Patrick Weiler
October 28th, 2014 08:09 PM

Serrated Leaf Motif
The topic of the salon is whether or not there are Khamseh kilims masquerading as Qashqai, Lur or Afshar kilims. Most rug authors and researchers (Is that an oxymoron?) admit that the Khamseh wove kilims. It's just that they don't really know what they look like or what to look for. Some of the possible clues are dovetailed tapestry construction, darker colors than Qashqai, darker warps and heavier feel.
The designs, however, mostly are so common to each other that it is not an easy call. We often look to rugs from a particular tribe for designs which then may have been transferred to their kilims, but this is mostly a dead end. The rug designs which seem to be most likely copied in kilims are those of the gabbeh. And the transfer may be the other way around, from kilim to gabbeh. My reasoning for this is that, until fairly recently (in rug years, anyway) SW Persian tribal kilims weren't very popular in the west. They were used domestically in SW Iran as covers over bedding at the back of the tent and has hangings. They made rugs to sell, but kilims for themselves. And gabbehs were similarly made for home use. So it seems that these domestic production designs were used in both of these formats, gabbeh and kilim, and the more formal, commercial rug designs had commercial cachet and a ready market.
Around a hundred years ago and more, kilims, especially Caucasian, did sell in the west. We have photos showing them in places such as Olana, the home of the Hudson River School painter Frederic Edwin Church. There is a Caucasian hanging as a portiere in this photo:

So, as opposed to looking for Khamseh rug designs which were also used in their kilims, let's look at pile designs which are common to multiple traditions.
One is the Bird on a Pole design, which we find all over the place.
Here is another one; the Serrated Leaf and Star Wars X-Wing design.

Put them together and you get rows of serrated leaves with rows of the X-wing between them.

The above is an Afshar carpet. It is probably over a hundred years old, a "squarish" 48" wide x 52" tall, (122x132cm) with the typical soft orange, brick red, three blues, green, brown, black and white colors. It has asymmetric knots, two wool wefts and depressed very light brown wool warp. The rows of serrated leaves are offset from each other by half a unit, giving the rug a dynamic, almost disjointed look.
Below is an Afshar bagface with the same design, 34" x 25" (84x64cm) and construction - but with that soft orange color weft and the same warp color as the rug, but a bit stiffer handle.

Next is a Khamseh version. It is more square, with darker coloration, 27"x25" (69x64cm) 8x8 knots per square inch and symmetric knots, no warp depression and a softer handle. The warps are very dark brown and the wefts are a dark reddish brown - as though a dark brown wool was dyed red.

Here is another Khamseh version of the same motif, but with a medallion in the center. It has a bit lighter brown warp, 27"x23' (69x59cm), 7x8 symmetric knots per square inch and no warp depression.

And to really mix things up, this probably later, modified version. It looks a bit younger than the other pieces.

This one is 27"x23" (69x59cm), with 6hx10v symmetric knots per square inch, two shots of very dark weft and light brown somewhat depressed warps. It has that Khamseh coloration, but more of an Afshar construction except for the symmetric knots. However, Murray Eiland, in his first tome on Oriental Rugs, writes "Formerly the knot could be used to distinguish village and tribal rugs, but this feature is less reliable now, as there has been extensive intermarriage (with local Persian villagers). The tribal rugs were traditionally woven with the Turkish knot, while the Persian villagers used the Persian knot."
This may, in fact, be a tribal Afshar with the Turkish knot.

I am not familiar with kilims in this design, but I am sure if there are any, someone will point them out. Again, typical rug designs were not often transferred to kilims. One reason is that the structural limitations of kilims lend them to have simpler, more rudimentary designs. And also, possibly, because these simple designs were quicker and easier to make.

Patrick Weiler

Filiberto Boncompagni
October 29th, 2014 06:56 PM

Hi Patrick,

There is a Caucasian hanging as a portiere in this photo
Partially Caucasian, I would say. The inferior parts are from a Verni or Sile - whatever: the Caucasian S-dragon carpets, generally made in soumak technique. They seem to be sewn to an Anatolian (Kurdish perhaps?) kilim.


Patrick Weiler
October 29th, 2014 09:03 PM

Pieced it together, did you?

I never saw that a section of Anatolian kilim was attached to the lower Dragon brocaded or soumak pieces. His decorator probably thought he wouldn't notice.
I took trip a year or so ago to the Ahwanee Hotel in Yosemite Valley. The hotel opened in 1927 and has many rugs and kilims. Some were original furnishings, including some nice Caucasian and Persian kilims still hanging on the walls.
My next project is to put together a Traveler's Report with some pictures of them. Here is one kilim hanging outside the elevator:

This one looks like a later piece. They even have sections of a flatweave Turkmen tent band hanging in one room.

Patrick Weiler
November 3rd, 2014 11:45 PM

Herati motif deconstructed
The Serrated Leaf design, often found in Khamseh weavings - especially in their bags - is actually the Herati design, not the "Serrated Leaf and X-Wing Star Wars design" (as it might become if I repeat it often enough).
It seems to have its origin in the Greater Khorasan region, which counts Herat (hence the name Herati Pattern) in Afghanistan and Mashad in Iran as large population centers.

"Khorasan in its proper sense comprised principally the cities of Balkh, Herat, and Taloqan (now in Afghanistan), Mashhad, Nishapur, and Sabzevar (now in northeastern Iran), Merv and Nisa (now in southern Turkmenistan), and Samarqand and Bukhara (before in Turkmenistan, now in Uzbekistan). Some believe that at certain times Khorasan covered a wider area, which included parts of Transoxiana, Soghdiana, Sistan, and extended to the boundaries of the Indian subcontinent."

Here is a drawing from an oriental rug motif guide on Jessie's Oriental Rugs web site, in Massachusetts:

And here is a tiny chanteh face with a section of the herati design. The chanteh design shows the "secondary gul" flower that sits between the herati motifs in an allover herati-design carpet, and four serrated leaves which would be from adjacent herati motifs if one were to expand this design onto a larger format.

If you're paying attention, (which is not the case as has been shown already in this salon, yes I mean you :cheers:) you can see that this piece has been repaired, as evident from the different-colored warp thread at the top compared to the bottom. The two top corners for the width of that light warp and about an inch down have been rewoven.

It is about one foot square (30x30cm) in size and has 7h x 9v asymmetric, open right knots per square inch, about 63 per square inch. Not too fine, yet it has a somewhat stiff handle due to some of the warps being moderately depressed.
With a mixture of both Afshar-like weave, yet Khamseh coloration and design, this piece fits the description of the Khamseh in general, squeezed in between the Afshar and Qashqai yet proudly different from both.

Patrick Weiler

Jim Miller
November 4th, 2014 06:04 AM

Leafs and stars

I think this may be another variant of the herati design, except that the center of the motif (stars in my case) are offset in each row rather than aligned in vertical lines. This closes the space between the motifs and puts the serrated leaves on top of one another.

It was advertised on line as Qashqai, but also described as 100% silk, so I don't have a lot of faith in the seller's attribution. It has symmetrical knots and is single wefted with salmon colored wefts. The sides look like that have been re-overcast, so no clues there. Here is a closeup of the back.

Any ideas on the correct attribution?


Patrick Weiler
November 4th, 2014 06:42 PM

Not Khamseh

I think you are right about the modified herati pattern design. Very simplified and an elegant piece. I wouldn't say Qashqai, but NW Persian. No offset knots to indicate Kurdish, although they are fond of the little rabbit-head/cloudband motif. The construction is too regular to indicate Shahsavan.

Patrick Weiler

Joel Greifinger
November 5th, 2014 01:27 AM

De-construction Zones

As you noted, deconstructing the herati pattern was all the rage with tribal groups throughout Persia. However, different groups seem to have had favorite ways to chop it up and geometricize it. You've posted a number of pieces with the most prominent Khamseh version, widely referred to as the "Serrated Leaf and X-Wing Star Wars" design.

Next door in Kerman, the Afshar were producing a variation, mostly for use on their khorjin. Here's a typical example:

The next one has a herati variant with that Afshar look, but also has the type of boteh minor borders that gets shared among the various South Persian tribal weavers. In this case, the combination of ivory and grey warps might tell us more. But what?

I've only found a couple of full-sized Afshar rugs that have their characteristic deconstructed herati as a field pattern. The first is from Tanavoli's Afshar monograph. It is one of the few dated Afshar pieces, the date woven in being 1245 (1829-1830):

the other is probably from the latter end of the 19th century:

If there are other Afshar rugs out there with this field design, I'd sure like to see 'em.

Meanwhile, the Kurds (both in NW Persia but also around Varamin) were performing their own variety of herati deconstruction, often appending a characteristic elem:

and this one where outrageous fortune has unfortunately caused all but a trace of the elem to disappear:

This type of Kurdish design was the subject of some very heated discussion in a number of archived Turkotek threads, since it seems to cause some viewers to see dragons, fighting animals and other zoomorphic formations:


In terms of Persian tribal bags, this more naturalistic, floral version seems the exception:


Patrick Weiler
November 5th, 2014 03:04 AM

Dragon me into something?

Those aren't dragons. They're worms. As in you've opened a can of worms.
There are several versions of this modified herati pattern, from several different tribes and regions.
There is this version sold at Bonhams which has a superficial resemblance to those Kurdish pieces, especially the first one in your post, but I think this one is based on a tree-of-life design:

And this one from Spongobongo, where the only thing left of the herati device is the leaves in this deranged arrangement.

And one on ebay claiming to be a "Scorpion Design".

One from the Bakhtiari region, (a rugrabbit piece) with serrated leaves, but no herati imagery otherwise. Was it derived from the herati design, or was it just a single motif taken at random from the design pool?

This isn't even any of the ones in my collection. Yet.

Patrick Weiler

Patrick Weiler
November 5th, 2014 07:53 PM

Leaf Gone Wild
Bijar weavers are renowned for their interpretation of the herati motif in rugs.
Here is one from Pinterest in a medallion format:

This one is probably the smallest Bijar version, in a chanteh:

It is 14"x12" (36x30cm) and 120 knots per square inch, with fully depressed warps. Not very many Bijar bags are found, much less a small one like this.

Here is one from KurdOnline, with no dimensions, also different in design than our featured Star Wars weavings:

And, an interesting development, here is a Bakhtiari bag that was with John Collins at one time. John has exceptionally good taste in weavings and is especially well known for his expertise in Bijar work:

The flatweave border at the bottom is familiar.
Now, back to the topic at hand, here is a very small piece, 9"x10" (23x26cm) with some moth damage and worn selvages, but with what may be a combination of the herati with the tree design:

It has 5hx8v flat symmetric knots for 40 knots per square inch, quite coarse, but charming. I would place it in NW Iran, but there aren't enough clues to confirm who wove it.

Another tree version, with a diagonal border and wider format preferred by Afshar, but with a flat back, a variety of weft colors including a very dark blue, 7hx9v for 63 symmetric knots per square inch. Also probably NW Persian:

You can see remnants of the red/green closure ropes at the bottom.

And to round things off, this hybrid design version. These two pictures were taken with a flash, so the orange border is actually a brick-red and the piece looks washed-out. It has the same coloration as the other three pieces above.

This one has 8x8 symmetric knots, 64 per square inch, very moderately depressed warps, 14"x11" (28x36cm). All three of these last pieces seem to be from the same tradition.

Patrick Weiler

Patrick Weiler
November 5th, 2014 10:48 PM

Back again

I was looking back at your nice bagface and noticed that it may be single-wefted. It is not easy to see without looking closely.

Researching Varamin bags, I have found that they are generally single wefted, and often contain offset knotting. And the minor red border has similarity to a common Varamin border. Your bag has the typical, moody blue field, but it is hard to tell if it also has that Varamin glorious orange

Patrick Weiler

Jim Miller
November 6th, 2014 04:30 AM

It is single wefted. Light red or salmon colored wefts. It is a distinctive color. I had thought about Veramin, based on the single wefts, but based on my limited experience I usually think of Veramin as have a finer weave and having more somber and darker colors.
It has a dark blue background to the field and a orangish read in the background of the smaller border and in many of the motifs in the lighter border. But I can't say if these are typical Veramin colors.

Joel Greifinger
November 7th, 2014 01:12 AM


There is this version sold at Bonhams which has a superficial resemblance to those Kurdish pieces, especially the first one in your post, but I think this one is based on a tree-of-life design

I don't think it's a superficial resemblance; I think they're kin. Here is the one you posted:

(You really shouldn't hang them sideways, it strains the structure :banana:)

and here are some others that have versions of the spotted trees. I figure all of them to be Kurdish:

and, here again, as in the Kurdish version of the 'Khamseh border' that I posted about in the 'Other Often Appropriated Motifs' thread, it appears in a chuval version.


Joel Greifinger
November 14th, 2014 05:52 AM

Spotted Trees in Outer Space
This bag face may be a somewhat different variety that nonetheless belongs in the previous Kurdish group. The complementary-weft weave above and below the closure tabs (which also occurs in some of the others in the last post) may provide a more specific clue to its origin, but not one that I can decipher:

The field alternates between a deep blue ground and a corroded brown that gives the motifs a stark 3-D presence that's tough to capture in a photo.