Golden textiles from Gokstad

Contents Archaeological Textiles Review Editorial ATR is published by the Friends of ATN, and hosted by the Centre for Textile Research in Copenhagen. Articles Editors: Karina Grömer Mary Harlow Jane Malcolm-Davies Ulla Mannering Kayleigh Saunderson Elsa Yvanez Scientific committee: Eva Andersson Strand, Denmark John Peter Wild, UK Lise Bender Jørgensen, Norway Elisabeth Wincott Heckett, Ireland Johanna Banck-Burgess, Germany Tereza Štolcová, Slovakia Heidi Sherman, US Claudia Merthen, Germany Christina Margariti, Greece Textile fibres from the Caleta Vitor Archaeological Complex, northern Chile Tracy Martens 4 Problems with Greek clothing terminology Quentin Richard 10 Nubian textile features: wool fragments from Hisn al-Bab and a tunic from Fag el-Gamus, Egypt Anne Kwaspen 20 Raincoats or riches? Contextualising vararfeldir through multi-perspective experiments Julia Hopkin 31 Golden textiles from Gokstad Marianne Vedeler Cover: Charlotte Rimstad Reconstructed tablet-woven band from Hallstatt in Austria (Image: Silvia Ungerechts) Viking Age textiles and tapestries: drawings by Miranda Bødtker Lise Bender Jørgensen, Dagfinn Moe and Hana Lukesova The textile bog find from Vong in Denmark Ulla Mannering, Charlotte Rimstad and Irene Skals Print: Grafisk, University of Copenhagen To purchase a printed copy of the latest Archaeological Textiles Review, please visit: https://www.webshophum-en.ku.dk/shop/ archaeological-textiles-664s1.html Information about institutional subscriptions is also available here. ISSN 2245-7135 47 58 75 Reconstructions revived: a handweaver’s personal perspective Anna Nørgård 90 Margrethe Hald: the quest for the tubular loom Ulrikka Mokdad and Morten Grymer-Hansen 101 Layout: Karina Grömer and Kayleigh Saunderson, Natural History Museum Vienna Visit www.atnfriends.com to learn more about the organisation. 2 Projects EgYarn: Unravelling the thread: textile production in New Kingdom Egypt Chiara Spinazzi-Lucchesi 113 Knowledge sharing: a newly found 2,700-year-old tablet-woven band from Hallstatt, Austria Karina Grömer, Silvia Ungerechts and Hans Reschreiter 115 Archaeological Textiles Review No. 63 Projects Unravelling Nubian funerary practices: textiles and body wrappings in ancient Sudan Elsa Yvanez Metallic idiophones 800 BCE to 800 CE in Central Europe: their function and acoustic influence in daily life Karina Grömer, Kayleigh Saunderson and Beate Maria Pomberger 126 129 RECONTEXT: Reconstructing the history of Egyptian textiles from the first millennium CE at the National Museum of Denmark Maria Mossakowska-Gaubert 134 Unwrapping the Galloway Hoard Susanna Harris and Martin Goldberg 136 Fashioning the Viking Age: status after the first three years Ulla Mannering 139 All the goods of the earth: making and marketing in the pre-Mongol marketplace Corinne Mühlemann 145 Conferences THEFBO: the significance of archaeological textiles Mila Andonova and Karina Grömer 147 Textiles and Seals: carving out a new research field and weaving data together Marie Louise Nosch 149 Funerary Textiles in Situ: an interdisciplinary workshop Magdalena M Wozniak 153 The Association of Dress Historians’ Annual New Research Conference Lena Larsson Loven 156 Old Textiles – More Possibilities: the Centre for Textile Research‘s 15th anniversary conference Elsa Yvanez 157 The textile dimension: textile finds in archaeology Petra Linscheid and Sabine Schrenk 160 North European Symposium for Archaeological Textiles Erika Ruhl and Sanna Lipkin 162 Resources: Recent publications and News Archaeological Textiles Review No. 63 164 1 Editorial Dear Reader, Another year in the shadow of the pandemic has passed. 2021 has been a challenging year for many. We had hoped that the pandemic would now be under control but we will probably have to adjust ourselves to an academic world in which we share more data online and meet up in person on fewer occasions. Therefore, it is with great satisfaction that the editorial team can now place the current issue directly online on the Archaeological Textiles Newsletter homepage. It is still possible to order a printed copy from the webshop at the University of Copenhagen in Denmark (www. webshophum-en.ku.dk/shop/archaeological-textiles664s1.html) but authors will no longer receive a printed copy. Important changes have occurred in the editorial group in 2021. Eva Andersson Strand who has been part of the ATR editorial team since 2008, has resigned from this task and we thank her for her long and devoted participation. She will still keep her place on the board of the Friends of ATN/ATR society. Three new editors have joined the team: Mary Harlow from the UK, Elsa Yvanez from Denmark and Kayleigh Sanderson from Austria. Together we have developed an efficient editorial process which we hope authors and readers appreciate. At the same time, we are constantly working on optimising workflow procedures, guidelines for authors (now downloadable from the homepage) and contact with the authors. It is very important that authors follow these guidelines before submitting articles and project reports. In future, we will decline non-compliant articles owing to the heavy workload on the editorial group. Please also write in formal academic English, using the correct scientific terms. If English is not your first language, please consult a skilled English speaker before submitting. The editorial team will make a final language edit but we do not have resources to make extensive corrections or to rewrite articles or reports. Throughout the past year, ATR has received a constant flow of contributions, and we are open to receiving more articles and project reports, especially from authors who are not familiar with or used to publishing in the journal. We therefore encourage authors to share their knowledge through the Archaeological Textiles Review. We are an editorial board which loves textiles, preferably those found in archaeological contexts, but also encourages themes of more general value to textile experts. We now accept articles up to 10,000 words, (including abstract, references, captions, acknowledgements) and four to ten images with captions, and project reports up to 3,000 words with three to five images with captions. Articles are double-blind peer reviewed and project reports are reviewed by the editorial group. Please keep updated at our homepage https://www.atnfriends.com, where we also post news. This year’s issue contains a range of articles and project reports from different periods, with the Scandinavian Viking Age particularly well represented in both articles and project reports. There are rather fewer conference reviews than usual due to the pandemic Fig. 1: The ATR editorial team (clockwise from top left): Kayleigh Saunderson, Elsa Yvanez, Jane Malcolm-Davies, Mary Harlow, Karina Grömer and Ulla Mannering 2 Archaeological Textiles Review No. 63 Editorial but we welcome more reports about online events (including ones held in 2021). Altogether there are nine articles, and eight project descriptions. Enjoy reading this issue and please spread the word about it. ATR is now a true open-source journal, free to download, and this is only possible with the work and help of many volunteers and enthusiastic hands and minds. Remember – the deadline for every issue is 1 May each year (with conference reports welcome until 1 October or later by arrangement). Friends of ATN/ATR society As all dedicated ATR followers will probably have noticed, the Friends of ATN/ATR society (which is the legal institution behind the newsletter/review) has not managed to hold an annual general meeting since 2017. Unfortunately, the planned meeting for this year, which was to take place during the NESAT XIV conference in Finland, was also cancelled, as the event only took place online. In order to update the society statutes to the new online working conditions and the fact that ATR no longer has a membership group defined by the subscription to the review, the ATN/ATR board suggests making minor changes to the statutes. These will be put to the vote at the next AGM in 2022. The changes and the date of the AGM will be announced on the homepage of the society in good time. Membership of the Friends of the Archaeological Textiles Newsletter/Review society is according to the statutes of August 2007 given to the subscribers to ATN and ATR. Institutional subscribers are not given membership. Since 2020, ATN/ATR has been an openaccess journal. Access to it is no longer restricted to the membership and anyone interested in archaeological textile research can join the society. To apply for a personal membership (which is free until decisions about the annual fee for 2022 is made Friends of ATN/ATR society: Template for the membership application at the next AGM), please fill in the following template and return it to ulla.mannering@natmus.dk. The membership will be valid from the day the applicant receives a confirmation from the ATN/ATR board. The membership will be valid until written notice is received by the ATN/ATR board resigning membership, and as long as the membership obligations, including paying the annual fee (as and when applicable) are followed by the member. Access to the online AGM will only be sent out to members registered according to the above-mentioned procedure. Suggestions for the AGM’s agenda may be sent to the editors before 1 February 2022. If no further suggestions are received, the agenda will be as follows: • • • • • • • Election of a chair, if somebody so wishes The report of the board for the period since the previous annual general meeting Presentation and approval of the revised account of 31 December 2021 Minor changes to the statutes, and decisions concerning membership and subscription Election of four members of the board and one deputy member for the current financial year (January to December 2022) Election of an auditor and one deputy auditor for the current financial year Any other business For the statutes of the society, including the voting procedure, please consult the statutes section on https://www.atnfriends.com. Archaeological Archaeological Textiles Textiles Review Review No. No. 6363 The Editors 3 Articles Marianne Vedeler Golden textiles from Gokstad Abstract Viking Age textiles with interwoven gold threads are rare in Scandinavia. Two such textiles were found in a Viking ship grave at Gokstad in Norway. Apart from brief overviews, the materials, techniques and find context of these textiles have not been described in detail before. The silk and gold embroidery was made with stem-stitch and simple laid-work. A madderdyed silk thread was used to make the flower pattern, in combination with a thread consisting of 80% pure gold lamella with a core of silk. A similar gold thread was used in a narrow band found with the embroidery. The 4 mm to 4.5 mm wide band is probably a remnant of tablet weaving. Both embroidery and band were found hidden inside a hollow ridgepole holding up the roof of a grave chamber. The precious gold and silk materials and the time-consuming and specialised technology used to make the gold thread suggests that these textiles were very valuable. Keywords Viking Age, embroidery, tablet weave, gold threads Introduction The renowned Viking ship grave from Gokstad in Norway was discovered in 1880 and first published in 1882 (Nicolaysen 1882). The man’s grave, dating from about the year 900 CE (Bonde and Christensen 1993) revealed a variety of finds, including a very well-preserved Viking ship. The various textiles found in the grave have received little attention, apart from two articles summarising them and presenting an overview which date back to 1973 and 1988 (Hougen 1973; Ingstad 1988). This paper focuses on the materials, techniques and find context of an unparalleled embroidery and a golden band found at Gokstad. A peculiar find context Two textiles with gold lamella were found inside a burial chamber raised on the deck of the Viking ship. The timbered chamber was shaped like a tent, with a thick oak ridgepole running along the centre of the roof (fig. 1). The rafters were fixed to this horizontal beam, which was the highest point of the chamber (fig. 2). When the excavator, Nicolay Nicolaysen, and his team were about to remove this heavy ridgepole, it parted lengthwise into two equal sections. The massive tree trunk had been deliberately cut lengthwise, hollowed Fig. 1: The Gokstad burial chamber as shown in The Viking Ship Museum (MCH) at Bygdøy, Oslo, Norway until 2021 (Image: Kirsten Helgeland, Museum of Cultural History, University of Oslo) Archaeological Textiles Review No. 63 47 Articles Fig. 2: The hollow ridgepole holding the heavy roof of the chamber roof (Image: Marianne Vedeler) out and put together again, before it was raised in the chamber. Hidden from sight, inside the two sections, were blue clay and what would prove to be one of the Gokstad burial’s greatest treasures (Nicolaysen 1882, 48). One half of the ridgepole revealed exquisite textiles in silk and gold threads of which one was an embroidery so far without close parallels. The other half contained the remains of a band made of goldwrapped thread. In the museum catalogue, written in 1881, the gold textiles are described as: ‘Lumps of clay-mixed cloth that has been inter-woven with gold thread. On account of its current state, the only thing to be said is that it has been very fine and of dark color’ (Nicolaysen 1882, museum number C10459). Unfortunately, only very few fragments of the underlying textile described as a fine cloth of dark colour are preserved (fig. 3). However, the techniques, materials, function and provenience of the golden textiles may provide useful information for an interpretation as to why the textiles were hidden in this location. Embroidery The embroidery is small, measuring approximately 2.8 cm x 3 cm. The motif is floral, made up of two circles, one inside the other, that are punctuated by smaller flower shapes (fig. 4). Combinations of stem-stitch and simple laid-work have been used to make the pattern. Laid-work is formed by laying one or more threads on the fabric and sewing over them at intervals with another thread to secure them in place. By using this technique, most of the precious metal will be visible on the surface of the embroidery. This is a technique that was commonly used in Chinese metal thread 48 embroideries in this period (pers. com. Yarong 2013) However, it is also present in embroideries with gold found in the Valsgärde boat graves in Sweden, dating to the tenth century CE and in embroideries without metal thread found at the Oseberg ship grave from Norway dating to the early ninth century CE (Nockert 2006, 325). Embroideries with metal in laid-work are also present in several other Scandinavian artworks from the later Middle Ages (Franzén and Nockert 2012, 42). In the Gokstad embroidery, a gold-wrapped silk thread has been used in combination with threads of pure silk (Skals 2015). The high-quality materials used emphasise the high status of the grave for an elite member of Norse society. In sections of the embroidery, two gold threads lie parallel but are sewn separately with simple laid-work. Traces of red colour can be seen on the s-spun silk thread used to stitch down the gold lamella. There is approximately 1 mm between the couching stitches in the laid-work. An s-spun silk thread has also been used to form the stemstitch contours of the flower pattern. Two samples of silk thread have been analysed for dye. A high-performance liquid chromatography (HPLC) coupled to a diode array detector (DAD) was used. When the embroidery was studied in a simple stereo loupe, two colours were visible. One was a faint red colour and the other what seemed to be a golden colour. When the two samples were analysed, results showed that both samples had a very similar composition, consisting of alizarin purpurin, pseudopurpurin, and anthragallol. These are markers of madder dyestuff. But the quantity of alizarin and anthragallol turned Fig. 3: Box containing remains of clay and fibres. A wool fibre was found here with one of the gold threads. The original museum label states “Remains of gold in-woven cloth” (Image: Marianne Vedeler) Archaeological Textiles Review No. 63 Articles Fig. 4: Gold and silk embroidery found inside the ridgepole (Image: Ellen Cathrine Holte, Museum of Cultural History, University of Oslo) out to be much higher in the golden than in the red sample. The differing quantities of these components suggest that different recipes were employed to obtain different colours on the thread (Łucejko et al. 2021, 2284). Madder can produce a wide range of colours from yellow and orange to scarlet red and almost brown (Eastaugh et al. 2008, 250). It is therefore difficult to know exactly what tint the embroidery once had but it is probable that it was made in two shades of red in combination with the gold thread. The embroidered floral pattern does not have any near parallels in Nordic finds. Stem-stitch and laid-work were used in silk embroideries from the ninth century CE Oseberg grave but the patterns in these embroideries are either animals enclosed in medallions, geometrical patterns, crosses or a combination of animals and leaf vines (Nockert 2006, 325–337). Stem-stitch was also one of the techniques used in the wool embroideries found at Bjerringhøj/ Mammen in Denmark dated to the late tenth century CE. However, no similar flower-patterns were identified. Instead, there are masks, animals and various acanthus leaves (Hald 1950). Embroideries detected in one of the two tenth century CE boat graves from Valsgärde in Sweden revealed a pattern of foliate derivation (Graham-Campbell 1980, 102). Gold band The other textile found inside the roof beam was a narrow band, approximately 4 mm to 4.5 mm wide and made with gold lamella (fig. 5). Five strips of the band were recorded, measuring 63.1 cm in length. It is probably remnants of a tablet-woven band, but only the weft threads of gold remain. The lamella is made of narrow strips, about 0.4 mm to 0.5 mm wide. These were wound in an S-twisted spiral around a core, which is now missing. The core was probably made of silk, as is the core of the gold thread used for the embroidery. The band also once had a warp made of another material which is now lost. Only the metal threads in the weft are preserved. It is therefore not possible to reconstruct the precise pattern or the exact weaving technique used or even be sure it was made with tablets. However, there are several features which suggest that the Gokstad band is the remains of a tablet weave. The tablets used in tablet weaving are usually squareshaped frames made from antler, bone or wood. These simple, punctured frames provide a surprisingly wide variety of technical opportunities. Although relatively few bands with metal thread dating to the Viking Age have been found, a variety of tablet-woven bands made of wool are preserved (Geijer 1938; Hald 1950, 227–242; Nockert 2006, 141–159). The ones that are in relatively good condition show a variety of techniques and a high degree of artisanship. They suggest that brocaded bands were most common during the Viking Age (Nockert 2006, 142–143). Weft threads can be woven to float so that most of the thread is on display on the surface of the fabric; this effect is called brocading. The band from Gokstad is quite narrow. It was possibly part of a broader band, forming a narrow brocaded patterned area within it. Other examples of this type, which date to the same period, also survive. Excavation of grave Bj 962 in Birka, Sweden, revealed Archaeological Textiles Review No. 63 49 Articles a man’s burial dated to the tenth century CE. A tabletwoven band with a chequerboard pattern made with silver-wrapped thread was found on a penannular brooch in the chest region of the deceased. The silver lahn (metal wrapped thread) in the band was used as a brocaded weft that formed the chequerboard decoration (Larsson 2007, 175). There are other examples of tablet-woven bands in which the metal-wrapped threads are almost covered by the warp; for example, five bands worked with silver and silk threads dated to the end of the tenth century CE were found in the Hørning grave in Jutland, Denmark (inventory number NM C31320, Hedeager Krag and Ræder Knudsen 1999, 164–165; Hedeager Krag 1999, 429–433). In one (band IV), the silver threads are partly covered by silk. The band is approximately 5 cm wide, and therefore much broader than the golden band from Gokstad. Preserved patterns on surviving tablet-woven bands dating to this period are all geometrical. In Ness in Norway (Arntzen 2011, 33–42), Önsvalla in Sweden (Nockert 1982, 204–206) and Tampere Vilusenharjo in Finland (Tomanterä 1978, 17) geometrical patterns were visible in the bands when documented. In the case of the Önsvalla band, the pattern forms stars and zig-zag patterns made with alternating gold and silver threads worked into the 1.3 cm wide band (Nockert 1982, 204–206). A number of tenth century CE bands made with another type of metal thread have been found in Birka, Sweden. These are made in drawn metal wire without a fibre core but, as with the types of bands discussed above, geometrical patterns dominate (Geijer 1938; Nockert 2006, 142–143). The same type of geometrical patterns can also be found in a number of tabletwoven bands made of wool. These were probably made locally (Nockert 2006, 150–151). In Oseberg in Norway, a complete tablet loom with tablets and work-in-progress fabric on it were found in the grave. There can be no doubt that such bands were produced in Scandinavia. The question is whether the metalwrapped thread was imported or locally produced. It is possible that the tablet-woven bands made with metal-wrapped thread were woven in Scandinavia, but with imported threads made of silk, gold and silver-wrapped lahn. Gold-wrapping technology used in the threads from Gokstad Metal thread was used in combination with other textile fibres in both the embroidery and the narrow band found in the Gokstad grave. The metal content used in the thread as well as the production techniques can provide valuable information about their provenance. There are several ways to make metal thread. In general, four kinds of gold and silver threads appear in fabrics from the early and high Middle Ages (Hardin and Duffield 1986, 233). Drawn metal wire, flat hammered strips, cut metal-wrapped thread and metal/membrane-wrapped thread. Two of these are made entirely of metal, either as drawn metal wire or as flat hammered strips. This type of thread dominates the finds from Birka in Sweden and was used in parallel with metal-wrapped silk thread (Larsson 2007, 146). In a tenth century CE find from Vangsnes in Norway, passementerie with drawn gold wire was found with thread made of flat strips (Hougen 1973, 79). The threads from Gokstad are of another kind. In Fig. 5: Remnants of a tablet-woven band from Gokstad: a - with gold-wrapped silk thread used in the weft; and b - the warp and weft (?) threads that held the gold threads in place on the surface are not preserved (Image: Ellen Cathrine Holte, Museum of Cultural History, University of Oslo) 50 Archaeological Textiles Review No. 63 Articles Fig. 6: Detail of the gold-wrapped thread used in the embroidery (Image: Calin Steindal, Museum of Cultural History, University of Oslo) both the embroidery and in the band, the metal was wrapped around a fibre core forming a thread termed lahn. There are several ways to make a metal-wrapped thread with a core of fibre. One was to hammer out or cast plates of gold or silver into extremely thin sheets. The metal sheets were cut into narrow strips called lamella and wound or “spun” around a core of silk (or linen) with a distaff. Another method, found in high medieval textiles, was to use gilt membrane strips that were created by beating gold sheets onto an animal membrane (Bravermanová et al. 2020, 126, 143). These were then cut into lamellae and wound or “spun” around a core of silk. In addition to these two types of metal wrapping, some Persian threads, mostly used in the 14th century CE, were made of gilded silver. In this type of thread, a thin layer of gold was laid over the silver strip. The process of “water-gilding” used an amalgam of mercury to fix the gold onto the silver surface (Hardin and Duffield 1986, 245). To answer what kind of materials and techniques have been used to make the gold thread from Gokstad, the metal thread was investigated using scanning electron microscopy (SEM) and optical microscopy studies. The aim of these studies was to identify the chemical compounds of the metals and the method used for making the thread. The optical microscopy study revealed that the metal-wrapped silk thread from Gokstad is not rounded but consists of a flat lamella with a thickness varying between 11 µm and 25 µm (fig. 6). Sharp edges along the strips suggest that it was cut from a metal sheet (Karatzani 2008, 412). The gold lamella was cut into approximately 130 µm to 170 µm wide strips and then wound around a core of silk forming an s-twisted spiral. The same method was used to make the metal thread in both the embroidery and the band. Fig. 7: Trace elements on the surface of a metal thread from Gokstad showing a high percentage of gold (Au) mixed with a smaller percentage of silver (Ag) and a very small percentage of copper (Cu) (Image: Calin Steindal, Museum of Cultural History, University of Oslo) Archaeological Textiles Review No. 63 51 Articles The next question was whether the lamellae were made of gilded silver, pure gold or a combination of metal and animal membrane. The gold thread from Gokstad contains a large quantity of gold. The SEM measurements show between 82% and 88% of gold mixed with a smaller quantity of silver (between 10% and 18%) which produced a soft and flexible thread. A very small amount of copper is also present but there are no traces of mercury (fig. 7). The investigation revealed no traces of animal membrane. Cross-section SEM energy dispersive X-Ray spectroscopy (SEM/EDS) was used to obtain information about whether the threads were gilt or not. To detect any trace of layers in the lamella, a scan was made of the metal edge. If the gold had been beaten onto a membrane or if different layers of gold and silver had been used, traces of the different layers would be visible, at least in fragments where some of the organic core is preserved. The scan showed that gold and silver is evenly distributed right through the metal. The conclusion must therefore be that there are no traces of a foil made of other metals or substances in the thread. It is made from one single plate of gold mixed with a smaller amount of silver and an even smaller amount of copper. There were also significant quantities of silicates found on both the inside and outside of the metal thread. This could have come from natural ochre in the earth of the grave, or perhaps traces of an unknown dyestuff. The gold lamellae were once wound around a core of organic fibre, which has now decomposed. This is especially so for the tablet-woven band. For the embroidery, the situation is better. Here, both the silk thread used in the embroidery and some small traces of the core inside the gold thread have been preserved. Fibre analysis of the fragments inside the gold lamella showed they were very degraded. It is most probably red-dyed silk (Skals 2015, 2, fig. 9). Red colour is visible under the fluorescence microscope but the fragment is too small to be analysed by HPLC. In nearly half of all sites dated to Viking Age Scandinavia where silk was found, silk thread was used in combination with gold or silver thread. Silver metal thread, as well as gold, was used in both embroidered textiles and tablet-woven bands (Vedeler 2014, 55). The gold and silk-wrapped threads have several advantages over the solid drawn gold wire. First, the production demands less of the precious metal, and a lighter thread can be made. Such a thread is relatively flexible, making it easier to work on a textile fabric. The silk core makes the thread flexible and less likely to break than those without such a core. 52 Nevertheless, a wrapped gold and silk thread is still subject to abrasion. The question of provenance The method used to make the metal thread found at Gokstad corresponds with analysis of other historical samples, among those from an Islamic cemetery in Valencia dated to the 11th to 13th centuries (Martínez et al. 2018). The gold and silk threads found at Gokstad were probably imported but a thorough and broad comparison with other findings from the eighth to tenth centuries must be made to investigate the question of provenance further. Bands made with metal-wrapped organic thread comparable to that in Gokstad are known from many European archaeological sites dated as early as the seventh century (Pritchard 1988, 152). Similar bands are also found across the Nordic area in modern Norway, Sweden, Denmark, and Finland. Either silver or gold was used, although in some cases, both metals are found in combination with silk (table 1). The place of production for metal threads from Scandinavian Viking Age graves has so far been based on visual comparisons conducted before new electrochemical methods were available (for example, Geijer 1972, 275). Comparative studies of Scandinavian material are therefore limited. The tablet-woven band and the embroidery from Gokstad may have been produced at different sites, even if the metal-wrapped thread used in the two is very similar. It is also possible that the gold and silk thread was produced with silk from one area and gold from another. The pattern and style of the Gokstad embroidery is different to other known embroideries found in the Viking world and may have been imported as a finished product. The floral pattern is unlike the animal styles and geometrical patterns that are typical of Scandinavia. However, the tablet-woven bands found in Viking Age graves are generally assumed to be Scandinavian products. Tablet-woven bands with geometrical patterns have been found in some Early Iron Age graves, and the tradition continues in Scandinavia into the Middle Ages testifying to the presence of a specific Scandinavian language of design and technique (Hald 1950; Hougen 1935; Nockert 2006; Raknes Pedersen 1988; Schlabow 1976). With this in mind, it is highly likely that the tablet-woven band from Gokstad was produced locally made of imported raw materials. There are several possible places of origin for the silk. Metal-wrapped silk thread became popular in China from the eighth century CE. Gold for the decoration of silk textiles was extensively used in China during Archaeological Textiles Review No. 63 Articles Table 1: Published graves dating to the Viking Age in Scandinavia containing “spun” metal-wrapped thread Archaeological Textiles Review No. 63 53 Articles the Song (969 CE to 1279 CE), Liao (90 CE to 1125 CE) and Jin (1115 CE to 1234 CE) dynasties (Chen Yan Shu 2011). Both hammered gold sheets wrapped around a silk core and gold sheets beaten onto animal membrane were used, especially in north-western China during the ninth and tenth centuries CE. A collection of beautiful silk fabrics with gold embroideries from this period is preserved in the Peking University Collection, Beijing. Gold thread is most often used in laid-work. One example of a gold and silk embroidery is a pair of exceptionally well-preserved boots decorated with gold and silk thread from Ye Mao Tai village, dating to the Liao Dynasty (916 CE to 1125 CE). A geographically closer provenance for metal thread would be the large Persian production area, where many of the woven silk fabrics found in Scandinavian graves were made. Persia itself had very little gold and had to import it for use in textile production. Written sources from the late Middle Ages and Early Modern times mention India, Aleppo, Constantinople, Armenia and Georgia as sources of gold imported to Persia (Hardin and Duffield 1986, 233). Near Eastern metal threads contain little copper, as does the Gokstad thread. Unlike the gold thread from Gokstad, almost all gold and silver threads dating to early medieval period Persia which have been investigated were metals attached to an animal membrane (Hardin and Duffield 1986, 239). This makes it less probable that the gold and silver threads from Gokstad originated from this area. Embroideries of gold and silk were probably also produced in England during the Anglo-Saxon period. Bands trimming the Anglo-Saxon embroideries attributed to St Harlindis and St Relindis are examples of textiles with “spun” gold threads (Spies 2000, 60-61). The same stitches were being used there as in Gokstad, although the surviving designs are different (Lester-Makin 2019). It has been suggested that silk embroideries found at Oseberg, a grave from the same district as Gokstad, have an AngloSaxon origin (Nockert 2006, 337). However, there are several difficulties in tracking the origin of these silk embroideries and their provenance remains uncertain (Lester-Makin 2019, 85) A survey of early tablet-woven bands suggests that all preserved examples with spun gold threads dating from the eighth to the 11th centuries CE have a Germanic, Anglo-Saxon, Scandinavian or possibly Islamic provenance (Spies 2000, 60–61). The question of the origin of the materials used in the production of the bands is complex. Documented trade routes from central Europe to Scandinavia also raise the question of the potential production of gold threads 54 in Europe, which needs further exploration. An answer to the question as to where the Gokstad gold thread was produced may be provided by a thorough electrochemical comparison with other preserved finds dating from the eighth to the tenth centuries CE. Use of the band and the embroidery The band and the embroidery were probably sewn onto a textile at the time they were discovered because Nicolaysen reported that he had found a long piece of what he thought was silk (Nicolaysen 1882, 48). Unfortunately, very few traces of such a textile exist today, and the Museum of Cultural History has no documentation on any later separation of fragments. The textile was, according to the museum catalogue, mixed with blue clay. The gold-wrapped thread and the silk used in the embroidery and in the tablet-woven band were obviously more sturdy than the fabric to which they were probably once attached. However, when analysing the gold lamella, a small fragment of thread attached to the gold sample was found. There is no doubt that this is a tiny piece of wool. The characteristic cuticle in the fibres is preserved, and in many cases so are the medulla (Skals 2015, 2, fig. 4). This small fragment of wool could be remnants of the ground fabric to which the band and the embroidery were once attached. What Nicolaysen thought was silk was a piece of fine wool cloth. Tablet-woven bands have been used for several purposes in the Viking Age. Apart from decoration on clothing, they were used as borders on the tapestries found in the ninth century CE Oseberg grave (Nockert 2006, 385; Vedeler 2019). None of these has remains of any metal-wrapped thread. However, examination of similar finds in Scandinavia in which metal-wrapped thread has been used makes it clear that when attached textiles are preserved, they have been interpreted as parts of clothing (table 1). They are, however, not associated with a specific type of garment. Embroidery and tablet-woven bands with metal-wrapped thread were most probably used as decoration on more than one type of clothing. At Valsgärde in Sweden, at least two graves contained metal-wrapped thread. In grave number 12, dated to the tenth century CE, there were fragments of silk samite interpreted as clothing (Grave Nos 12/918 and 12/973, Bender Jørgensen 1986). They are embroidered with a silver-wrapped silk thread. A total of 12 cast bronze buttons were also found. Another grave, similarly dated to the tenth century CE, contained a silk samite with metal embroidery depicted as either a collar or cloak trimming (Grave no. 15/5915, Bender Jørgensen 1986; Graham-Campbell 1980, 102). Traces of brocading in wool and gold have Archaeological Textiles Review No. 63 Articles also been found (Bender Jørgensen 1986, 161). Another example comes from Önsvala, Nevishög in Sweden. Two graves at this rich cemetery revealed textiles with metal-wrapped thread (Grave 24 and 2, Larsson 1982, 146–149, 155–167). In both cases, the textiles were interpreted as clothing (Nockert 1982, 204–206). Embroideries and bands with metal-wrapped thread were sewn onto a variety of different textile qualities. The tablet-woven silk band with silver and gold threads found in a man’s grave at Tampere Vilusenharju, Finland, is an example where the band was attached to a twill fabric (Tomanterä 1978, 17). A brooch attached to the band indicates that it was part of the man’s clothing. In Birka, Sweden, 53 very rich graves have so far revealed textiles with metal thread, all except two with thread made of metal wire (Larsson 2007, 163175). Most of these are tablet-woven bands, which seem to have been used in the same way as the bands with metal-wrapped organic thread. Over 50% of these tablet-woven bands were found around or in connection with the deceased’s head and were probably part of various forms of headdresses for both men and women (Larsson 2007, 176). Another 20% of the bands from Birka were found with a knife, a sword or other grave gifts. Approximately 20% were found in the body region often near the shoulders or breast. The latter were probably used as decoration on clothing (Larsson 2007, 177). Fragments of gold embroidery found together with silver passementerie at a ship burial from Ladby, Denmark was also interpreted as remnants of clothing (Hedeager Krag 2004, 81). A grave found at Fyrkat in Denmark contained a gold-wrapped thread attached to a blue wool cloth. The textile could be remnants of an embroidery with an unknown function (Grave 4, Østergård 1977, 103). A 1.5 cm wide band made with silk and silver threads found at Ness in northern Norway attached to the back of a man’s skull (Arntzen 2011, 33–42), and could be remnants of a headdress. These and several other examples indicate that embroideries and various bands made with metalwrapped thread were used as decoration on different kinds of clothing and on different parts of the garments linked to both men and women. What was the function of the textiles from Gokstad? How were they used? The golden textiles were certainly an extraordinary find - not least because of their original context. Why place the most precious of textiles hidden from sight inside the ridgepole? It must have required extra work to hide the textiles in this way. The significance of the location deserves a study of its own. Nevertheless, the peculiar find context of the Gokstad embroidery and band also raises many questions about their purpose and use before burial. Finding textiles with ridgepole could lead to the conclusion that the textile had been used as a wall decoration, and that this function was continued inside the grave chamber by simply fixing the textile to the beam to hold it in place. Looking at the quality and nature of the textiles this seems highly unlikely. Furthermore, the golden textiles were found inside the hollowed pole, not attached to the outside. This makes it even more unlikely that the textile was used as a wall decoration. Similar finds across Scandinavia suggest that textiles like these were used as decoration on clothing. In the case of the Gokstad embroidery, it was probably originally sewn onto a garment made of fine wool. Considering the context and the quality of the golden decorations, it seems more likely that this was part of a garment hidden from sight between the two parts of the ridgepole. Conclusion The materials and techniques used in the golden textiles from Gokstad resemble a very limited number of other finds from high status graves dating to the Viking Age in Scandinavia. They bear witness to good artisanship and prosperity. The embroidery, formed in a floral shape, was made with red silk and gold thread. The silk thread may have been in two shades of red, while the gold lamellae was also wrapped around a core of red silk. The preserved band is probably a remnant of a narrow tablet weave, but only the weft threads of gold remain. It is, however, difficult to trace the place of production, especially of the gold thread. The spun gold thread used in both textiles could have been imported as thread and then used to craft the textiles locally but these could also have been produced in a number of different areas discussed above. The question of where the materials originated is complex, and their provenance remains uncertain. Both textiles were probably originally used as decoration on clothing. Small traces of remaining fibres indicate that the garment was made of fine wool cloth. The deceased was probably not dressed in this at the funeral. The peculiar find context inside a hollow beam in the burial chamber raises questions about the meaning and purpose of these textiles in life and death. Acknowledgements The analysis of gold, fibres and dyes was funded by the Gokstad Revitalised project. Many thanks to Hartmut Kutzke for reading the text on gold technology and for Archaeological Textiles Review No. 63 55 Articles operating the SEM in cooperation with Calin Steindal at the University of Oslo. Thanks also to Irene Skals at the National Museum of Denmark for doing the fibre analysis, and to Jeannette J. Łucejko and Illaria Degano at the University of Pisa for analysing the dyestuffs. My deepest thanks to Wang Yarong at the Chinese Academy of Social Sciences, who informed me about the use of gold-wrapped silk thread in Chinese finds and who also gave me the opportunity to see the collection of Tang silk at the Peking University in Beijing. I also owe thanks to two anonymous reviewers who gave me valuable advice and tips for improving this article. Bibliography Arntzen, J. E. (2011) En ”palisadehaug” fra yngre jernalder i Nord-Norge? Nicolay 114, 33. Bender Jørgensen, L. (1986) Forhistoriske textilier I skandinavien, Prehistoric Scandinavian Textiles. Copenhagen: Det kongelige nordiske oldskriftselskab. Bonde, N. and Christensen, A. E. (1993) Dendrochronological dating of the Viking Age Ship Burials at Oseberg, Gokstad and Tune, Norway. Antiquity 67, 575–583. Bravermanová, M., Brězinová, H. and Vichová, J.B. (2020) High Medieval textiles of Asian and Middle Eastern provenance at Prague castle, Czech Republic. Archaeological Textile Review, 62, 126–143. Chen Yan Shu (2011) 陈彦姝, 清华大学美术学院来源: Song, Liao and Jin silk adorned with gold. Decoration Magazine 1/12. Beijing: Tsinghua University. Eastaugh, N., Walsh, V., Chaplin, T. and Siddall, R. (2008) Pigment Compendium. A Dictionary and Optical Microscopy of Historical Pigments. London/New York: Routledge. Franzén, A. M. and Nockert, M. (2012) Prydnadssömmar under medeltiden. Stockholm: Kungl. Vitterhets Historie och Antikvitets Akademien. Geijer, A. (1938) Birka III. Die Textilfunde Aus den Gräbern. Uppsala: Kungl. Vitterhets historie och antikvitetsakademien. Geijer, A. (1972) Ur textilkonstens historia. Lund: Gleerup. Graham-Campbell, J. (1980) Viking Artifacts: A Select Catalogue. London: British Museum. Hald, M. (1950) Olddanske tekstiler.Komparative tekstil- og dragthistoriske studier paa grundlag af mosefund og gravfund fra jernalderen. Copenhagen: Det kongelige nordiske oldskriftselskab. Hardin, I. R. and Duffield, F. J. (1986) Characterization of Metallic Yarns in Historic Persian Textiles by Microanalysis. In H. Needles (ed.) Historic Textile and Paper Materials. Advances in Chemistry. Washington DC: American Chemical Society, 231–252. 56 Hedeager Krag, A. (1998) Dress and power in Prehistoric Scandinavia c. 550-1050 AD. In L. Bender Jørgensen and C. Rinaldo (eds), Textiles in European Archaeology. Report from the 6th NESAT symposium, 7-11th May 1996, Borås, GOTARC Series A, Vol. 1. Göteborg: Institutionen för Arkeologi, Göteborgs Universitet, 125–130. Hedeager Krag, A. (1999) Fränkisch-byzantinische Trachteinflüsse in drei dänischen Grabfunden des 10. Jahrhunderts. Archäologisches korrespondenzblatt 29(3), 425–444. Hedeager Krag, A. (2004) New Light on a Viking Garment from Ladby, Denmark. In J. Maik (ed.), NESAT VIII, Priceless Invention of Humanity. Textiles, Acta Archaeologica Lodziensia 50(1). Lodz: Lodzkie Towarzystwo Naukowe, 81–86. Hedeager Krag, A. and Ræder Knudsen, L. (1999) Vikingetidstekstiler. Nye oppdagelser fra gravfundene i Hvilehøj og Hørning. Nationalmuseets Arbejdsmark 1999, 159–170. Hougen, B. (1935) Snartemofunnene: studier i folkevandringstidens ornamentikk og tekstilhistorie. Norske Oldfunn VII. Oslo: Oldsaksamlingen. Hougen, B. (1973) Gulltråd fra Gokstadfunnet. Et fragment. In E. Kivikoski, P. Sarvas and A. Siiriäinen (eds), Honos Ella Kivikoski. Suomen Muinaismuistoyhdistyksen aikakauskirja 75. Helsinki: Finska fornminnesföreningen, 75–81. Ingstad, A. S. (1988) Textiles from Oseberg, Gokstad and Kaupang. In E. Munksgaard, L. Bender Jørgensen and B. Magnus (eds), Archaeological Textiles. Report from the 2nd NESAT symposium 1.–4.V 1984. Arkæologiske skrifter 2. Copenhagen: Arkæologisk Institut, Københavns Universitet, 133–48. Karatzani, A. (2008) Study and analytical investigation of metal threads from Byzantine/Greek ecclesiastical textiles. X-Ray Spectrom. 37, 410–417. Larsson, L. (1982) In Gräber und Siedlungsreste der Jüngeren Eisenzeit bei Önsvala im südwestlichen Schonen, Schweden. Acta Archaeologica 52, 129–196. Larsson, A. (2007) Klädd krigare. Skifte i skandinaviskt dräktskick kring år 1000. Uppsala: Uppsala Universitet. Lehtosalo-Hilander, P. (1982) Luistari I. The Graves. Helsinki: Suomen Muinaismuistoyhdistyksen Aikaklauskirja 82(1). Lester-Makin, A. (2019) The Lost Art of the Anglo-Saxon World. The Sacred and Secular Power of Embroidery. Oxford & Philadelphia: Oxbow Books. Łucejko, J. J., Degano, I. and Vedeler, M. (2021) Textile dyes from Gokstad Viking ship’s grave. Heritage 2021(4), 2278–2286. Martínez, B., Piquero-Cilla, J., Montoya, N., DoménechCarbó, M. T. and Doménech-Carbó, A. (2018) Electrochemical analysis of gold embroidery threads Archaeological Textiles Review No. 63 Articles from archeological textiles. Journal of Solid State Electrochemistry 22(7), 2205–2215. Nicolaysen, N. (1882) Langskibet fra Gokstad ved Sandefjord. Kristiania: Alb. Cammermeyer. Nockert, M. (1982) Textilfunde von Önsvala, Nevishög KSP, Schonen. In L. Larsson (ed), Gräber und Siedlungsreste der Jüngeren Eisenzeit bei Önsvala im südwestlichen Schonen, Schweden, Anhang II. Acta Archaeologica 52–1981, 202-208. Nockert, M. (2006) Brickväv och brickband, Siden, Silkebroderi. In A. M. Christensen and M. Nockert (eds), Tekstilene. Osebergfunnet bind IV. Oslo: Kulturhistorisk museum, Universitetet i Oslo, 141-149, 277–337. Pritchard, F. (1988) Silk braids and Textiles of the Viking Age from Dublin. In E. Munksgaard, L. Bender Jørgensen and B. Magnus (eds), Archaeological Textiles, Report from the 2nd NESAT symposium 1.-4. V. 1984. Arkæologiske skrifter 2. Copenhagen: Arkæologisk Institut, Københavns Universitet, 149–161. Raknes Pedersen, I. (1988) The Analyses of the Textiles from Blindheim, Giske. In E. Munksgaard, L. Bender Jørgensen, B. Magnus (eds), Archaeological Textiles. Report from the 2nd NESAT symposium 1.–4. V. 1984. Arkæologiske skrifter 2. Copenhagen: Arkæologisk Institut, Københavns Universitet, 116–125. Rimstad, C., Mannering, U., Jørkov, M. L. S. and Kanstrup M. (2021) Lost and found: Viking Age human bones and textiles from Bjerringhøj, Denmark. Antiquity 95,1–18. Østergård, E. (1977) Tekstiler. In E. Roesdahl, Fyrkat. En jysk vikingeborg. II. Oldsagerne og gravpladsen. Copenhagen: Nationalmuseet. Schlabow, K. (1976) Textilfunde der Eisenzeit in Norddeutschland. Göttinger Schriften zur Vor- und Frühgeschichte 15. Neumunster: Karl Wachholtz Verlag. Skals, I. (2015) Identifikation af textilfibre fra Gokstad-fundet. Report project 50318. Copenhagen: Nationalmuseet i København. Spies, N. (2000) Ecclesiastical Pomp Aristocratic Circumstance. A Thousand Years of Brocaded Tabletwoven Bands. Jarretsville: Arelate Studio. Tomanterä, L. (1978) Appendix 2 in Terhi Nallimaa-Luoto: Tampereen Vilusenharju – nuoremman rautakauden kalmisto Pirkanmaalla. Master’s thesis. Karhunhammas 3, Turku: University of Turku. Vedeler, M. (2014) Silk for the Vikings. Ancient Textiles Series. Oxford: Oxbow. Vedeler, M. (2019) The Oseberg Tapestries. Oslo: Scandinavian Academic Press. Yarong, W., Chinese Academy of Social Sciences, Beijing, China – personal communication, 24 October 2013. Internet sources UNIMUS database (undated) https://www.unimus.no/ portal/#/ (accessed 1 November 2021) Archaeological Textiles Review No. 63 Author: marianne.vedeler@khm.uio.no 57