What knot?

[Phil Smythe Member Find all posts View Profile Close Phil Smythe]

This thread is a carryover from a question I asked in another thread. For my own uneducated novice benefit I am trying to ascertain how to differentiate knot types as seen in photos of the backs of carpets, that is when one does not have the rug in hand, only a photo. Sorry to bore you experts with what to you may be obvious, but one has to ask questions to learn after all.

I recently had the opportunity to sit in while a fragment of a rug with asymmetric knots open left was taken apart and photographed (close-up photo of back of those knots below). Extrapolating from what I see there I ‘thought’ the knots in photo marked 1 below (cropped from a photo in this thread (http://www.turkotek.com/VB56/forum/m...pair-of-torbas) were asymmetric knots open right (because of how the knotting in each knot appears to ‘cross’ in the opposite directions than the knots in the ‘open left’ photo do).

However I am told that the rug whose knots are show in photo 1 (below) are symmetric. But what my uneducated novice eyes are seeing there is not what I expected a symmetric knot to look like, that is I had expected the two nodes of a symmetric knot to be more evenly ‘butted-up’ against each other, rather than appearing ‘crossed’ as they seem to me to be.

So my question is, if the knots seen in photo 1 below are symmetric, how does one differentiate between the look of those and asymmetric open right knots and / or what do asymmetric open right knots look like (in photos of the back of a carpet)?

Any help regard in my question then would be MUCH appreciated.

PS. Please note that my question concerns only what the back of the knots look like as shown in photos, not how you can tell what knot is what when a rug is in hand.

[Steve Price Administrator Find all posts View Profile Close Steve Price]

Hi Phil

I don't think there's any way to determine knot type from the back. To find knot type, start by running the back of your hand along the direction of the warps. The direction that offers less resistance points to the end that was woven first. With that end at the bottom, find a spot where two colors meet and fold the rug back along the weft that goes through that spot. This exposes the base of the knots, and it's easy to tell which tuft goes with which knot collar at places where two colors meet. Symmetric knots have both tufts in one collar. Asymmetric knots have one tuft in the collar, one tuft next to the collar. The side with the tuft adjacent to the collar is the side that's open.

Steve Price

[Joel Greifinger Administrator Find all posts View Profile Close Joel Greifinger]

I don't think there's any way to determine knot type from the back.

Agreed.

Joel Greifinger

[Dinie Gootjes Member Find all posts View Profile Close Dinie Gootjes]

Hi Phil,

I completely agree with Steve and Joel, twisting of the knot nodes has nothing to do with the kind of knot used. The question then is, what does cause it ? I first thought it could be the presence of warp depression, but I this Kurd has considerable warp depression and perfectly straight knots:


I have an idea that the straightness of the knot nodes is caused by the direction in which the weaver pulls the knot when making it, left, right, or straight down. So that is independent of the type of knot. One example would be this Kurdish rug, in which the (symmetrical) knots are clearly pulled to the left. The nodes are virtually dancing .



But one robin doesn't make a spring (what is that anyway?

[Dinie Gootjes Member Find all posts View Profile Close Dinie Gootjes]

Grrr, I am losing part of my reply, and the edit function doesn't seem to work for me. I should still be registered as a 'junior member' I guess

[Dinie Gootjes Member Find all posts View Profile Close Dinie Gootjes]

Still losing sentences, I give up. It should be clear what I meant.

[Phil Smythe Member Find all posts View Profile Close Phil Smythe]

Hi Phil

I don't think there's any way to determine knot type from the back. To find knot type, start by running the back of your hand along the direction of the warps. The direction that offers less resistance points to the end that was woven first. With that end at the bottom, find a spot where two colors meet and fold the rug back along the weft that goes through that spot. This exposes the base of the knots, and it's easy to tell which tuft goes with which knot collar at places where two colors meet. Symmetric knots have both tufts in one collar. Asymmetric knots have one tuft in the collar, one tuft next to the collar. The side with the tuft adjacent to the collar is the side that's open.

Steve Price

Thank you for the feedback Steve, I was aware of what you say when you have the rug in hand.

However, as I state in my post, I was looking for a method when you only have photos.

But thanks anyway.

By the way I notice that although I have been signed up here for some time I am still just listed as 'registered'. Although that moniker doesn't seem to make any difference re posting etc, what criteria qualifies one as junior member, member, etc? Number of posts, time signed up, etc?

[Phil Smythe Member Find all posts View Profile Close Phil Smythe]

Hi Phil
I don't think there's any way to determine knot type from the back.
Steve Price

Agreed.
Joel Greifinger

Steve / Joel,

As alluded to in my original post, I recently had the good fortune to be present when a large collection or rugs were being photographed (and several that were fragments taken apart). Mostly antique Chinese but about 15 or so were ‘Central Asian’ / not Chinese of any sort.

Now as I assume you know, all or 99% of Chinese rugs are made with the asymmetric knot open left (in the sample lot I saw of well over a hundred rugs, all were open left). And when one used the methods Steve suggests in his post to discern the type of knot, all / 100% of those rugs (i.e. well over one hundred) had identical shaped knot structures when viewed from the back (save for those with warp depression of course, which were in the minority. Still open left of course, but different shape.)

When the central Asian rugs were photographed, the photographs of the backs of the ones with a knot open left all had basically an identically ‘shaped’ knot structure to the Chinese open left, but all the rugs with the asymmetric open right knot all had the opposite shaped knot structure - when seen from the back. (There were no carpets with symmetric knots, and hence why I made my original post to try to see how one could differentiate between those and asymmetric knots in photos).

Now I don’t know if what I saw first hand was just coincidental, but given there was not one mismatch using any of the ‘methods’ (that Steve suggests) to establish knot type from the front vis a vis the back, it seemed a solid sample with a 100% non deviation on the ‘look’ of the back of the two asymmetric knots types. That is if the knot was found to be open left it had a distinct structure / shape on the back 100% of the time, and if found to be open right it had a distinctly different shape on the back 100% of the time.

Anyway, I will try to get a couple of the photos the backs of knots from the Central Asian rugs from that photo session to post to show what I mean re the 'difference' but it may take a few days, or more, as I do not have them in hand but think I can get a least a couple for use as example.

[Steve Price Administrator Find all posts View Profile Close Steve Price]

Hi Phil

At the very least, you'll have to know which end of whatever is in the photo was woven first.

Steve Price

[Chuck Wagner Administrator Find all posts View Profile Close Chuck Wagner]

Hi all,

Certainly, the definitive information is gained by examining the front of the rug, including the simple but important top-from-bottom distinction. Assuming that top from bottom is understood, I think it's more accurate to say that you cannot be positive about the knot type when viewing from the back, mostly because on rare occasions symmetrical knots are tied with higher pull tension on one side that the other, which gives the nodes a -slight- offset that is visible from the back. Otherwise, for asymmetrical knots, the lower of the two nodes is on the open side.

Regards
Chuck

[Phil Smythe Member Find all posts View Profile Close Phil Smythe]

Pardon delay in getting these photos back folks.

So, below is an example of what I alluded to in my above post re the knot shape on the back of a rug.

The top image shows the backs of asymmetric knots open left. Of well over a hundred open left knotted rugs that I witnessed photographed, every single one had this shaped knot when viewed from the back. Note how the left node ‘appears’ to come out from under the right node, and the general direction of the threads in the yarn. Certainly no coincidence in this smapling given the number of carpets involved in the ‘sample’(100+).

However, there were only a half dozen or so rugs photographed with an asymmetric knot open right, but every single one of those had a knot shaped like in the bottom image when viewed from the back. Note how the right node ‘appears’ to come out from under the left node, and the general direction of the threads in the yarn. Not really a big enough carpet sample to definitively say whether a coincidence or not, but………………………..

At right the white line shows how the yarn wraps around the warps (black dots) and which way the respective knots open (when seen from the front).

So as long as one knows that the carpet was woven with an asymmetric knot, then discerning what kind of asymmetrical knot it is, whether ASOL or ASOR, is or certainly seems to be possible from a photo of the back.

Unfortunately there were no symmetrically knotted rugs in what I saw photographed, but it seems what Dinie and Chuck have pointed out, that is the shape (of the back of a knot) may very much depend on which way the weaver pulled the thread down (i.e. directly down, or left, or right) on the symmetric knot. This then seems to throw a spanner in the works regarding differentiating symmetrical knots from asymmetrical knots when viewed from the back.