It's always nice to find out things... especially when they are connected to medieval textile tools. So when, a while ago, a colleague handed me a reproduction of a weaving knife (or beater, however you opt to call it) from Viking Age Dublin to try out, I was more than happy to do so. It fit in perfectly with my plans to do the tablet-weaving shenanigans... and so I sat down and wove and gave it a try.

Well. Before, I was firmly in the camp of "I just use my finger to beat in the weft, and that's good enough for me". Let me tell you, though: I am not ever going back to that. Well, maybe for a narrow, simple band, where weaving is fast and I'm handling the tablets with one hand and the shuttle with the other and it's more like just speed down the warp, and that's it. But for things like the twill patterning, or wide bands? Aaah. It does make such a difference - and it actually makes a huge difference, for me, to pressing the weft in with some other tool.

In the past, I had tried to use the shuttle edge, or a special small shuttle with a "beating sting" attached to it (which I have mislaid and cannot currently find, which tells you already how much I use it...), or a ruler, or a bone folding tool. None of them cut the mustard for me - they were one more thing to handle, and my finger did not need picking up and setting down in contrast to these all, and made about the same job.

The difference? Size and shape of the new thingie. This is how it looks:



You will note the very broad, very curved blade with the curious indentation at the very tip - things that my colleague found slightly weird. There is a second find from Viking Age Dublin with this shape, so one might surmise that it's not a fluke, but that there is some reason behind it. I had no clue about this possible reason, only a guess regarding the curved blade: It might allow to press the weft in very precisely yet gently, using a rolling motion of the curve across the fell.

Well. That guess certainly proved correct - the tool's form is just perfect to roll it across the width of the band, making sure all the fell is pressed back firmly and evenly. As there is only a smallish contact point, it is easy to apply a little more pressure where necessary and a little less where the warp is softer. Calling this thing "beater" does it no justice at all, by the way - movement and action are all nice and soft and gentle, as befits a delicate tablet weave. With the warp under just enough tension to work well, but not more, beating in the shed would not be a good idea, as it might disorder the tablets. It also puts more stress than necessary on the warp, as well as on the tensioning method - so the curved blade is a total boon.

[caption id="attachment_4054" align="alignnone" width="640"] Gently pressing in the weft.


For this, a smaller tool would actually work for me - but I have also found it wonderfully useful for correcting mistakes.

When I need to go back a pick, I loosen up the last weft, pull it towards the tablets, and then I open the shed it runs through with my fingers. I then usually stuck something in there - my finger (which binds one hand, so not very good for longer or more complex corrections), some stick, or whatever (a bone folder, for instance). The purpose of this is to keep the old shed open to individually turn each tablet back the way it came, re-establishing the position before the mistake. Obviously, it's a) helpful if you see how the threads run over your tool, and b) important that it does not slip out too early. Both of these were, hm, let's call it sub-optimal with a stick or a bone folder.

With the almost 30 cm of total length, and the relatively long blade part, the tool easily accommodates my warp for these corrections. The back of the blade is wide enough to stretch the shed apart nicely so I can see what is happening, and the total length is just long enough to either set the handle's end on my knee or hold it between my legs, making sure it stays put.

[caption id="attachment_4051" align="alignnone" width="640"] Fixing a mistake by weaving back - the crossings from the previous turn are carried towards the tablets.


Most often (fortunately!) the mistakes are not so bad that I need to go back for the whole pick - they are just one or two tablets that accidentally turned into the wrong direction. Or sometimes, I'm not sure whether the edge of a pattern section is where it is supposed to be. In these cases, I pick out the tablet cord in question close to the fell and follow it back to the stack of tablets. For this, something more pointy and accurate than a finger is very helpful. Bone folders will work, as will needles... but, guess what works even better?

Yes. The pointy tip of the tool. With the indentation making the tip of the blade effectively a lot narrower than the curve would dictate, it is just perfect to pick out the offending place, stick it in and follow the yellow brick road, excuse me, the tablet cord in question back to its origin.

[caption id="attachment_4053" align="alignnone" width="640"] Catching the questionable pattern section edge to check whether the tablets stand correctly.


It works much, much better than the bone folder. Or a ruler. Or my finger.

It's also not a bad thing that this tool is totally beautifully made, and very smooth, and beautifully decorated with carvings, and made from service tree wood, which is my absolute favourite wood ever.

So worth the extra step of picking up the tool! And in the future, my fingers will not see that fell again... because I'm not going back, nothankyouverymuchindeed.

Things going on?

Prep for the Textile Forum - writing lists, making plans, figuring out stuff, and generally getting really excited.

Also... more weaving. Currently I'm trying to get merging lines down pat, so the idea was to weave a tree (inspired by one of the Arlon motifs) and follow that with a bee (which is perfect for trying a few other things that serve as design elements).

Well. Merging lines. That means you have one line coming from the left, and one from the right, and the main turning direction of these lines is not the same, so one basically has to eat the other one. In theory, I got this. In practice, I thought I had it at one point... and then I wanted to repeat it and I got horribly, horribly muddled up. Maybe it was too many things at once...



...because I wanted to try starting multiple diamonds at the same time on the free side of the band. Well. That did not go as planned.

You can see in the picture that things got really, seriously wonky. Ah well. I'll make the tree a little taller... and try again. (In theory, I know exactly what I have to do. In practice, somehow, things sort of get out of hand occasionally...)

I hope you're not bored to death yet with my posting of tablet-woven animals but, well, you guessed it - I've made another one, and I have nothing better to do than share it here:



Starting this was a little bit more difficult than I had anticipated, if only because I was not able to tell left from right properly, and had originally planned something that does not work - so it took a few tries before I was able to set off.

I also learned that there are some things that have to be done in sets of four, not sets of two picks, which is why there is some wonkyness at the base of the neck (the bit at the bottom would have been avoidable, but I was too lazy to un-weave two picks).

Despite all of this, I managed to get out at the end with much less muddle than I had at the end of the horse, so things are getting better - and with the new rule to add to the set, there should be even less muddle next time.

And now I'm wondering what I should weave next... though I probably should do some other things first for a change. These critters eat a lot of time!

So after the success of the tiny wonky animal on the narrow band, obviously there was no other choice than trying out the new system on the wide band.

And, well, what shall I say? I was mostly successful, and I can now give you this:



Woven on a 42 tablet wide band, with no pattern drawing. Not even a sketch.

Actually, I did make several sketches, but in the end, I followed none of them. I was sitting there, weaving, and that certain part of my brain went "I think I should do this here" or "wonder if I can't give that horse a mane" or "this doesn't look wide enough yet".

Obviously, it did not go all smoothly all the time - I did botch up at a few different occasions, once so badly I actually had to un-weave two picks.  The tablets were a bit out of sync after one point (damn you, directional change at odd weft counts!) and that did make going on, and matching lines together, a lot harder than it should have been. So in the end, I had some parts that were running smoothly and some wonky places between them where I sort of just fudged it. The smooth parts were partly in sync and partly not, so the wonkyness went both ways, and it was rather a relief to find that the head blob had finally gotten long enough, as had the front legs, so that I could just turn everything pink again and call it a day. Or a horse.

However, being able to go on and actually finish the critter proves, to me, that the system does work very well - because the last times I got out of sync in a similar way, or even less direly, everything got so confusing that I basically gave up and restarted with a clean slate. This time, it was possible to muddle my way through, and the next critter will surely be less wonky again. I've also realised that there are some more hints (that can probably be turned into rules) to prevent things like this from happening again.

So. Achievement unlocked: Twill patterns without pattern.

And what kind of critter will I weave now?

For ages and ages, I have dreamt of this. Well, admittedly, my dreams were a little bit more spectacular, but still - I've done it.

I have woven an animal.



It's not looking very spectacular, and it's not a very complicated animal (which would actually be rather difficult to achive on only 12 tablets), but it is one. And I was utterly, utterly happy to have managed this...

... not because it is the most beautiful little animal that was ever woven (far from it!). But because it means that I have finally cracked that nut and figured out the system to weave patterns freehand on a twill background.

This was done without any pattern or sketch, starting from a base with all diagonals running into the same direction, using the newly-rediscovered system to change diagonal slants. It's a humble, ugly little animal that looks a bit like it is vomiting, but I'm very, very proud of it.

So... now I will progress. To more, bigger, more complex animals that might actually resemble real living beings so much that you might be able to recognise what it's supposed to be...

I haven't progressed to the wide playband as quickly as planned - I got a little held up figuring out some more things with the narrow one, as there are (as always!) one or two little more details that need some more verification.

Making the direction change at the end of the pattern section, where the block of tablets turning together to form the different-coloured stripe serves as orientation, though, works very well. It also has the nice side effect of keeping appearances neater - probably because the slightly increased wonkyness of the fabric is counteracted by the firmer structure of the several-turns-in-one-direction band.

[caption id="attachment_4019" align="alignnone" width="640"] The two light triangles on the left are woven with the directional change at the start of the pattern, the ones to the right with the change at the end.


Or it was sloppier weaving on my part in the section before... or it was all the mistakes made. Hm. Whatever the reason - whether fewer mistakes or better weaving - it does look nicer now, and it is so, so much easier to work.

After figuring this out, I played around some more... and obviously I ran into some new, exciting (well, more or less exciting) problems. One of them was quickly solved, thanks to the input of the Most Patient Husband of Them All - I had accidentally done a second direction change after five picks instead of six. Odd numbers are, in this case, not a good idea, and five is right out. Aaaah. So you don't only have to tell left from right, count up to two (or multiples of two), and tell dark from light, you also have to tell odd from even.

That, though, is something I should be able to do. In fact, it's not hard at all - you just need to look at whether your shuttle is on one side or the other side of the band, and then remember that there is a "good side" and a "bad side" (which has no cookies, but horrible things in store for you) and only do direction change shenanigans when the shuttle is on the good side.

So next time I sit down at the band, there will be some Y shapes... and hopefully my current theory will work out.

Also up for getting solved? How to pour all these little rules and theories and things to know and keep track of into a format that will let others follow me when I talk about all this patterning craziness at the European Textile Forum...

Sometimes, you puzzle over something and puzzle over it and try to find a good way to keep track of what is happening and how to count to two and keep right from left. At least that happened to me in the occasionally ongoing tablet weaving project, where I'm still trying to figure out how the weavers back then managed to weave those intricate patterns without swathes of graph paper.

So. I finally sat down once more and tried to count to two, keep track of those double-turns, start with double-turning the correct tablet, and keep left from right. Like so many times before, I was not able to do it. Again.

Short explanation of these semi-coherent ramblings: When tablet-weaving patterns on a twill base, you need to make sure your base slants in the correct direction. If your slant is wrong, the pattern comes out wonky.

This means that if you want to weave a triangle, you have to change base direction somewhere inside it, or the second side of the triangle will look really, really shitty.

There are two ways to change that slant, one more modern and one historical. The modern one is easy - you weave to the middle of your triangle (or triangular section of a larger pattern or motif), then swap the turn direction of all your tablets, done. This works, but gives you lots of float lines in the middle of the pattern.
For the historical one, you only change turn direction of every second tablet, and you do that sequentially. For this, you either let a tablet stand idle or double-turn it, and one of my problems was knowing which one to pick to start the sequence.

Sigh.

And then, suddenly, I was wondering about the rule to know which tablet to start with. Because there must have been a rule, depending on, probably, slant direction and number of tablets in the pattern part. And then I was thinking that this would mean thinking ahead and figuring out beforehand how many tablets would be involved in that pattern part... and that would mean even more counting, and more keeping track of things, so more brain work. Which is, according to my theory, something you want to avoid.

Not because I'm thinking folks back then were stupid. No, because you want to have as much brain capacity free for keeping track of the things you have to keep track of, and on making your pattern look nice and shiny. So my rule for the tablet weaving system is that everything that can be made a rule and shoved into the backbrain will be made a rule.

Such as the direction the tablets in their packs are turning (always towards each other) and the direction the stripes are running when I turn the whole pack into one direction. That does make things easy because I only need to look and I know what my base pattern slant is, for instance.

As I was wondering about these possible, yet still-unknown-to-me rules for the double-turn tablet, I suddenly thought about trying to do the direction change not at the start of the triangle... but at the end. After I have decided, spur-of-the-moment, that the thing is large enough now and I am going to close it.

Well. What can I say except "D'oh" and "why didn't I think of this much earlier". Turns out that this makes finding the tablets that need to double-turn so, so much easier - it basically reduces all the problem to "look at your tablets, check if they stand correctly, and if not, make it so". Which is totally fitting in with how I figure that a lot of the original weaving was done, and makes things much more relaxed and much more logical.

So now I'll be going for a little more practice with this stuff, first on my 12-tablet trial warp... and then on the larger, 42-tablet trial warp. Because sometimes you just need some room to play, right?