Posted by Filiberto Boncompagni on 01-24-2003 06:10 AM:

Caucasian Prayer Rug

Thank you, John, for another interesting report on a TM rug morning.

I think I can help the owner of the last rug in you presentation, the prayer rug dated 1311.

Let’s look at Ralph Kaffel’s "Caucasian Prayer Rugs", pages 72/73. Rug # 28, labeled as:
KARABAGH -dated AH 1312 (1894) 0.84 X 1.17m(2'9" X3'10")

Amazingly similar, isn’t it?
Here is what Kaffel wrote about it:

"This is one of those atypical rugs which occasionally emerge from Caucasia or Turkey. While no precise analogy to this rug has been identified, either in literature or by anyone with whom I have discussed it, Danny Anavian, a New York antique rugs dealer, has examined the piece and proposed that it is Kurdish work from the environs of Shusha in Karabagh. The offset depiction of the flowers in conjunction with the vertical stripes, the use of a particular shade of yellow, and the overtly Islamic character of the rug (with its stylized mosque), as well as some technical characteristics, have led him to this conclusion. Shusha was a major stop on a trade route between Persia and Caucasia that was used by Kurdish travellers, and the presence of large Kurdish communities in the area is indicated on a number of ethnographic maps.
The configuration of vertical stripes is typical of south Caucasian Karabagh rugs, while the polychrome zigzag border is consistent with Kurdish work. The inclusion of a mosque is quite rare, however, and the horizontally drawn lacy filigree is another unusual feature. Whilst Caucausian prayer rugs with depictions of mosques are uncommon, there is a group of late Shirvan rugs that feature very realistic depictions of famous mosques."

Well, we have a very good analogy here!

And here is the

Warp brown and white wool, mixed together; Z2S; 18 threads per inch (72 per dm) Weft brown wool; Z2S; 12 knots per inch (48 per dm)
Pile symmetrical knot; 108 knots per sq. inch (1728 per Ends plainweave with shades of ivory and brown weft
Sides 2 cords overcast
Colours dark red, cadmium red, gold, yellow, dark green, blue-green, indigo, periwinkle blue, salmon, peach, dark brown, mid-brown, tan, light brown, black, ivory (16)

Kaffel Collection"

Yes, it belongs to Mr. Kaffel.

John, I’d like to see an higher resolution scan of the TM morning rug if you have it.
For what I can see from your picture the colors should be the same, keeping in mind the variations that can occur in different lighting conditions, during the process of developing the film, printing the photo and scanning it.

Were the two rugs done by the same weaver?
The Mosque’s dome is rounded - as it should be - in the older one, though, while in Kaffel’s one the dome is rendered with straight lines…
Or possibly they come from the same workshop/village?
I hope the owners of the two prayer rugs will read us and provide with more input.
Best regards,


Posted by Marvin Amstey on 01-24-2003 05:33 PM:

Hello Filiberto,
Great find for comparison. It is also a good example of a fallacy in using style to suggest age: round dome - older rug; straight-sided dome - newer rug. We hear this kind of reasoning frequently, and you have provided dated examples of a one year difference! I hope folks will remember these examples and not be so dogmatic with pronouncements about style and age.

Posted by Filiberto Boncompagni on 01-26-2003 02:26 AM:

Hi Marvin,

The difference in drawing could also suggest that the 1312 rug was woven by a less experienced weaver…or it was the same weaver who simply changed her style?!
Who knows?
Only a comparison of the rugs structure can give us some clues.
Let’s hope John will be able to contact the owner of the TM one. I’m curious…


Posted by Filiberto Boncompagni on 01-27-2003 10:35 AM:

John sent me a better image. I put it here followed by my scan for better comparison:

Borders are identical. End finishes could be the same, I’m not sure.
The date and the inscription are more naturally curvilinear in the TM rug.



Posted by Richard Farber on 01-27-2003 02:37 PM:

Dear Mr. B.

very interesting with way that the creators of these carpets dealt [or in fact didnt deal] with the problem of moving the border around the corners . . . [somethings that is usually beautifully solved in most susanis]
The fact that it hasnt been solved might be indicitive of the fact that these are carpets to be made quickly for the market . . .

what do you think ???


R. F.

Posted by Yon Bard on 01-27-2003 04:28 PM:

Border resolution is never a consideration in tribal and cottage-industry rugs. It's present only in workshop rugs made from cartoons. If anyone knows of exceptions, I'd like to see them'

Regards, Yon

Posted by Steve Price on 01-27-2003 05:12 PM:

Hi Richard and Yon,

I suspect that the reason border resolution is done smoothly in suzanis is that the design can be drawn on the ground cloth in advance, so the layout can be done before any of the embroidery begins. In pile weavings, the design has to progress row by row, from bottom to top, and resolving the upper corners is usually impossible.

I don't know that it's correct to say that border resolution is never a consideration in tribal and village weavings. I think it is just that it can't be done in any reasonably convenient way. The borders at the end woven first are often resolved, suggesting that when it can be conveniently achieved, the weaver does it.


Steve Price

Posted by David R. E. Hunt on 01-28-2003 07:06 PM:

Turkic Tradition?

Filiberto and all- If my memory serves me correctly, this resolution of corners, or maybe it would be better to say failure to resolve corners, and this abbreviation/interruption of border design is a distinguishing characteristic of rural Turkic weaving tradition. If I can find the source I will quote, but I think that a quick overview of tribal weavings from indigenous turkic sources will demonstrate a high percentage of correlation- In short, a lot of turkic rugs do this- Dave

Posted by Steve Price on 01-28-2003 07:20 PM:

Hi David,

Failure to resolve corners is characteristic of any rug not woven from a cartoon. That is, any tribal or village rug is likely to show this, urban workshop rugs don't.

The reason is fairly straightforward. Rugs are woven from one end, a line of knots at a time. So, the weaver generates the field and borders as she proceeds from the "bottom" (beginning) to the "top" (end woven last). Making the borders turn the corners is not difficult at the "bottom", and generally the borders are resolved at that end. When she gets to the top of the field, the border design is usually not at a point where it can smoothly turn a corner, so it doesn't.

In workshop rugs done from cartoons, the cartoon is drawn to make everything symmetric in both planes. In fact, the cartoon is often only of one quadrant of the rug, which is repeated as its mirror image horizontally and then again vertically. The rug not only has perfectly resolved corners, it has two axes of symmetry.

I hope this helps. One way for a beginner to recognize a workshop carpet is to look for unresolved borders at the top.


Steve Price

Posted by Yon Bard on 01-28-2003 08:35 PM:

Steve. glancing through any number of illustrations you'll observe that very few Turkoman, Caucasian, and even Persian tribal rugs have resolved borders even at the bottom; and in the very few in which the bottom border comes out with an even number of motifs, the design fails to 'turn the corner' in an orderly fashion. The logic that you use for explaining why the borsder isn't resolved at the top applies equally well to the bottom, since after all knotting starts at one end and proceeds horizontally to the other. And, since the warps had been laid out in advance, the weaver doesn't have the option to adjust the width of the piece in order to get a resolvable border-an option she would have in principle as far as the length of the rug goes.
I am convinced that border resolution was not anything they even remotely strove for. Otherwise, how could they have slept at night knowing how dismally they had failed?

Regards, Yon

Posted by Patrick Weiler on 01-28-2003 08:56 PM:

I've got you cornered


I have Heriz and Tabriz rugs with unresolved corners, both top and bottom. The Tabriz is 17' long, so it was probably made in a large workshop. Shiraz, Hamadan and Qashqai, all have unresolved corners top and bottom. From this I conclude that just because a rug is made in a large town workshop does not mean the corners will be resolved.
I have a crude Kurd with resolved corners, but the border is just rosettes. So, I think one of the problems is not that it is easier to resolve the corners at the bottom of the rug, but it is changing the orientation of the motif or design from a horizontal to a vertical attitude that is difficult, top AND bottom.

Patrick Weiler

Posted by Steve Price on 01-28-2003 09:13 PM:

Hi Yon,

I haven't done any data gathering, but have casually skimmed through some published photos this evening, with an eye on corner resolution. My impression is that there are a number of village and tribal rugs with awkward corner resolution at both ends, but in many the awkwardness is only at one end. It seems particularly evident in Kazaks, perhaps because their borders tend to be fairly large scale. Is the awkward end always the end that was woven last? No way to tell from a photo.

As for how the weavers found inner peace after making awkward corners if they really wanted smooth transitions, my guess is that they simply accepted the limits of their abilities, even as you and I do.

This seems like a question that someone like Marla Mallett or Mike Tschebull, who have spent time with village weavers, could answer easily. Do they resolve borders when they can, or are they totally indifferent to the matter? Inquiring minds want to know.


Steve Price

Posted by Steve Price on 01-29-2003 06:33 AM:

Hi Yon and Patrick,

I believe Patrick hits the nail pretty close to head on when he notes that some border designs lend themselves more readily to appear to smoothly turn a corner than others do, and after looking at more photos last night I agree that very often, corners are not resolved cleanly at either end.

Do the weavers care? I suspect that sometimes they do, sometimes they don't. This may be a local, rather than global aesthetic. Here is a rug in which the weaver obviously did care.

It's plate 93 in Kerimov's book on Caucasian rugs, attributed to Shusha (Karabagh's capitol). The top is at the right side of the image. Notice that the main border is a series of stepped polygons, linked end to end. Notice, too, that the corners turn cleanly at both ends. Finally, notice that this is achieved by making the next-to-last horizontal polygon smaller than the others. That is, it is obvious that the weaver cared enough about the corner resolution to modify the border motif a bit.

I've seen a number of Karabagh rugs (one that hangs on a wall in my home) with the same border and the same little trick to make the corners resolve. Here's a closer look at the one I own, where the change in motif dimension is very obvious:

Lest you think that the irregularity is an internal elem, the irregularity is not at the end where the weaver began, but at the end where she finished. I know this for sure on the one in my home.

So, I think the issue is more complicated than it might appear at first and probably warrants a fairly extended discussion of its own.


Steve Price

Posted by Filiberto Boncompagni on 01-31-2003 01:21 PM:

Dear all,

As it turned out, John found the owner.
The rug belongs to Mrs. Sharon Larkins-Pederson.
She asked to Steve Price to have a look at it. This is the report:

"I just spent a little time with Sharon's rug. It is very nice, and the
colors match those on my monitor rather well.

It is less finely woven than its lookalike, with 56 symmetric knots per square inch (7 x 8).
The warp is two ply of tan wool, and there are two shots of tan wool weft between each row of knots. The edges have two cables overcast with the weft. The ends each have some plainweave, an inch or so.

It feels like a Kazak, loose and floppy and with long pile. .
On the other hand, it doesn't have the usual red wefts of late 19th century Kazaks (and I believe that the inscribed date of 1310 - 1892 or so - is likely to be the time it was woven). So, if asked to make an attribution,
I'd hedge my bets and say Genje."

Interesting. Kaffel's rug has a density of 108 per sq. inches, almost the double.
My non authoritative opinion is the two rugs were made by different weavers belonging to the same environment (family, village, tribe, Kustar, whatever).
The weaver of Mrs. Sharon’s drew it in a more handsome way (I’m speaking of the Mosque Dome, the crescents and the inscribed date), in spite of doing it with half of the knots number.
What do you think?

P.S. - I officially invite Mrs. Larkins-Pederson to tell us more about her rug on these threads.

Posted by R. John Howe on 02-03-2003 05:43 PM:

Dear folks -

With Steve Price's assistance as impressed (perhaps oppressed) technical analyst, Filiberto and I have been pursuing further information about this TM rug morning piece that so resembles one in Ralph Kaffel's book.

Filiberto invited Ms. Larkins-Pederson to tell us a bit more, but knowing that folks are often reluctant to join our proceedings here, I wrote and ask her if we could at least quote an additional snippet we had on how she had acquired this piece.

She responded as follows:

Here is a paragraph which you may use as you like;

The rug belonged to Janis Carter who was , by her own description, "Queen of the B Movies" in the 40s and 50s. She was a very well educated (Case Western) woman who was 'discovered' by Howard Hughes when she was in a review in New York. She made several dozen movies but the one that most people recall is the one in which she played John Wayne's wife. I believe it was a famous war movie. She left Hollywood and moved back to New York and was the original 'Revlon Girl' and was also the first woman to host a live television show. She met Jules Stulman and married him and retired from active working. They traveled a lot because he was a dealer in rare and exotic lumber. He also was very philanthropic in Israel's cultural life so they traveled there often as well. As a result, they often stopped in Istanbul and picked up interesting items and this is one of the items that she left to me. I met her when I was working at Duke Medical Center and we became great friends, one reason being that she thought I looked like I could have been her daughter. (She never had any children). She died of a heart attack in her sleep several years ago. Although I know nothing about rugs it was clear that it was a prayer rug so I never felt comfortable having it on the floor and walking on it.. My sister, Carol Ross, who is a volunteer at the Textile Museum in DC offered to lug it to Syria when she was posted there and had it cleaned and some repairs done. It's been hanging in my hallway ever since. It is nice to know it's probable history and I appreciate your making that opportunity available. And, as I told Mr. Price, if my retirement account keeps tanking I may have to sell it! But I hope not! In any case, I have written all this down and put with the information you had on the site and put it with the background information on other special items I own so whoever ends up with it someday will have this background.

Best regards,

Sharon Larkins-Pederson"

Thanks very much for this interesting information.

This is the kind of tale that might encourage Ralph Kaffel, who sometimes looks in, to say something about his own similar, but twice as fine, piece.

We'll see.


R. John Howe

Posted by Marvin Amstey on 02-04-2003 12:02 PM:

For Ms. larkins-Pederson: when were you at Duke? Maybe we were there around the same time; 1960-64 for me.