By Katherine Larson and Marta Kløve Juuhl

Editor’s note: This article appeared in Archaeological Textiles Review No. 64 (ATR), published in 2022 by the Centre for Textile Research, University of Copenhagen. A wealth of detail on how this technique may have been woven will be of interest to warp-weighted loom enthusiasts, but anyone interested in solving puzzles may also enjoy this article. Links to the article and its foundational study appear below, but first here is a summary from one of the authors, providing you with the background and highlights of studying double-cloth on the warp-weighted loom.

The coverlet-width warp woven for the 2022 double-cloth study. From ATR No. 64

Go Big or Go Home – The Importance of Textile Width

By Katherine Larson

Double-cloth, also called double-weave (dobbeltvev in Norwegian), has a history in Scandinavia. The somewhat complicated weave structure meant that this was not an every-woman technique, but for a knowledgeable few it did provide an effective means for creating free-form patterning on a simple loom: Set up two warps of contrasting color, exchange the warps in selected places, and now you were weaving a bridal procession instead of plain old stripes and plaids. What a break-through moment for the weaver who first figured this out! Some manipulation with pattern boards would be required to hold the contrasting warp layer apart for a weaving shed, but if you were clever enough to conceive of double-cloth in the first place, that part would be easy. 

This 17th century Norwegian reversible double-cloth textile, a fragment of a once longer piece, depicts riders on horseback, possibly representing a bridal procession. OK-10878, National Museum, Oslo.

Double-cloth has been known in parts of Scandinavia since at least the Viking Age. The occurrence of Persian double-cloth fabrics within the same time frame suggests that our clever weaver of the north may have seen an exotic textile from a distant land and set about adapting her home loom to reproduce this brilliant new idea.

A narrow medieval double-cloth fragment from Rennebu, Norway. The figures on the left possibly depict two women weaving on a warp-weighted loom. Museum of Cultural History, UiO, Oslo.

The ‘home loom’ of the north, of course, would be the warp-weighted loom, which was used for weaving reversible double-cloth for centuries in Scandinavia. Even when the technique’s flashier cousin came along – non-reversible double-cloth, woven on the treadle loom – the reversible technique still prospered for a while. A significant body of 18th century coverlets provides ample evidence for both of these double-cloth traditions, but while the non-reversible technique survived long enough to be documented, the reversible version fell out of use before anyone thought to describe the process. Yet over 80 coverlets in the reversible technique, many in excellent condition, provide mute testimony to the successful use of the warp-weighted loom to produce relatively complicated textiles. 

Non-reversible double-cloth lends itself to finer pattern elements than the reversible technique. Details of double cloth coverlets: reversible (left) OK 01696, National Museum, Oslo; and non-reversible (right) NF !910-0433, Norwegian Folk Museum.

The initial study

The coverlets of this gone-but-not-forgotten weaving tradition provided the starting point for my decade-long study to consider how reversible double-cloth might have been woven on the warp-weighted loom. Over the course of many research trips, I found that details inadvertently ‘recorded’ in the coverlets by the weavers themselves provided a surprising amount of information about how the loom may have been set up, while experimentation with various possibilities on the warp-weighted loom served to rule out a few ineffective options (most notably the natural shed). Based on these findings I proposed a possible method for weaving double-cloth on the warp-weighted loom. 

Studying a double-cloth coverlet at the National Museum in Oslo, 2009. Photo: K. Larson

Studying the medieval Lomen double cloth textile. Photo: K. Larson

As with any warp-weighted loom research, Marta Hoffmann’s groundbreaking documentation of plain-weave methods on the warp-weighted loom (The Warp-Weighted Loom 1964) provided the foundation for this research. But double-cloth is more complicated than plain-weave, and even Marta Hoffmann’s carefully qualified speculation, that double-cloth patterns may have been picked by hand instead of with the use of additional sticks, did not stand up well to observation, experimentation and speculation. 

The word speculation deserves repeating here: not for naught was Hoffmann so careful when straying from known facts. While my proposed method for weaving double-cloth was based on reasoned guesses and experimental results, it was still speculation. It was not entirely surprising, then, when a flaw emerged, all of which underscores the importance of Hoffmann’s documentation of an actual living weaving tradition. 

The 2022 study

Words to live by: What works at one width sometimes does not work at another. The error in the proposed weaving method seems obvious in retrospect, but textile width was just one of many factors in the initial study. Warping a loom for double-cloth takes a long time, and the study’s experimental textile (30 cm) was designed to require minimal setup when traveling to various locations. This facilitated what was then a top priority: consulting with knowledgeable warp-weighted loom weavers while testing different loom configurations. Since simply forming a reasonable shed was an early stumbling block, the width of the experimental weaving was low on my list of concerns. However, once an apparent solution to serious warp-passage problems presented itself – a slight fanning of the warp threads accomplished by stretching the spacing chains – shed formation improved dramatically and I moved on to consider many other factors in weaving double-cloth. The element of textile width was not revisited. 

Evidence of pattern storing is found in Scandinavian reversible double-cloth textiles, and therefore experimentation with transferring stored patterns was part of the study.  This proved to be a straightforward process on the warp-weighted loom, especially in comparison to the treadle loom. Photo: K. Larson

The flaw in the proposed warp-fanning method emerged only after the initial study was complete. In contemplating the next logical step in understanding double-cloth, I decided to weave a piece based on the medium-width (~ 80 cm) procession textiles. This group of five decorative panels depicting a procession on horseback is generally thought to slightly predate the coverlet tradition. Unfortunately, shortly after I began pattern weaving on my procession textile it became obvious that the normally beneficial effects of stretching the spacing chains did not extend to the center of this wider warp. 

Procession double-cloth in progress. Photo: K. Larson

Warp-fanning had worked well for relatively narrow double-cloth textiles, and since four of the six surviving medieval Scandinavian examples were 30 cm in width or less, it is possible that this method served as a useful way to weave the earlier decorative panels. However, it clearly would not have accommodated even the modest width of the procession textiles, much less that of the coverlets, which ranged from 123 to 175 cm in width. I was back to square one.

I discussed my ongoing double-cloth research with Marta Kløve Juuhl, curator (now retired) at Osterøy Museum near Bergen.  Marta, an experienced weaver and a recognized expert in warp-weighted loom weaving, was intrigued by the idea of weaving double-cloth at full coverlet width. Thrilled to find someone else curious about this question, I arranged with Marta to start a project on one of the looms at Osterøy Museum in late February 2020. Unfortunately, the pandemic interrupted any idea of further joint work, but we continued to discuss the project regularly by email.

In warping the loom at Osterøy Museum, we used the same elements of loom setup as those in the initial study, with the exception of warp fanning. As expected, weaving did not go smoothly, requiring continual clearing of the sheds, a familiar problem. Nonetheless the small amount of weaving accomplished at Osterøy was instructive, identifying heddle length as the first issue to address. 

Setting up the coverlet-width warp at Osterøy Museum, February 2020; sewing the warp to the beam. From ATR No. 64

Preparing to attach the backmost row of weights to the warp. From ATR No. 64

Initial pattern weaving. From ATR No. 64

Fortunately, bringing the warp home to Seattle required cutting all those laboriously tied heddles. Since they would have to be retied anyway, I conducted a small interim experiment before putting the coverlet-width warp back on the loom. This study indicated two things: heddles held slightly taut by their warp threads, but also tied at different lengths for the forward and back layers, reduced heddle tangling during shed changes. The heddles were retied at the new lengths (all 1,494 of them) and heddle tangling receded as a problem.

Weights were the next issue, with the coverlet-width warp requiring a total of 100 weights at 1 kg each. While relatively narrow stones are plentiful in Norway, most of the field stones in the Northwest are nicely rounded thanks to thousands of years of glacial action. These were deemed too thick for the compact weight rows required. Suitably narrow weights were instead achieved by using coins, eight rolls of pennies per weight. Acquiring these coins at several banks felt distinctly odd, loading money into a reusable shopping bag while wearing a pandemic mask. Multiple banks were visited during this penny-acquisition spree due to withdrawal limits caused by a nationwide coin shortage, a problem no doubt exacerbated by the needs of warp-weighted loom research.

With the warp reestablished on the loom, the most likely avenue for improving loom function seemed to be the disposition of the weight rows. Four different configurations were tested, but one clearly outperformed the other three. In that configuration all weight rows were placed behind the shed rod, with the two rows of the forward layer held separate from their backmost neighbors by being tied at intervals to the shed rod. 

Four weight row configurations were tested in the 2022 study. Option ‘d’ was the most successful, with the two forward rows placed behind, but attached to, the shed rod. From ATR No. 64

In the final configuration, spacing chains of the two forward weight rows were tied to the shed rod. The chains formed slight arcs, providing additional stretching of the warps that was somewhat reminiscent of the earlier warp-fanning method. This effect may have assisted in warp passage. From ATR No. 64

This unusual weight-row disposition was inspired by an irregularity I had noted in several coverlets during the initial study. The method for establishing use of the warp-weighed loom in the first place had been a program of measuring irregularity in warp thread spacing. This irregularity is mostly disguised by the profusion of pattern in the coverlets, but is quite evident when measured. Oddly, several coverlets had a series of widely spaced warp areas that were somewhat evenly dispersed across the textile. It seemed possible that this represented a pull on the warp threads – or rather on the spacing chains – to hold the weight rows separate in some fashion. The success of the final configuration, with the forward weight rows tied to the shed rod, lends support to this possibility.  

Reversible double-cloth coverlets woven on the warp-weighted loom were the last chapter in a centuries-long tradition in central Scandinavia. Since available evidence points to narrower textiles being the primary decorative intent of medieval double-cloth, perhaps the appearance of the horizontal treadle loom had a role in encouraging this final chapter. The adoption of the treadle loom happened unevenly in Norway, but occurred at about the same time as the coverlets were woven. With the new loom’s superior capability to produce yardage, it seems likely that once it was acquired, the warp-weighed loom would have been abandoned. Yet for those with knowledge of the double-cloth technique, the now-empty looms may have represented an opportunity. After all, the slow pace of weaving a wider double-cloth textile would no longer be an impediment to the all-important need to produce yardage. 

The use of materials for double-cloth was also revisited in our 2022 study. Double-cloth has an inherent problem: weaving (and continually exchanging) two plain-weave warps in a space best suited to one. Medieval Scandinavian weavers addressed this problem by using a (smooth) layer of linen vs. a layer of colorful wool, but the Norwegian weavers of both the procession panels and the coverlets had transitioned to a new set of materials. In these textiles the linen layer was replaced by a sheep-brown warp that, as evidence from the initial study indicated, was predominantly composed of hair from the dual-coated northern European short-tailed sheep. This relatively smooth fiber likely functioned as well as linen, and the resulting textiles, now all sheep’s wool (or nearly so), may have had a more appealing hand than those made with a layer of linen. 

Perhaps a similar spirit of innovative was at work in solving the problems that arose when weaving double-cloth at a wider width. Since the warp-weighted loom’s signature natural shed was (presumably) ineffective for double-cloth, the weavers may have reimagined the function of this basic loom part, using the shed rod as an attachment point for holding the double-cloth layers apart.

Or perhaps not. We cannot know for sure how double-cloth was woven on the warp-weighted loom, we can only experiment and speculate. 

Detail from Procession, 2019. Photo: K. Larson

For those interested in weaving double-cloth on the warp-weighted loom, I’ll be interested to hear what you find.  kllarson@uw.edu

Katherine Larson lives on Bainbridge Island, Washington. She has a doctorate from the University of Washington, where she is an affiliate faculty member in the Department of Scandinavian Studies, and she is the author of The Woven Coverlets of Norway.

Read the full academic article by Larson and Marta Kløve Juuhl from Archaeological Textiles Review:  “Norwegian double-cloth: warp-weighted loom experiments in a complicated technique”

The background research that forms the basis for the 2022 article can be read with these links: the study was published in the Norwegian Folk Museum yearbook (2015); and an account of the research process appears in an earlier Norwegian Textile Letter article (2012). 

The entire issue of Archaeological Textiles Review No. 64 is available online and can be accessed here.  Articles in this issue that might be of interest to readers of the Norwegian Textile Letter include experiments with 2/1 twill on a two beamed loom based on textile fragments from 400 BCE Scotland, rare finds of linen garments from a 17th century Swedish grave, and reconstruction of a tablet-woven band from 3rd-4th century Germany.

Border from Procession. Photo: K. Larson

March 2023

Help support wonderful articles on Scandinavian textiles with a donation to the Norwegian Textile Letter. Thank you! Tusen takk!