This project is inspired by a second half 10th century grave find from the Shestovica cemetery in Chernigov, Ukraine (Figure 1). The find and several others in a new series of graves that were exposed during excavations of sites for new building developments are discussed by Kirill Mikhailov, who had permission from the excavators to study the grave objects. The extant piece that was my inspiration was a narrow band, one of several garment fragments from the Viking age Shestovica cemetery, excavated in 2005 under direction of V. P. Kovalenko. The band is c. 10 mm wide, tablet woven of silk and silver thread consisting of drawn silver wire. Its position in the grave relative to the male burial and the other textile fragments suggests that it formed the edging of a wool mantle. Kirill Mikhailov discusses its similarity with gold and silver brocaded bands found in a large number of burials from the Viking age trading center Birka in current day Sweden (discussed extensively by Agnes Geijer).
Figure 1: Tablet-woven band from the grave chamber, Shestovica cemetery, Chernigov, Ukraine (Photo K. Mikhailov).
Technique, material and design
The band that inspired this project is a tablet woven silk band with a silver brocaded pattern. Its materials and technique are very similar to those of many Birka bands. They consist of: (a) the basic ground weft that binds the warp threads together, hidden by them in the usual way, and (b) a second, brocading weft of double drawn silver wire, creating the actual pattern or design. The brocading weft consists of a pattern of floats and hidden sections by passing the weft underneath one or more warp threads of one or more cards. I chose to use gold thread instead of silver, which better suited the purpose of the project. The use of gold thread in brocaded bands, although much more rare than silver thread, is supported by contemporary Viking age grave finds.
The pattern was obtained from a blog post on "Aislings Welt". The author of the blog obtained the pattern by examining the photograph of the band and weaving several attempts until it visually matched the extant piece. The pattern was woven on 19 cards with four threads per card. The warp was store bought 60/2 silk, the ground weft was store bought silk sowing thread and the brocading weft was store bought gold thread (#7 metallic gold around a silk core). This resulted in a width of about 12 mm, close to that of the extant piece. The warp tension was approximately 7 g/thread, resulting in a pattern where the diagonal lines were at close to 45 degree angles with respect to the centerline of the band, consistent with the extant piece. The final woven length of band was c. 22 inch.
The extant piece clearly shows the use of a double brocading weft. Looking more closely at the bottom right edge in Figure 1, you can see that the brocading wefts are woven together as a pair and extend to the edge of the selvage. While this may be fine for trim on a mantle, I did not like the idea of the occasionally somewhat sharp kinks due to loops at the edges of the band for a headband that is worn either directly on the forehead or over a thin veil. Hence, I chose not to extend the brocade all the way to the edge of the band. I also chose to weave each of the brocading wefts form alternate sides, one going from left to right, the other from right to left in each pick. This gives more even selvages and greatly helps to keep both brocading wefts flat and parallel in each pick.
According to Nancy Spies the most common turn type of the brocading wefts between each pick is to dive to the back of the band between one or more of the outer cords and back to the front of the band on the return pick between the same cords (as shown in the left panel of Figure 2). It is common with this turn type to find that the metallic brocading weft, over time, cut the ground weft, which explains why many early bands are missing their edges. I also found that this turn type tends to pull the very thin ground weft through the band, leading to sloppy selvages. To solve this, I chose to turn the brocading weft around the second-last cord on each side as illustrated in Figure 3 for the left selvage. The treatment at the right selvage is simply the mirrored version of the left one. This turn type is a slight variation of the turn type shown in the right panel of Figure 2.
Figure 2: Two turn types of the brocading weft between successive picks as discussed by Nancy Spies.
Figure 3: Schematic of the turn type of the brocading weft between successive picks at the left selvage as used for the current project.
Weaving the band
The band was woven on a warp weighted tablet weaving loom that I designed and build for earlier projects. The loom is designed to use on the top of a table, with the warp ends weighted down and dropping freely over the fixed rod at the end of the loom. The other end of the warp is attached to a ratcheted rod, which stores the finished band. The initial warp length was about 36 inch, enough to create the required length of band plus room to compensate for the take-up during weaving and to experiment a bit or create a sample piece for later reference.
The warp was created one card at the time, threading back and forth twice per card. The cards were warped alternating S and Z (to eliminate fouling of the warp threads during turns). The brocading wefts passed under two of the four threads of the cord creating a smooth and uniform underside. There is no evidence in the extant piece that the twist direction was changed along the band (although the piece is too short to be conclusive on this issue). I choose therefore to rotate all cards collectively forward one quarter turn for each successive passage of the ground and brocading wefts. As discussed above, the two brocading wefts were passed from alternate sides as can be seen in Figures 4 and 5. The unfinished warp ends were periodically untwisted to eliminate the build-up twist.
Weaving the band.
Figure 4: The first inch.
Figure 5: About halfway finished. The finished band is rolled up on the nearest rod.
Figure 6: Close ups from the top (left) and the bottom (right) of the band.
Figure 6 shows that the loops of the brocading wefts are at the underside of the band and lie pretty flat. This is a direct result of the particular treatment of the brocade wefts between successive picks as show in Figure 4. However, as you can see, occasional small kinks in the metallic threads are hard to avoid, so to protect the veil and/or forehead from the edges of the metal brocade, a strip of folded linen was sewn onto the under side of the brocaded band. Furthermore, I made the strip a little longer than the finished brocaded band so that it also serves to tie the band around the head. Figure 7 shows the finished headband as worn by Her Majesty just under the crown.
Figure 7: The finished headband as worn by the recipient.
Find a downloadable version of this post at: https://www.academia.edu/37766326/Gold_brocaded_headband_inspired_by_a_find_from_the_Shestovica_cemetery
Aislings Welt (2015) Shestovica cemetery, Chernigov, Ukraine http://aislingde.blogspot.com/2015/12/shestovica-cemetery-chernigov-ukraine.html (Last accessed: November 4th, 2018)
Collingwood, Peter (1982) The Techniques of Tablet Weaving, Echo Points Books & Media (Vermont)
Geijer, Agnes (1938) Birka III: Die Textilfunde aus then Gräbern, Kungl. Vitterhets Historie och Antikvitets Akademien (Uppsala).
Mikhailov, Kirill (2008) New finds of Viking Age textiles in Ukraine and Russia, Archaeological Textiles Newsletter 47, pp. 17-20.
Spies Nancy (2000) Ecclesiastical Pomp & Aristocratic Circumstance: A Thousand Years of Brocaded Tabletwoven Bands, Arelate Studio (Maryland).