Posted by Steve Price on 06-26-2003 09:02 AM:

Symmetric and Asymmetric Knots

Hi People

One interesting thing about Turkmen pile decorated tent bands is that they are almost always done with symmetric knots, although only the Saryk and Yomud routinely used symmetric knots in other pile weavings. That fact is curious enough to have been the subject of many comments in print.

I've found only three published tent bands with asymmetric knotting. While my search certainly hasn't been exhaustive, it includes a lot of examples. One curious thing about the three exceptions that I found is that they all appear in the same book (Cassin and Hoffmeister's Tent Band /Tent Bag) and are the only tent bands in it.

Does anyone have a sensible explanation for the nearly exclusive use of symmetric knots on tent bands when the weavers (presumably the same weavers) used asymmetric knots for every other pile piece they made? Or are there lots of other asymmetrically knotted tent bands around that you know about but I don't? Or are nearly all tent bands actually Saryk or Yomud work?


Steve Price

Posted by Steve Price on 06-27-2003 09:59 AM:

Hi People,

The two pile decorated tentbands in Tent Band/ Tent Bag (numbers 39 and 40) were erroneously described in the text as being asymmetrically knotted. They are both knotted symmetrically. The source of this information is Peter Hoffmeister, who owns them.

The all pile tentband in the book (number 1) is asymmetrically knotted, as described in that book and in Wie Blumen in der Wuste, where it is plate 80.

It appears that there are no known tentbands made with pile decoration on a flatwoven ground in which the knotting is asymmetric. At least one of the rare all pile tentbands is asymmetrically knotted.


Steve Price

Posted by R. John Howe on 06-27-2003 05:29 PM:

Steve -

I am hoping that we will be able lure Marla into this discussion, but in a side conversation she has given one reason why symmetric knots might generally be used in these mixed technique tent bands.

She wrote in part:

"...symmetrical knots are far more secure and practical when combined with the warp-faced weave."


R. John Howe

Posted by Vincent Keers on 06-27-2003 06:59 PM:

A wake up call,

Well this really messes things up.
Why warp-faced? Because we see warps at the back?

What about brocading? 3-1-3

And what about a soumack construction like when forgetting to clip the loops but getting on with the wrapping in an 1-0-1-0-1-0-1 (3 warp construction) etc. so no piles.

And because 1 weft is inserted, an a-symmetrical knot would be dumb. Because this results in one row of knots the "left" warp is up' And the next row of knots the "right" warp is up. So sometimes the loops around the warps are on the lower warp and next row they are on the higher warp?

I've been reading your book. But this all really makes my world very chaotic.
Why calling these bands warp-faced?
And the spoonbags weft faced?
Because more wefts?

Best regards,

Posted by Christoph Huber on 06-28-2003 06:04 PM:

Dear Vincent

You see the warps on the back as well as on the front (only where there is no pile of course.)
And there is no problem with warps being 'up' since the knots are wrapped around alternate warps so that both of the warps used for any knot are on the same level - no matter how many wefts are involved.
I suppose one of the problem with asymmetrical knots would be that they often wouldn't stay properly in place until the wefts are inserted, since the ratio between the thickness of the pile-threads and the space between the two warps is rather different on an ak yup compared to a "normal" rug.

Best regards,

Posted by Vincent Keers on 06-28-2003 07:50 PM:

Dear Christopher,

Thanks for helping me out. I think..
One weft does split the two sets. But I think i see now that the wefts are sinious. But if I do this on my "do it yourself kit" or "give it a try" kit, things look different. That's because i was hammering the wefts down to compact.

And...I think a warp-faced structure is a warp faced structure. This means to me, that the warps have to show something, something to do with the design. So the face of the weave is domminated because of the warps. Not because the weaver didn't hammer down the wefts in a weft-faced structure.

The 3warp construction is strange in rugworld but brocading does feel like the same rythme and brocading is used by these weavers.

So can we say that this 3warp construction comes from brocading? 3 up, 3 down etc. or 3 up, 1 down, 3 up, 1 down etc because this gives offset brocading.

The warps being closely packed, can't be the only reason for this 3warp construction.
(This is what Marla said, didn't she?)

Best regards,

Posted by Vincent Keers on 06-28-2003 08:53 PM:

Here it is:

The first, 1 in a lifetime, do it yourself kit.

Get yourself such a great tool. Now you can weave your own socks, underwear or whatever. Cold ears? No problem.

You see?

And, while we are at it.
It seems that, if the wefts are fixed at the sides, the result is that the warps show up in between.

So, my conclusion from this is, that those romantic pictures from weavers making them outdoors, could be ....well....a bit strange.
44 feet, the wefts not fixed at the sides?
Brocading, weaving, knotting? No sides fixed?

Posted by Christoph Huber on 06-29-2003 05:32 AM:

Dear Vincent

To my understanding - although never having it tried out so far - the warps really HAVE TO be much nearer. In fact, you have in a tent band twice as much warps as would be needed for a rug of this width.
The result is that the WARPS are sinuous while the straight weft remain invisible.
On your model you could, just to get an idea of what happens with warp and weft in a tent band, insert wefts made of wire. So these "wefts" would stay straight without the need of keeping the warps too near together.
I don't think that you have to bother too much with the knotting itself. It is exactly the same as on a normal rug - with the sole difference that the shed was open while knotting. Look at the second picture on Marla's pages - my whole understanding is based on this drawing!

Have fun experimenting!

Posted by Richard Tomlinson on 06-29-2003 06:51 AM:

vincent, is that a loom or a mousetrap?

Posted by Vincent Keers on 06-29-2003 07:07 AM:


Dear Richard,

For some this may look like a mousetrap.
But you can order one.
So you can see what you're talking about.
Think this would help a lot.
Like inviting sheep and goats on the next ICOC,
give all invited this "Do it yourself" kit.

Best regards,

Posted by R. John Howe on 06-29-2003 07:20 AM:

Hi Richard -

I know you're just kidding Vincent (and who better to kid?) but the sort of thing he is doing here is very, very important.

Lots of things that are hard to describe clearly, become transparent when one tries to do them.

The old training example we used to use was to describe to another person (without examples or demonstration) how to tie one's shoe laces. Or pretend I'm a Martian and teach me, with words only, how to smoke.

Very hard to do with just language but much easier when actually tried.

When I was at the Maryland Sheep and Wool Festival recently, I saw that they were selling little looms like this on which folks can experiment with particular weaves.

A very good idea and practice.

The Apostle Paul says somewhere "Who is this who darkeneth countenance with words?" Unfortunately, it's all of us most of the time.


R. John Howe

Posted by R. John Howe on 06-29-2003 04:54 PM:

Dear folks -

Vincent is right that you can buy small looms to demonstrate to yourself (with reference to a book like Marla's) how particular structures are woven.

Here without attempting to break over our non-commercial stance is one source that offers them.

Notice that the first one seems pricey but look down a bit and there's one for about $6.50.

Vincent's idea that perhaps they should be given out to registrants at some rug conference is, to me, a very good one.

Or if that couldn't be afforded, maybe someone might conduct a session the admission price of which would be to purchase such a loom and the substance of which would be to lead folks through the weaving of some standard weaves.

Would let us "walk the walk" a little, rather than only "talking the talk."


R. John Howe

Posted by Vincent Keers on 06-29-2003 07:46 PM:

Hi John,

But this is a cardboard box.
This is something people can make themself.
1 box, some cotton, some wool.
It's very lightweight so you can put it on your lap.
You can stick in needles anywhere you like.
You can put the rest of the gear, in the box.

And when it's done. Cut everything out and start over again. And if the box gets to porous, make a new one. This one is 20 minutes work. Next one is 15 minutes.

So next ICOC. Working class: Make your own loom, because making a loom is the real start.
Next thing is: Make a mousepad on your own loom.

Best regards,

Posted by R. John Howe on 06-30-2003 05:49 AM:

Vincent -

I think you are right that building your own mini-loom is an even better place to start, but I think you under-describe the materials needed (eg., I see some bits of wood in yours) and I think you underestimate the initial difficulty of this task for many who have never woven or repaired textiles.

Twenty minutes for you, maybe, but I think a bit longer for most of the rest of us the first time without clear guidance.

But you are right that this is something worth doing. Very clarifying.


R. John Howe

Posted by Vincent Keers on 06-30-2003 11:08 AM:

Hi John,

Yes, I'm very handy.
No wood. Only the lead-pencil as wedge for splitting the warps in 2 sets.
2 Cardboard plates with some sellotape for getting all the warps a bit higher.
And it's very, very basic. Transportable so I can leave for better meadow-land in a wink.

Dear Christopher,

Tentbands aren't made with wire as wefts.
As you can see in the third image, at the left side, I fixed the wefts at the sides because i forced them through the cardboard sides (This can be seen in the second image).
This makes the warps more sinuous and the wefts more stiff. (The warps show up in between the wefts)
I still do not get the picture of a weaver making a band with wefts that need to be stiff without getting a band that gets more narrow and narrow as work's in progress, if the sides aren't fixed.
Even if the warps are very close, this only would make the process more frustrating because than the warps will not allow the sinuousity we need.

So I think, most of these bands are factory made and that's why it's hard to give them a tribal attribution?

But, I'll give it another try.
Because everyday in this life I learn something new.

Best regards,

Posted by Vincent Keers on 07-03-2003 05:35 PM:

Hi all,

So what did I learn?
I will never weave a tentband..that's sure.

First I'll show you the warp/weft structure

9 cotton warps per cm. 25 per inch
The weft shots are stiff, so this forces the warps to go sinuous.

Here's the rug!

Yes, you may burst out in a hysterical laughter.
But Arnold Schwarztenegger said: "Me do like it....."

First, this is symmetrical on the open shed, so 3 warps wide. Offset, so skipping one warp in between the knots. Next line, same thing, only on the second open set of warps.

I was disapointed, because the fabric began to shrink under my hands as you can see. But I had to keep tension on the wefts.
I didn't like what I saw, and again, I think this can't be done if the wefts aren't fixed at the sides.

So I changed the knots.
Here it is all symmetrical on 2 warps and all in offset. The problem with this is.....

because the weft is stiff, the warps will behave depressed. One weft hides half of the symmetrical knot. Marla shows a Senneh structure on her site that does the same.
This is why, a-symmetrical knots do not give a strong structure if on weft is inserted. It would make the fabric floppy. The symmetrical knot can be used as a real knot, because both warps are encircled. The strengthens the fabric. A-symmetrical doesn't have this "warp-binding" quality.

In short:
Tentbands could have been made in symmetrical structures. No problem. Because the one weft structure and the offset symmetrical knotting keeps the fabric flexible and soft.
So the open shed, 3warp, construction is because the weavers had a fixed program.
This program was brocading.

This concludes my private investigation in the open shed/3warp construction.

Any questions?


Posted by Steve Price on 07-04-2003 10:09 AM:

Hi All

I hope everyone noticed what's just transpired here. Vincent has done some hands-on experimenting and generated a hypothesis for why tentbands are woven with the peculiar structure they have. Specifically, they are done with symmetric knots, over alternate warps instead of over consecutive warps. There is only one known exception, an all-pile tentband done with asymmetric knots.

The use of alternate warps for the knots makes the fabric very flexible, a property much more important for a tentband than for the more common kinds of Turkmen textiles. It would be interesting to know whether the one all-pile asymmetrically knotted band is less flexible than the others. The use of symmetric knots makes this structure much more stable than asymmetric knots would.

I think this is a significant advance in thinking, and at least provides a plausible (albeit not rigorously proven) explanation for this very odd structure. If anyone had a rational explanation before this, I missed it.


Steve Price

Posted by Vincent Keers on 07-04-2003 08:22 PM:

All are keyboardless

Hi Steve,

But what's most important is the one weft structure. Because 3warp knots on a 2 wefts stucture isn't flexible as a 1 weft structure.
And this 3warp knots seem to be very conveniant with 1 weft. Because the weft goes
under/over/under the warp.
loop-/skip-/loop the 3warp knot does
and this fixes the piles.
And the next weft does strengthen it because it does over/under/over and pulls the center warp up against the bottoms from the piles.

In short:

If all the symmetrical knots are in offset and one weft, the fabric is flexible as it can be.

Asymmetrical and 1 weft can be a problem.
But I dare not to describe it.
Think the result is: Piles sticking out all directions. East, west, north, south.
Maybe I'll give it a try.

I expect 2 wefts in the all asymmetrical pile band. So less flexible. And if I'm wrong, I'll start all over again.

Did you have a good look at your spoonbag? Does it have 1 or 2 fine wefts? The weavers fool us by showing red/brown and white woolen wefts in the plainweave parts. But the moment they start knotting those tiny knots, they change to dark, very fine wefts. One straight and one sinuous.

And in the end something pops up in my brains.
While I was knotting, and knotting and doing things a bit different. Changing weft ease, changing knots and designing a Tippy in a river? "POP" there it was. It's more easy to change your technics, than to change the design. Nobody in my tribe is mad at me because I was experimenting with the technics. But if I was supposed to design a boat on the river? What would be the result?

Best regards,

Posted by Chuck Wagner on 07-05-2003 10:49 PM:

Hi Vincent,

That's a nice bit of work, which I'm still digesting. A note on the
40 foot band looms: Those things are often Non-Stationary;
the warps are staked out over a long distance and the loom is slid
forward as the work progresses. So there isn't a long distance
between the weaving work and the tension control. There's a
photo of a tripod loom in one of my books somewhere that I
have scanned. When I FIND it I'll send it to Steve to
add to this post. This photo illustrates this feature very well.

That doesn't negate your point, but it indicates that the effect
you describe may be less severe than one might expect.


Chuck Wagner

Posted by Vincent Keers on 07-06-2003 06:15 PM:

Hi Chuck,

Hope you can find the picture.
I've seen some pictures. But the thaught, these weavers made piled tentbands, never crossed my mind. Neither did I see it.

It's because these 3warp knots have the same binding quality as the 2warp symmetric knot, it's very hard not to tie the knots to strong, like I did in my test-rug. So the problem isn't in the straight weft, it's more in the knot. And because the knots pull the 2 (from 3) outer warps together, the test began to shrink. So the wefts got shorter too.
When I fixed the wefts at the sides, the problem was solved because than the wefts helped me holding the warps in their place and this prevented me from tightening the knots to much.

So, I'm not convinced these partly-piled tentbands could have been made outdoors.
But, I'm not a weaver.
And maybe those non-stationary looms had a sort of side-beems that fixed the 3warp/weft tension problem?

Best regards,

Posted by Patrick Weiler on 07-06-2003 11:59 PM:



Given the nearly universal structure, could these tent bands have been commercially made in an urban setting for a variety of wealthy Turkmen consumers?
We know that some suzani were at least partially commercially made (the designs may have been drawn on the ground structure by professionals).
Wealthy Turkmen may have commissioned their "upscale" tentbands, too.

Patrick Weiler

Posted by Chuck Wagner on 07-07-2003 10:53 AM:

Found it

Hi Vincent,

Here's the picture I was referring to before; it's NOT a Turkoman
setting, but rather, a nomadic family in Iran. But the loom setup
is the point.

First a note about photo credit: I can't remember which book I got
this from, and a lot of my books are now back in the US. My most
likely guess is Jim Opie's Tribal Rugs, but it may have also come
from a book my wife bought in Tehran.

Now, on to looms. As you can see, this tripod rig is designed to move
ahead as the weaving progresses. The heddle stick is on a
hinged tether that can be rotated toward (or away from,) the
weaver, or tied securely to the frame.

So, tension control is probably better than one might expect given
the rustic setting. But I have my doubts about achieving the level
of consistency seen in the bands in the Salon, or the ones I own
for that matter.

As Pat hinted, they look TOO good to be rustic. And it makes sense
that such bands would have a special status. They're too fancy
and too delicate to be taken down and put up frequently without
getting damaged along the way. These bands are wrapped around
the frame of the yurt, facing inward, as a decorative/utilitarian
item. The frame moves in windy conditions and abrades whatever
is in contact with it.

There are LOTS of flatwoven tentbands around, but relatively
few piled ones. I suspect they're wedding gifts, or something
like that, and had a special set of circumstances associated
with their use. So, these may well have been made by a pile
tent band specialist in a semi-commercial mode, or by weavers
making their own goods but in a settlement environment.


Chuck Wagner

Posted by Steve Price on 07-07-2003 11:10 AM:

Hi Chuck and Pat

The pile decorated Turkmen tentbands are trappings for the yurt in which newlyweds live for the first year or so of their marriage, according to the anthropologists who report Turkmen practices in the 20th century (Jon Thompson and Peter Andrews, for instance).

Most likely, that was true in the 19th century (and earlier) as well. If so, it accounts for the pretty good condition of most of them - they simply weren't subjected to long use.

Many exist as fragments, but it seems likely that most (perhaps all) of those were cut for practical reasons (easier to display the smaller items; more profit to be made by selling 20 panels than by selling one long band; things like that).

I don't recall reading any accounts mentioning specialized weavers who made pile decorated tentbands as a sort of inter-tribal business. Perhaps someone else knows of relevant information bearing on this.


Steve Price

Posted by Vincent Keers on 07-07-2003 08:51 PM:

Hi Chuck, Steve,

Thanks for the image Chuck.
Don't think they are weaving a partly piled band.
Looking at the the picture again I think that if I put some wheels under my cardboard "do it yourself" loom, this would have helped the weavers a lot.

Hi Steve,
Did you read accounts on weavers making partly piled tentbands outdoors?
I read that the brides made them with the help of mothers. But I read a lot. And most things I read seem to be wrong.
How long would it take to make a tentband?
30 days? 45 days? 90 days?
Think if the weaver does 4000 knots a day she's done ok. No slavery working conditions.
How many knots in 1 band? 120.000?
30 days.

I do not understand this.

But, I'm not a weaver or a bride.

Best regards,

Posted by Patrick Weiler on 07-07-2003 09:01 PM:

Got your goat?


It appears that the weavers are making one section of black goat hair fabric which will be sewn together with others to make a tent covering.
You can see a couple of these black goat hair tents in the background.

Patrick Weiler

Posted by Steve Price on 07-07-2003 09:14 PM:

Hi Vincent

I really don't know how many knots are in a finely woven tentband. If I can find time, I'll do some counting and try to estimate it .

I don't recall reading whether these were made outdoors of under a shelter. Perhaps someone else has access to this information. I'd expect this to be in Peter Andrews' The Turkmen Tent, but I don't have a copy of it.


Steve Price

Posted by R. John Howe on 07-08-2003 09:05 AM:

Steve, Vincent et al -

The most substantial piece of writing Peter Andrews has done that treats the Turkmen tent is in his two-volume "Nomad Tent Types in the Middle East," (1997). I think Steve may be right that there is another earlier piece by Andrews with a title somewhat like the one that he suggests. It may have even appeared in some form in an early Hali. There is also a catalog on an exhibit, "The Turcoman of Iran," that Peter Andrews and his wife Mugul curated and that traveled England in 1971-72.

I have both of the two volume book and the catalog and have been looking through them a bit this morning and will mine them in a few postings.

In response to Vincent's question of how fine the mixed technique tent bands are I confess that I have heard different estimates, including that some may have a knot count that reaches 600 knots per square inch. (That seems plausible since Andrews reports that these mixed technique tent bands are said to take two to three years to produce.)

Marla and Christoff would surely know something of the knots counts in these bands, since they have been analyzing them closely.

And in the catalog mentioned above, Andrews reports that one mixed technique tent band fragment included in the exhibition had a knot count of "H. 13 X V. 12" (156 knots/sq inch) and that it has a single pick of weft between rows of knots and is woven with a symmetric knot tied on alternate raised warps.

I need to do some scanning and then I'll post a few more things.


R. John Howe

Posted by Steve Price on 07-08-2003 10:15 AM:

Hi John

The Andrews book I was thinking of was the two volume set that you have - I was just confused about the title.

I haven't had a chance to count the knots on mine. However, Vincent asked me how many warps there were per inch last week, and I counted 28. Since knots only go over every second warp, the horizontal knot density is 14 per inch in the piled areas. I'd estimate the piled area to be about 50% of the total area, although it would be pretty labor intensive to determine it accurately.


Steve Price

Posted by Vincent Keers on 07-09-2003 09:17 AM:

Hi all,

3.3 kots per cm horizontal. 1 warp per mm.
That could make 5.5 per cm vertical.
19.25 kots per cm2.
192,500 knots per m2
Tent band surface = 3.60 x 192,500 = 693.000 knots
50% = 346,500 knots.
2000 a day (It's a bride) 173 days.
6 months.
If mother helps: 3 months.

Best regards,

Posted by R. John Howe on 07-09-2003 04:28 PM:

Vincent -

Speed on these Turkmen tentbands might be lessed by the scale of the space within which the weaver has to work, but it is known that experienced weavers often tend to weave 10,000 to 12,000 knots in a 10 hour day.

At the Smithsonian Folklife Festival last year we had in the Ersari demonstration tent a partly completed rug (about 100 knots/sq. inch) which would be about 4 feet by 6 when completed. Such a rug would typically be woven by two women sitting side by side and it was estimated that weaving at the rate indicated above it would take the two of them about four months to complete this 4 X 6 rug. It shows concretely how labor-intensive rug weaving is (and that doesn't consider all the hand steps before the weaving).

I suspect that Andrews' reported estimate that it might take 2-3 years to weave one of these mixed technique tent bands might well be accurate. Not steady work but a long period of careful work in constricted circumstances.


R. John Howe

Posted by Vincent Keers on 07-09-2003 06:50 PM:

Hi John,

Yes i know, but a bride isn't a slave...I think......
ok...not yet......
So I gave her time to play around, etc. because she's only 12 years old. Think 2.000 knots a day is ok as long there's not a well to do boy lurking around. If such a guy is found, the knotcount will go up because than pressure rises.
If she makes a 10,000 knots the band is ready in 35 days. And the moment she makes 12,000 knots, she'll be weaving those bands for the rest of her life

But, what ever, I do not see how these piled tentbands are made outdoors. Not in 3 years, not in 35 days. What kind of loom did they use? How did they get the designs? More questions than answers for now.

Best regards,

The rug you saw:
240,000 knots total
8,000 per women per day.
That's 15 days.
Think they are married, so that's why it takes them 4 months.

Posted by R. John Howe on 07-10-2003 07:27 AM:

Vincent -

Yes, I think your calculation shows that weaving is likely done sporadically around other household tasks.

The rate of weaving may be approximately correct but 10-12 hour days of continuous weaving could not be the case if it takes two weavers 4 months to complete a 4 X 6 rug of about 100 knots/sq. inch fineness.

The particular rugs in question were/are woven in homes.


R. John Howe