by employee at SHM Creative Commons Erkännande-Ickekommersiell-Dela Lika 2.5

In the following, I will combine information from both analyses, but will group the fragments according to Hägg’s description, as she takes into account details like seams and wear and tear on the bands.

Tablet-woven band B 2 a

The fragment is 18-19 cm long. It consists of two pieces of tablet-woven band that have been joined together by a seam, where a couple of cm of the end of each piece have been folded towards the back, and then stitched together (Hägg 1974 p. 77).

There is a fragment of silk twill (S 4) sewn along one side of the band. A silk band runs across the end of the tablet-woven band, enclosing both it and the silk twill fragment (Geijer 1938, p. 85). There are small traces of silk twill along the other edge of the tablet-woven band as well, indicating that it originally were framed by silk strips on both sides (Hägg 1974 p. 77).

The other end of the tablet-woven band ends in a torn edge (Hägg 1974 p. 77).

Start of tablet woven bands B 2 a
Photo: Geijer 1938 Taf 19 (section) - fragment B 2 a

The silk twill (S 4) is of a type common in the Mediterranean region at the time. It has two warp systems: one warp that binds the weft, and one pattern-forming filling warp that serves as support, and when changing the weft threads. The outside of the fabric resembles ordinary twill (three-weave weft twill); while the inside is different. The same type of fabric is found in other graves at Birka (Geijer 1938 p. 58).

This twill type may be plain or have a pattern (created by using different colors in the weft), depending on the choice of the weaver. However, neither Geijer (1938) nor Hägg (1974) mentions whether the silk twill from Bj 824 A had traces of a pattern.

B 2 a. GRAB 824. (...). Brettchenband mit Silberschuss. In der Musterpartie 9 ganzseidene Br. Unterseite dicht und unregelmässig. Am Rande waren 2 ganzseidene Schnüre und dazwischen 2 oder 3 jetzt vermoderte. Im Ganzen 17-19 Br. Schussdichte 15 per cm. Breite 12 mm. 2 Stücke von 12 bzw. 19 cm. Am Rand ist ein Seidenfragment (S 4) angenäht, aus dem auch ein querverlaufender Streifen hergestellt ist, der das eine Ende der beiden Borten mit daran sitzender Seide umschliesst.
Geijer 1938 p. 85

B 2 a (...) har ena änden infällad under ett kantband av siden, som ligger i rät vinkel mot brickbandet, 824 A:1. Andra änden saknar ursprunglig avslutning. Bandet är ca 18 cm långt och hopfogat av två delar. Vid fogen är bandändarna nedvikta mot avigan med någon cm, vikkanterna lagda intill varandra och sammansydda. Ett större stycke av bandets sideninfattning finns bevarad. Den består av två, aviga mot aviga lagda remsor med invikta kanter. Mellan långsidornas vikkanter har brickbandet infogats och fasts. Små rester av sidenkypert längs bandets båda sidor visar, att det varit infattat från båda håll. Det ovan nämnda kantbandet av siden i rät vinkel mot brickbandet har alltså fållat in inte bara detta och den bäst bevarade infattningsremsan, utan även en likadan remsa längs brickbandets andra sida.
Hägg 1974 p. 77

Tablet-woven band B 2 b

B 2 b consists of several fragments. The first fragment (9 cm long) has one end covered by a silk band, in the same manner as B 2 a, while the other end is torn. The next fragment is torn at one end, while the other end has been folded towards the back (probably as a part of joining it to another band piece). The last fragment may have been folded at one end, while the other end is torn, with trailing threads. All fragments have small remains of silk strips along their edges, indicating that they were framed in silk (Hägg 1974 p. 78).

B 2 b. GRAB 824. Taf. 19:1. Brettchenband mit Silberschuss. In der Musterpartie 9 ganzseidene Br., ziemlich grob. Am Rande waren 4 bis 5 Br., von denen nur die innerste aus Seide übrig ist; infolgedessen bildet jetzt der Silberschuss am Rande eine Reihe dichtgedrängter Ösen. Insgesamt 17 oder 19 Br. Unterseite dicht, aber unregelmässig. Schussdichte 14 bis 15 per cm. Breite 13 mm. Länge 9 cm sowie kleine Fragmente. Im Muster usw. sehr ähnlich dem vorhergehenden. Vielleicht ist es dasselbe Band, obwohl es so verschieden wirkt — teils wegen der starken Zersetzung im ersteren Band, teils deswegen, weil die Randschnur in dem letzteren verschwunden ist.
Geijer 1938 p. 85

B 2 b. Parstycket till detta [B 2 a], B 2 b, föreligger i tre separata fragment, 824 A:2. Det första är i ena änden infällat med ett kantband på samma sätt som B 2 a. Andra änden är söndertrasad. Nästa del saknar ursprunglig avslutning i ena änden men har en fogvikning i den andra. Det tredje och sista fragmentet är söndertrasat i ena änden och har vad som möjligen kan vara en fogvikning i den andra. Alla tre har rester av sideninfattning längs kanterna.
Hägg 1974 p. 78

Tablet-woven bands with silvered and gilded weft (B 22)

The slim tablet-woven bands with silvered weft were not the only bands in the grave. There were also fragments of bands of a different type, where the weft alternated between silvered and gilded thread in sections along the band.

One band is preserved almost in its entirety. It is 9.5 cm long (only 8 cm had a metal weft) and 12 mm wide. There were at least 29 threads in the warp whereof 17 formed the pattern section, and 6 threads (only four currently survive) were at each side (Geijer 1938 p. 88).

The band has a narrow strip of silk taffeta (S 1) sewn along all sides, with the possible exception of one of the long sides. The underside is dense, and somewhat irregular, and Geijer (1938 p. 88) notes that its appearance indicate that the whole piece may have been sewn to a fabric that have since disappeared.

In addition to the fairly well-preserved band, there were several fragments that may have formed a similar counterpart to the band (Geijer 1938 p. 88).

Tablet woven bands B 22
Tablet woven bands B 22
Photo: Geijer 1938 Taf 23 (section) - fragment B 22

None of the fragments have been drawn on the grave plan, and their position in the grave is thus unknown.

The silk taffeta (S 1) is a tight and even plain weave, with 28 threads per cm in both warp and weft. Each silk thread in the weave consists of almost 150 cocoon threads that have been slightly twisted together. The regularity of the thread, even though the cocoon threads have not been spun together, indicates a remarkable skill in the handling and sorting of the silk (Geijer 1938 p. 60-61).

B22 GRAB 824. Taf. 23:2, Abb. 20 b. Brettchenband abwechselnd mit Gold- und Silberschuss in Partien zu 6-15 cm. Die Musterpartie hat 17 vollständige Br., die verschiedene Flechtenmuster bilden. Am Rande ursprünglich 6(?) Br., davon 4. erhalten. Zusammen mindestens 29 Br. Die Schüsse 20-25 per cm, das Gold am dichtesten. Die Unterseite dicht, etwas unregelmässig. Die Schnüre haben dieselbe Drehungsrichtung. Breite 12 mm, Länge 9.5 cm. Der Metallschuss deckt nur 8 cm der Länge des Bandes. An beiden Enden ist ein Stück des Brettchenbandes zu sehen. Das Ganze, möglicherweise jedoch nicht die eine Längsseite, ist mit einem schmalen Taftstreifen eingefasst (S 1). Das Aussehen der Unterseite macht es wahrscheinlich, dass das ganze Stück ursprünglich an irgendeinem jetzt verschwundenen Stoff angenäht war. Fragmente gleichen Aussehens haben wahrscheinlich hiezu ein Gegenstück gebildet.
Geijer 1938 p. 88

S 1 a. GRAB 824. Taf. 13: 3. Taftseide. Sehr ebenmässige und enge Leinwandbindung. Dieselbe Zahl - 28 St. per cm - und dieselbe Sorte Fäden in beiden Richtungen. Laut Materialuntersuchung (siehe Beil. I) ist der Seidenfaden sehr ebenmässig und besteht aus nahezu 150 Kokonfäden, die nur leicht zusammengedreht sind. Die Seide ist nicht entbastet - d. h. die Kokonfäden besitzen noch den schützenden Seidenleim (Serizin). Die Tatsache, dass die Kokonfäden nicht zusammengesponnen zu werden brauchten, deutet im Verein mit der Ebenmässigkeit des Materials auf eine bemerkenswerte Geschicklichkeit in der Behandlung und der Sortierung der Seide hin. In der Analyse wird die Annahme ausgesprochen, dass die ursprüngliche Farbe gelb gewesen sei - eine Ansicht, der die Verfasserin doch kaum beistimmen kann. Der eben beschriebene Seidentaft findet sich in Form einer 2-3 mm breiten Kante rings um ein ungefähr 1 dm langes Stück Band mit Goldschuss (B 22).
Geijer 1938 p. 60-61

Other textiles - linen loops and broken diamond twill

Both oval brooches had preserved small fragments of linen loops (Hägg 1974, p. 45). These loops would originally have held up a strap dress, a garment that may have been called a smokkr by the Norse (Ewing 2006 p. 37). In addition, there is remains of silk that might be a tool-band for the knife or scissors (Hägg 1974 p. 78).

Textiles preserved in the oval brooches
Illustration: Hägg 1974 p. 130

In addition to the loops, there were remains of a broken diamond twill (W 10) preserved underneath one of the oval brooches. The same fabric was found on the pair of scissors (Hägg 1974, p. 65). Hägg believes that it may have belonged to a garment worn on top of the serk, but underneath the smokkr.

Finally, on top of the oval brooches were a rough wool cloth (W2) that may have covered the deceased in the grave (Geijer 1938 p. 167, Hägg 1974 p. 65)

<...> skall här bara nämnas, att båda spännbucklorna har öglor av linne kring nålhållare och nålfäste, vilket visar att kjolen även i denna grav bör ha varit av den vanliga hängseltypen, 824:3 och 4.
Hägg 1974 p. 45

824. Mot undersidan av brättet på ena spännbucklan ligger rester av diamantkypert, W 10, fig. 43 och 824:4. Saxens skänklar har fastrostade fragment av samma W 10. Tredje spänne saknas. Diamantkyperten kommer från framsidan av ett plagg, som burits över särken men under kjolen. Över spännbucklorna fanns rester av en grov ylleväv (W 2), som kan härröra från ett täcke.
Hägg 1974 p. 65

Interpreting the facts

Given the fragmentary evidence and schematic grave plan, it is perhaps not surprising that there are different interpretations of what kind of garment(s) that were worn by the deceased woman in grave 824 A.

Interpretation by Geijer: backcloth or cloak

Based on the silk twill (S 4) that has been stitched to one side of B 2 a, Geijer proposes that both tablet-woven bands with silvered weft may have adorned the edge of a rectangular piece of silk twill. The resulting garment may have been fully or partially lined with another fabric that have decayed in the grave (Geijer 1938 p. 167). Whether the entire garment was made of silk or just trimmed with it cannot be determined from the few fragments that have been preserved (Geijer 1938 p. 139-140).

The grave plan shows the tablet-woven band fragments lying across both brooches. Geijer suggests that the bands were attached to a garment that hung over the shoulders and the back of the woman and was pulled down by the two brooches. She further suggests that the silk band running across the end of B 2 a, enclosing the tablet-woven band and the silk twill fragment, was part of a strap that was tied around the neck (Geijer 1938 p. 167).

Cloak held up by the oval brooches
Illustration: Geijer 1938 p. 149

She notes that the tablet-woven band fragments with alternating silvered and gilded weft (B 22) may have formed a pair of bands, whereof one is almost fully preserved. The bands were framed with silk taffeta and probably sewn onto a fabric that have later deteriorated in the grave (Geijer 1938 p. 88). However, as the position of the bands are not drawn on the grave plan, she makes no attempt to interpret what type of garment they may have belonged to.

GRAB 824. W. 10 Jhdt. 2 ungef. 20 cm lange Silberbänder, B 2 a, die den Rand eines vermutlich rechteckigen Seidenstückes (S 4) geschmückt haben, das ganz oder teilweise mit einem anderen, jetzt vermoderten Stoff gefüttert war. An dem einen Ende war das Band sowohl wie der Stoff mit einem Seidenstreifen eingefasst, der in einen Riemen fortgelaufen zu sein scheint (den man z. B. um den Hals knüpfen konnte, Abb. 47). In der anderen Richtung sind die Bänder nicht ursprünglich abgeschlossen. Sowohl die jetzige Form als auch der Plan geben an, dass das Kleidungsstück, auf dem die Bänder angebracht waren, über den Schultern oder dem Rücken lag und mittels der beiden Spangen nach unten gezogen wurde.

Ausser diesen auf dem Plan erkennbaren Stücken: ein 9 cm langes, ähnliches Stück, jedoch besser erhalten (B 2 b), und die Fragmente B 22 mit abwechselndem Gold- und Silberschuss, die mit Seide eingefasste Paarstücke ausgemacht haben. Auf der Oberseite der Spangen finden sich unbedeutende Reste groben Wollstoffes, gleich W 2.
Geijer 1938 p. 167

In einigen Fällen, u. a. im Grab 838, war die Spange zuoberst am Stoff selbst befestigt, dessen üppige Falten zeigen, dass er von der Schulter heruntergezogen worden war. Auf ähnliche Weise war vermutlich das silberbandverzierte Kleidungsstück befestigt, das von der Frau im Grab 824 getragen wurde; ob das ganze Kleidungs-stück aus Seide oder nur damit besetzt war, lässt sich jetzt nicht entscheiden. Wahrscheinlich sah Form und Befestigung ungefähr wie die schematische Zeichnung Abb. 47 aus.
Geijer 1938 p. 139-140

Interpretation by Hägg: a decorated tunic or jacket

Hägg (1974 p. 78) notes that the grave plan is of limited use when interpreting the textiles in the grave. It appears detailed, but a closer examination shows that there is discrepancies between what is drawn and what has been found.

Closeup of brooches in the grave plan Illustration: Section of grave plan

Firstly, the large silk twill fragment stitched to B 2 a has not been drawn on the plan at all.

Furthermore, the band lying either on top of or under the oval brooch to the left has been drawn with both ends torn. This cannot be band B 2 a, as one of its ends are covered by a silk band.

The other band, B 2 b, has some segments that are torn at both sides. However, one of the B 2 b segments is covered by a silk band at the end, and it is likely that this fragment would be found at the end of band, similar to B 2 a. Thus, neither B 2 a nor B 2 b are torn at both sides (Hägg 1974 p. 78).

Finally, the grave plan is not clear on whether the bands lie on top of the oval brooches, or underneath them. The oval brooch on the right side has been turned around in the grave, but the leftmost brooch appears to have been found in its original position. However, the band on the left brooch could be interpreted as being drawn on top of the brooch – while at the same time, the pattern of the brooch is visible through the band (Hägg 1974 p. 78).

In conclusion, Hägg decides to treat the grave plan as a schematic overview of the rough position of the finds, but uses other methods to decide on the exact position of the textiles. She proposes that the torn edges on segments of B 2 a and B 2 b are likely to be caused by wear from the edge of one of the oval brooches and positions the band fragments accordingly in relation to the brooches (Hägg 1974 p. 78).

Suggested position of oval brooches and bands
Illustration: Hägg 1974 p. 130 (sections). B 2 a (left) and B 2 b (right)

When examining the oval brooches, Hägg found no evidence suggesting that they had pierced a garment. Instead there were linen loops that indicate that the deceased wore a smokkr, fastened the same way as in other finds from Birka, and a possible tool-band of silk that may have fastened the knife or scissor (Hägg 1974 p. 45, 78).

Detail - start of tablet woven bands B 2 a In addition, both tablet-woven bands have small traces of silk fabric along both sides, indicating that they were framed with silk strips, and thus may have been decorating the middle of a garment instead of the edge, as Geijer supposed. The part of the silk band that Geijer believed may have run on as a strap that were tied around the neck, is instead enclosing the small silk twill fragment on the other side of B 2 a (Hägg 1974 p. 77-78).

Photo: Geijer 1938 Taf 19 (section) - fragment B 2 a

It is uncertain whether the silk-bands that enclose the ends of the bands and silk strips were turned towards the middle of the body (for example running along the edge of a slit or opening at the front of the garment), or whether they were turned towards the sides (Hägg 1974 p. 78).

Likewise, it is unknown where the two other tablet-woven bands (B 22) were in the grave. However, the best preserved of these bands has a shape and size that would fit with the bands being a pair of cuffs (Hägg 1974 p. 78).

Hägg tentatively concludes that the garment was a tunic or jacket of some kind, possibly made of woollen broken diamond twill and decorated with silk and tablet-woven bands (Hägg 1974 p. 87).

Små rester av sidenkypert längs bandets båda sidor visar, att det varit infattat från båda håll. Det ovan nämnda kantbandet av siden i rät vinkel mot brickbandet har alltså fållat in inte bara detta och den bäst bevarade infattningsremsan, utan även en likadan remsa längs brickbandets andra sida. Geijer, som bortsett från spåren av en tvåsidig infattning, uppfattade det hela så, att kantbandet skulle ha övergått i ett självständigt band (som t.ex. kunde knytas kring halsen, jfr rekonstruktionsteckningen Abb. 47). Den del av sidenbandet, som ser ut att fortsätta i ett band, är i själva verket resterna av kantbandet för den andra infattningsremsan.
Hägg 1974, p. 77

När det gäller att finna ut de exakta fyndlägena för parstyckena B 2 a och b. är man hänvisad till gravplanen (fig. 48). Tyvärr är planen över denna mycket viktiga grav ett typexempel på vad som inledningsvis sagts om det textila materialets styvmoderliga behandling i arkeologiska sammanhang. Den mycket eleganta planen visar gravinnehållet i detalj (t.o.m. ornamentiken på ena spännbucklan är antydd) med undantag för alla rent textila delar! De silvervävda brickbanden är vid ritandet sedda utan infattning och kantning av sidenkypert. Möjligen borde detta ses som en varning för att fatta planens upplysningar alltför bokstavligt också i andra avseenden. På planen är det samman hängande bandet över kvinnans vänstra axel söndertrasat i båda ändarna. Det bör då inte vara identiskt med B 2 a, eftersom detta fortfarande bildar ett sammanhängande band med ena änden bandkantad och inte upptrasad. Är det istället 2 b, måste den bandkantade änden, för att det hela skall stämma med planen, ha varit infogad någonstans mitt på brickbandet. Det verkar föga sannolikt och det korresponderar dåligt med konstruktionen av föregående band. Planen stämmer uppenbarligen inte på denna punkt, som jämförelsen med de existerande fragmenten visar. Istället bör brickbandens söndertrasade ändar ha tillkommit genom nötning mot någon av spännbucklorna. De söndertrasade bandändarna skulle enligt det förslaget egentligen ha legat som teckningen, 824 A:6, visar.

Geijer menade (jfr citat ovan samt hennes Abb. 47), att brickbanden B 2 a och b kantat ett plagg, som legat över axlar eller rygg och som hållits på plats av spännbucklorna, vars nålar på ålderdomligt peplosvis skulle ha stuckit direkt igenom tyget. I spännbucklorna, som är de enda spännen kvinnan hade (!), finns textillämningar av flera slag: linne i öglor från en hängselkjol av vanligt slag, diamantkypert, ett sidenband rakt ned från nålfästet i ena spännbucklan, antagligen ett upphängningsband för sax eller kniv, fig. 43 och 824 A:3-4. Det finns alltså inga spår av ett sådant arrangemang, som Geijer föreslår, inga spår av att nålarna - mot alla regler - stuckits igenom siden-brickbandsstycket för att på så sätt fästa det över axlarna. Höger spännbuckla med det ena brickbandet mot brättets undersida hade vält runt i graven och ligger inte längre i ursprungligt läge, jfr planen. Det gör däremot den vänstra. På denna sida har brickbandet ritats liggande ovanpå(?) spännbucklan (spännets ornamentik syns dock genom bandet).

Brickbanden B 2 a-b har varit infogade mellan sidenkypertstycken, som täckt delar av bröstet både ovanför och nedanför själva banden. Det är däremot ovisst, om de sidenbandskantade ändarna varit vända in mot kroppens mittlinje som kanter till ett sprund eller annan öppning i plagget, eller om de varit vända utåt sidorna. Likaså är det obekant, var de andra parstyckena, Geijers B 22, legat i graven. Dessa två, runt om avslutade parstycken, har emellertid form och storlek som skulle passa utmärkt för ett par manschetter, vilket det också snarast torde röra sig om i detta fall.
Hägg 1974, p. 78

Some thoughts on the different interpretations

Geijer (1938) is the first to attempt an interpretation of the garment(s) in grave 824 A. However, her focus is on analysing the quality and construction of the fabrics themselves (in this case the silk fabrics) and the patterns of the tablet-woven bands. It is therefore not surprising that Hägg, in her later analysis of the fragments, discovers new details that shed light on the construction of the garments in the grave.

Watercolour-drawing of interpretations

Illustration, by Hilde Thunem (2022). Possible garments, based on Hägg's and Geijer's interpretations. Silk twill (blue), tablet-woven bands B 2 a-b (grey) and silk band (red). In addition to the silk-decorated garment, there is a smokkr (brown) of unknown material. The colours in the illustration are merely for illustrative purposes, as no dye analysis has been performed on the textile fragments.

Did the oval brooches pierce the garment?

Oval brooches were not normally used to pierce the smokkr or other garments, as shown by the numerous finds of fabric loops connecting the brooches to the garments they held up. In the few cases where the brooches are piercing one or more garments, it is interpreted as part of the funerary arrangements, and not as a reflection of how the garments were worn in everyday life.

For example, there are two graves at Haithabu where the oval brooches were found piercing the underdress, probably to keep the brooches stable when the dead women were carried to their graves, fully dressed and on display (Hägg 1991, p. 278). Grave ACQ at Køstrup may have had a similar arrangement (Rasmussen and Lønborg 1993, 175-176). In grave B 6228 at Veka the overgarment fragments have stains indicating that the oval brooches were placed on top of the cloak, with the brooch needles piercing the cloak and fastening the loops of the smokkr underneath. It is interpreted as a display of the jewellery of the deceased in relation to the funeral (Lukešová 2011 p. 157-159).

The oval brooches in grave Bj 842 A contains the fabric loops that would have held up the smokkr, and possibly one or more tool bands for the scissors or knife. However, there is no evidence inside the brooches that indicates that they were piercing directly through whatever garment that bore the decoration of silk twill and silvered tablet-woven bands. While lack of evidence does not disprove Geijer’s interpretation, neither does it support it.

What are the similarities with other existing garments?

Geijer was not aware that the long tablet-woven bands were framed with silk strips on both sides. Arguably, one of the pieces of silk twill may have been narrower than the other, so the new evidence discovered by Hägg does not exclude the possibility that the twill and bands adorned the edge of a rectangular cloak.

However, the neck-band that Geijer introduces in order to explain the silk bands running across the end of the tablet-woven bands, requires that the rectangular cloak has cut some kind of opening for the head, with an edge that runs perpendicular to the edge of the cloak. This is a construction method that differs significantly from the straight edges found in early Iron Age cloaks. It is used in the cloak worn by the man found in Bocksten bog (dated 1300-1360 AD), but not, as far as I know in any of the preserved semi-circular cloaks of European monarchs in the period 650-1060 AD (Kania 2010). Again, it is not impossible to interpret the evidence the way Geijer does, but it lacks parallels in the existing Viking Age find material.

735:6 If we instead assume that the tablet-woven bands were framed with larger pieces of silk twill on both sides, it would probably not be placed along an edge. A similar technique, where tablet-woven bands have been stitched to silk twill that has been sewn onto a garment in wool, is known from other graves at Birka. One example is Bj 735, where the bands and silk fragment clearly form the front of a slightly tailored garment (possibly a tunic).

Illustration: Hägg 1974 p. 128. Interpretation of decoration in Bj 735.

Hägg's interpretation of Bj 824 A explains the silk band at the end of the tablet-woven bands either as decoration of a frontal slit in the tunic or jacket, or outlining the sides of the garment, similar to the vertical bands at Bj 735. Finally, although the evidence is uncertain, the size and shape of the silver and gold bands (B22) fits well with being decorative cuffs. If that is the case, they provide an extra indication that there was a tunic or jacket present in the grave.

In summary, the evidence is not conclusive. Either of the interpretations are possible, or at least not directly contradicted by the evidence. However, I find Hägg's interpretation of the garment as a tunic or jacket better supported, not only by the evidence in Bj 824 A, but also by the existence of similar types of finds from Birka.

Either way, given the uncertainties about the stratigraphy and position of finds in the grave, I would be very reluctant to use grave Bj 824 A as the sole basis of an interpretation of an overgarment. Doubly so, if one wants to introduce a new type of garment or a new way of wearing it.


Illustration sources