How to do tablet weaving - Interactive pattern designer - Patterns - Theory
I am not an expert on tablet weaving, or indeed any type of weaving. But the idea of tablet weaving fascinated me because I found it hard to understand how the patterns were produced, and why. So I investigated, and this website is the result! This page gives the theory behind tablet weaving. You don't need to read this page to do the craft, but it might interest you.
First, you need to understand how "ordinary" weaving works, before seeing how tablet weaving is different.
Usually weaving is done on a loom. This is just a machine to make the job easier, with different types, but the basic technique is the same for all. There are a number of threads running in the same direction (it could be up and down, or side to side - depends on the loom). These are called the warp threads. The number of them determines the width of the cloth produced. Then there is a single thread which goes back and forth across the warps to make the cloth. This is called the weft. Weavers talk about warp and weft quite a bit, so it's a good idea to remember which is which. If you tend to get them muddled up, then the weft is is the thread which goes from "weft" to "wight" and back again. The warp threads are held taut, which (nearly) rhymes with warp.
There are two basic types of weaving with looms, tabby weaving and twill weaving.
Tabby weaving is also known as plain weave, linen weave or taffeta weave. It is the simplest type of weave. The weft goes under a warp, over a warp, under a warp, over a warp, and so on. In the next row, the weft goes over the warps it went under last row, and under the warps it went over last time. Every two rows are the same. It produces a pattern like this:
Perhaps you are thinking "Which is weft and which is warp?" The pattern is symmetrical for both weft and warp, so it doesn't matter! Traditionally, the warp threads are shown as going up and down, so that would be the red threads, and the weft thread is blue. But turn the pattern round, and it could be the opposite.
The diagram above shows the threads spaced out, but usually, of course, they are pressed close together.
Twill weaving still has warps and a weft. But this time, the weft goes under 2 warp threads, over 2, under 2, and so on. For the second row it does not do the opposite. Instead, it goes over 1 thread, under 2, over 2, under 2, and so on. So the unders and overs are staggered by 1 each time. After 4 rows, it gets back to the start again.
The spaced out version is rather mesmerising! Here is the pressed together version. It shows the distinctive diagonal stripes of twill.
This style of twill is also symmetrical as far as the warp and weft. If you turn it round, you might get the stripes going the other way! But you see the same amount of warp and weft either way.
For completeness, I must add that there are variants of twill, with the weft going over and under more than 2 warps. There is also something called satin weave which has more still. I will ignore these. Look them up on Wikipedia if you're interested!
So why is tablet weaving different?
The most important difference is that in tablet weaving, you don't see the weft at all. The pattern is made entirely by the warps. Perhaps this happens in loom weaving as well, but I can't find anything about it. For tabby weaving and twill, the warps and weft are uder equal tension, so both types of thread appear equally. In tablet weaving, the warps appear in the pattern when the weft goes under them, of course. But when the weft goes over the warps, the warps on the rows either side slide over to fill the gap. All you get are warps appearing in the final cloth. I suspect this is because the warps are under less tension, so they don't pull up the weft to be visible.
Tablet weaving often produces stripes, and so looks the same as twill. But it is not. In twill, there are stripes of two colours, one being the colour of the warps, and the other the weft colour. In tablet weaving, you don't see the weft colour. The stripes are produced by different warp colours.
I have a theory (which may be entirely wrong) about why tablet weaving happened. All kinds of weaving depend on getting the weft to go over some warp threads and under others. If you have a way to automatically lift every other thread, for example, you can do tabby weaving a lot faster. Lift the odd threads, push the weft through, lift the even threads, push the weft back again. Looms can do this for you. Indeed, I had a toy loom as a child which made tabby weaving. However, it couldn't do twill weaving (except by labouriously putting the weft over and under the warps manually). Serious looms can do twill as well, by having complicated mechanisms for lifting different combinations of threads, but I don't think these were available in the past. However, in the past, someone may have wanted to produce a twill, and thought that if you threaded the warps through the holes in the corner of a square tablet, it would lift two of the warps (top holes) and leave the the bottom two alone. That would be row 1, and then you twist the tablet to get a different two warps to lift, for row 2. Then perhaps you could get a twill pattern. Unfortunately, tablet weaving doesn't work like that! You need to keep twisting the tablet, and that alters the order of the warp threads. In fact, each group of four warp threads gets more and more twisted if you twist the same way all the time. Definitely not twill, which has the warps staying in the same order and staying in a straight line through the cloth, not at all twisted. However, the tablet weaving does produce stripes (with the correct colours threaded on the tablet) so that was a result, I suppose. This twisting may answer another question that worried me - why does everyone insist that tablet weaving is only used for narrow pieces of cloth, like belts? Perhaps this twisting produces structural weakness in the cloth. It may be stronger along its length (as twisting makes threads stronger) but weaker, or more uneven across.
You may have noticed that I have not produced a diagram of what threads look like in tablet weaving. I am tempted to say "That's because the're a right mess!" The pattern on the surface of the cloth can be shown, and my tablet weaving designer page does that (it shows the underneath as well), but if you tried to show all the threads, including the weft, by spacing them out, it would look very confusing. The simple loom weaving above just has 2 levels of threads - what you see and what's underneath at that point. Tablet weaving has 3 - the top 2 threads of the tablet, the weft, and the bottom two threads. Then the tablet threads move from the left side of the tablet to the right and back again. Click here for diagrams of the tablets themselves and the threads going through the tablets, and the final pattern. But I blench at producing a complete diagram of what's going on!
There are various actions to produce different patterns. The examples below start with this threading of the tablets - all threaded from front to back (unflipped):
The most important part of tablet weaving is turning the tablets. You can turn them clockwise or anti-clockwise. With the correct choice of threads in each tablet, turning all the tablets the same way tends to produce stripes:
Turning the tablets the other way produces stripes going the other way:
So turning them one way, then the other, produces zigzags:
It also reduces twisting of threads (see above).
There is no reason why you can't, for example, twist half the tablets one way and the other half, the other way. A bit fiddly, but possible. That would produce stripes leaning one way for half the tablets, and the other way for the other half - that is - chevrons. However, there is a better way to produce chevrons.
The tablets don't lie at right angles to the weaving, instead they lie flat, along the line of the threads. That means that, starting from the weaving, some of the threads have a longer length before they reach the tablet than others. The threads are all really the same length, so if there is a longer bit before the tablet, then there is a shorter bit afterwards. You have to decide how you thread the tablets, so as make the longer/shorter bits behave properly. If you thread all the tablets from front to back, then they will all be "unflipped". (Which is "front" and which is "back" doesn't really matter, but decide for yourself which is front, and stick to it!) See How to do tablet weaving for pictures showing this.
It is quite possible to thread some of the tablets front to back, and some back to front. Or, during weaving, you can "flip" the tablets. That means turn them not as in a normal turn, but forwards (or backwards), so they lie flat the other way, and all the "short" bits beome "long" bits and vice versa. If you do this, it has the odd effect that a clockwise turn of that tablet seems to look like an anti-clockwise turn. So you get stripes going the opposite way. If you set up the pattern above, then flip the first two tablets, and turn all the tablets the same way, you get chevrons (without having to turn half the tablets differently!)
And if you turn the tablets one way for a few rows, then the other way, you get diamonds!
I don't find the stripes produced by the turning hard to understand. The colour of the stripe is staggered throughout the tablets, and turning the tablets makes the stripe colour appear on one side of different tablets and disappear on the other. You've got to get the pattern right at the start, but then it all seems natural, by the mechanism of the turning.
However, the flipping did confuse me for some time. But I worked it out as follows.
Let's start by looking at unflipped tablets. The diagram below is following just one tablet as it is turned 5 times clockwise. I've put a different coloured thread in each hole to make it clear. Remember, this is a single tablet, not 5 tablets! It shows how each turn makes each colour appear in a different place.
You will see that the red always follows the yellow (that is - it appears in the place that yellow was last time).
Now, let's flip the tablet. This makes the colours swap over as follows. (Whether this is the top two / bottom two, or left two / right two doesn't really matter for the theory!)
Now to see what happens when this new arrangement of colours get rotated clockwise:
The red now follows the blue (that is - it appears in the place that blue was last time). And if you look at each colour, it is behaving in the opposite way to the unflipped clockwise version. In fact, it is behaving as if it was an anti-clockwise turn! This can be very confusing. If you look at one colour by itself, say red, it is travelling through the tablet in a clockwise direction. And of course the colours all start in different places. But the order of colours has changed, which makes all the difference to the pattern. This is all very mathematical, connected with symmetry, and rotation, and reflection. By the way, you don't have to understand it! Just trust that flipping makes a clockwise turn look like an anti-clockwise turn.
You might wonder why we need to have two directions of turn and flipping, since flipping merely makes one direction of turn look like another. I think that it's partly for ease. If I want chevrons, it's a lot easier to set up the pattern, then flip half the tablets, then turn all tablets the same way without further thought. If you didn't flip, then you'd have to turn half the tablets one way and the other half the other way for every single row. Tedious! It could be that turning and flipping are used for more complicated patterns - I'm not sure about that.
© Jo Edkins 2019 - Return to Crafts index - Return to Tablet Weaving index