Experimental Archaeology and Textile Studies: Part 3

Here's what I did. I made a 16 thread tablet woven starting border. I made 144 warp threads. I cut one end of the warp thread loops to create 72 long threads. I took my needle and threaded each warp thread into the band individually. I sewed this long strip of warp threads onto my loom beam, then carefully separated the warps into front and back threads. 72 in the front, 72 in the back. I tied a loomweight to each bundle, resulting in two loomweights with 72 threads each. 

This was a clunky solution to the problem. I had to deal with an issue I hadn't considered, but now that it was a potential problem I needed to think about for the Stage 2 of the experiment, I had a chance to form a contingency plan for it. The important point here is that I still managed to arrive at the proper setup I required and everything was ready for heddling and weaving. 

Here's where experimental archaeology in academia is very tedious and tiring. If this parameter had been an essential variable, I would not be able to devise such a solution. If I was working with a scrap of textile that had this warp density and a tablet woven starting band, I would not be justified in 'faking' it. As it stands for this experiment, however, it gave me an excellent question to ask senior textile studies researchers what they thought but didn't otherwise impede my progress or time tabling.

The actual weaving process for the Stage 1 pre-experiment went very smoothly. The sample was about 6cm wide and about a meter long. And it looked like a linen seat belt. Lol. Maybe that idea is worth exploring later when the automobile industry looks into more sustainable construction materials for their electric cars... I used a wooden weaving sword to push the weft into place, but archaeologically, there is no evidence for this at my case study sites.  How did an Iron Age weaver push the weft into place? Is a tool even needed? Can an iron/bronze weaving sword do the job? Again, it's something to consider with future research because in this case, it does impact the cloth I was making. 

I finished the strip after a week of weaving 'in my spare time', cut it off the loom, and marveled at what I did. Here are some thoughts I had:

1. Is the heaviness of a loomweight important to consider by itself, or is it more important to consider how mass is distributed across the thickness more important? At the start of this experiment, it was generally understood that heavy loomweights = coarse fabrics, but a 0.7mm thread is about a fingering weight yarn.
2. If you can weave with heavy loomweights and linen threads, what happens if you use wool? Wool, as many of you know, is far more elastic than linen.
3. What is the upper bound of tension that a linen thread can handle which considers its gauge. In other words, how much tension can a 0.7mm linen single handle before it breaks?
4. How might people have use this very dense cloth? It turned out warp faced, but is that a desirable outcome for an Iron Age population?

And I had more thoughts afterwards too, but these were the ones that seemed most interesting to pursue for now. Because the Stage 1 experiment did what it needed to do to convince me that I had worked out the potential bugs, I could do some creative things with the finished piece. I washed some pieces and left others unwashed. I experimented with different hemming principles and finishing techniques. Here are those results:

This is the finished piece that was folded on itself to show how rigid and dense the cloth was.

Here is a close up of the finished weaving to illustrate how dense the warps are. You can barely see the wefts!

This is a steromicroscopic image (apologies for the quality) of one woven sample that was washed. Note how the yarns have flattened out slightly

This is another steromicroscopic image of the sample where I smoothed the surface with a polished stone. It is more flattened than the above washed sample and the resulting fabric is very soft to the touch.

On this sample, I smeared a stick of beeswax over the surface after the cloth was washed. I thought about how someone might try to waterproof a cloth after it was woven. This is one way to do that, another way might be to dip the cloth in hot beeswax and allow it to dry flat. Anyway, it has a very strange surface, like touching a candle!

I'm sure by now that you probably want more details about the experiment. I'm writing up the parameters of my experiment and the reasons why in more detail, along with the insights this experiment garnered, to be put into a publishable format for a journal (still deciding where to send it). So many nuanced decisions were made during this process and it does convey the complexity involved in textile production--no surprise there, but it does make it difficult to maintain transparency in my methods without having to write an entire book that describes my reasoning to the fullest extent possible. 

I have presented a 'lite' version of this experiment (Stage 2, actually) while I was in Edinburgh at the end of October and got some great feedback, including positive words from Susanna Harris (who helped examine the Must Farm textiles). I've also submitted an abstract to a student conference for experimental archaeology scheduled for March 2020 in Sheffield. I'm hoping that with all the feedback from these sources, I'll get the paper written and submitted to a journal by summer next year. I'm thinking of EXARC as my first choice, but I'm open to other options too.

That itself is another point to make. Publishing the results of experimental work in academia can be very time consuming as well. It could be a full year or so from the time I started the experiment until it is published, if I'm lucky. And here I am publishing some of the details on my blog inside of two months. Also, some of you may recall that I have talked about loomweights and warp-weighted experiments in the past. I did! I have a blog (Part 0, Part 1, Part 2, Part 3, Part 4, Part 5) and video post about it as well. Check those out to see how far I have come with my research and experimental pursuits. I say some wrong things but getting access to academic material was very hard for me in those days.

I'll post more updates about the progress of this project as I make new videos and write up more about the Stage 2 of the experiments. Let me know what you guys thought about this content and whether you thought it was too long/short, too detailed/too vague. Thanks for reading!

Experimental Archaeology and Textile Studies: Part 2

With all the parameters outlined for the experiment design and the research question formatted to be a suitable hypothesis for the experiment, now came the hard part: how does it all 'work' together? Within every experiment, there are several small decisions made as part of this process. For all the preparation made prior to beginning, there will inevitably be contingencies you did not entirely account for.

For example, I decided to begin the warping process by making a tablet woven starting band, performed in the manner as seen by the Lapp weavers from Norway (Hoffman, 1974). To produce their wadmal and grene, they often wove a starting band while simultaneously integrating the warp. Not all weavers that Hoffman studied performed this method, but it does have archaeological presence in other parts of Europe. 

You can see in this photo how dense I needed the warp to be. The starting band also struggled to cope with the heaviness of the loomweight, but this is not a reflection on the experiment, just an issue with how to attach the warp to the loom.

I planned for as many of the possible contingencies I could think of that might go wrong. So, I made sure I had plenty of time to warp. I also had my assorted tools, including scissors, a needle, and my cup of water (for drinking!). I measured my warping pegs on my Kromski Harp loom (32") to confirm that I would get the correct length for my warp (side note: I have no idea how Iron Age people dealt with their warping problems, but this works for the experiment). My tablets (made of thick cardstock) were threaded with the same linen and stretched between two warping pegs. Then I began weaving and warping. After about 10 warps (or five times around the setup), I quickly realized that I wasn't going to achieve the correct density of warp threads I needed. Not even by a long shot!

I had a contingency for which I had no plan. I ended up unweaving everything, calling in my partner to help me (how much 'help' did any individual receive during Iron Age textile production?!), crying, and feeling hopeless. I was time pressed and didn't know how smoothly the rest of the experiment would go if I was already waylaid early on in the process. But, you adapt to adversity. I decided that no matter how I set up the tablet band, the way the cards turn prevents you from achieving a dense weaving setup. How can I get the density I wanted and also get the warp threads onto the loom? 

Let me back up to describe what I wanted the setup to be for the warp tension. After deciding on the loomweight shape and mass, and the spindle and spun yarn I needed, I had to calculate the warp tension for the setup. To keep the tension per thread no higher than 30g per thread, I determined that this would result in 72 threads per loomweight, or 144 threads for the woven fabric (tabby weave setup). This warp tension of 30g was determined by the Center for Textile Research (CTR) experiments to be a suitable upper limit for optimal weaving with wool. Why 72? The CTR experiments also determined that the starting width of your setup should be approximately equal to the total width of your loomweights (again, for optimal weaving). For a 2kg loomweight and a goal of 30g of tension per thread, that meant 72 threads (of about 0.9mm maximum gauge) would be ideal. This is tricky because there's a point where you can't get the warp threads so close that they either overlap or their fiber halo (more prominent in wool threads than in linen threads) causes them to stick and abrade as weaving commences. I spun a yarn that was about 0.7mm and bought a yarn that was similar in gauge (0.6-0.8mm). 

Here is my handspun linen that I used as a test. I photographed this with a camera mounted onto a stereomicroscope. (That's why it is a little fuzzy)

I decided that it would be good to start the experiment at this known upper limit when testing the linen because I could more easily link this back to wool and the other experiments conducted previously. But I was still stuck at this point where I needed to have a dense warp setup and didn't know how else to get the experiment going. I did the next best thing, which was 'fake' it. And I know that sounds a bit unscientific, but this particular problem was unimportant for the main question I was looking at since a starting band was not the focus. I had assumed that I should start with a tablet woven border, but there is no archaeological evidence for this in my case study sites. By removing this assumption, I could deal with the problem of getting the correct warp density I needed. 

Continued in Part 3!

Experimental Archaeology and Textile Studies: Part 1

This post is part of a series and I'm breaking it into sections for easier reading. If you have questions about a specific aspect, it'll also be easier to spot the information you want to know more about in a comment.

First, there are some important differences between the way I do experiments for Expertly Dyed and the way I must do experiments in academia. I'll get into the reasons why in a later episode and will post a video/blog explaining those differences specifically.

For now, I wanted to share Phase 1 of a two phase experiment I conducted during late August and part of September this year. Normally when I setup an experiment for Expertly Dyed, I apply a lot of craft knowledge and intuition and do a fair amount of 'let's see if this works'. I think many fellow crafters (of any sort, not just those who like textiles) will agree that this is often a legitimate way of exploring a craft. In fact, it is also a way that we explore the world as infants. We do this casually as we create recipes while standing at the stove and when we reach for the garlic granules, on a whim, we might substitute onion granules to see what that's like. We do these things naturally and implicitly and rarely do we ever consider our actions in a meta sort of way. 

Experimental archaeology in academia is not done this way. It has to be very well researched and there needs to be a lot of setup before you begin, or else you wind up digging yourself into a hole without a ladder and an imminent flood quickly approaches. Variables must be isolated and controlled so you can ensure that you can investigate the relationship between your independent and dependent variables. So, in my Phase 1, I produced a very skinny strip of linen fabric using heavy loomweights on my supervisor's warp-weighted loom. It is seemingly a silly step from a crafter's perspective because it seems completely irrelevant to your overarching goal, which is simply to weave something. I know that weavers will do tests to check their setup before they embark on a large project, just the same as a knitter will check their gauge before knitting a pattern that relies on fit. 

However, the nature of my experiments (which I will describe in a continuation of this series) required that I test my experiment parameters before I actually began the experiment. It was designed to be overly cautious because I did not have the luxury of time and I had my finances to consider as part of the experimental design. I used the case study sites that I’m researching as part of my PhD, which are Danebury and the Environs Sites. I selected these sites because they were all excavated by the same principle investigator (Barry Cunliffe) and well archived, and they could give me a sense of the ways textile tools were used and treated upon deposition within a small bounded landscape.

Here are a few considerations that I made for the initial design:

1.     I wanted to use the heavier loomweights from my dataset, which hovered around the 2000g mark. I am not aware of any published experiments where loomweights of this magnitude are used; many non-textile scholars have questioned whether such heavy loomweights could perform as such. Further, I opted to use a clay weight as my model since I had access to air dry clay, a suitable proxy, rather than chalk (which is the predominant material use for Danebury loomweights). I modelled the shape to be triangular, in part because this shape has been contested as whether it was suitable as a loomweight or as oven architecture. To be brief, clay triangular weights at the Danebury sites are sometimes found associated with collapsed ovens, leading to the contestation that they were originally used as loomweights. In reality, these objects (and other textile tools) are found in a variety of contexts so it isn't a very clear distinction. A previous proof-of-concept experiment I conducted at the start of 2018 suggested their utility on the warp-weighted loom. Taking this idea further, I reasoned that it would be appropriate to use this shape on the loom to determine its suitability for this function.

2.     As quick as it may be to spin yarn with a wheel, it is not as fast to spin with a spindle—at least for me. Because spinning wheels are a more modern invention, I had to use more simple methods. Instead of spinning a thousand yards or more of yarn with a spindle, I opted to use another proxy. My lovely Golding ring spindle is approximately the same size, shape, and mass as a Danebury spindle whorl. Then came the question of material choice. I settled on flax for two reasons: 1) it hasn’t been investigated as a fiber source for Iron Age Britain—because wool is often the only fiber sufficiently discussed for textile use during this period—and was likely still important in select cases during the Iron Age; and 2) it can tolerate approximately double the amount of tension that wool can support before breaking—this is important if I’m investigating new territory with heavy loomweights. So I spent about 15 hours spinning flax with my proxy spindle. I found a comparable unbleached, wet spun flax yarn of similar gauge online and used that as my test yarn.

3.     It is uncertain whether a long-handled comb was used during the process of weaving, and we have next-to-no evidence of a weaving sword during this time. However, any weaver will tell you that you need some way of packing the weft. For my Stage 1, I only used a weaving sword to pack the weft, but I did commission an antler comb that was modeled from a Danebury example for the Stage 2 experiment.

4.     Needles, again, are a complicated matter. Needles can be used for a variety of tasks and there is no clear evidence of their use at Danebury for textile production. Some look like they would be great for sewing or darning, while others are certainly not. I did use a steel needle that was of a similar length and width (a typical darning needle) to attach the warp threads to the loom and for experimenting with hemming techniques.

Here is one of the air dried clay loomweights that I modeled from an Iron Age equivalent.

With that, I had my basic setup for the Stage 1 part of my experiment. I am gliding over some of the reasons I chose what I did and why I think my decisions are valid for the point of my experiment, which is this: Can flax yarns spun with Danebury spindles be tensioned with heavy loomweights without breaking?

Continued in Part 2!