Posted by Patrick Weiler on 02-02-2003 11:32 AM:

Pile sumak?


The second page of the Salon shows a Shahsavan khorjin face. You say:

"The next piece was, I think, also in Jerry's ACOR exhibition.
This is a Shahsavan sumak khorjin bag face. Joe said that at one time he owned both halves of this khorjin, but that he had separated them and sold one to Wendel Swan. A very similar complete khorjin set was published as Plate 113 in the 1996 ICOC "Atlantic Collections" catalog. Wendel pointed out from the audience that this is a quintessential Shahsavan design."

On my monitor it appears to be a pile woven khorjin face. One feature that shouts "pile" is the white outlines of the diamonds in the field.
Another is the white border with colored diamonds.

The Atlantic Collections Khorjin is pile woven.

Sumak or pile, I would like to have it in MY collection.

Patrick Weiler

Posted by R. John Howe on 02-03-2003 05:04 PM:

Pat -

Wendel needs to speak up just to confirm, but I think you may be right, that this is a pile piece.

As you may remember there is a kind of debate, with some folks claiming that it is doubtful that the Shahsevan wove pile pieces, and others thinking that there are not a few pile pieces that should rightfully be attributed to the Shahsevan.

I think the sumac labeling is just my error.


R. John Howe

Posted by Wendel_Swan on 02-03-2003 05:14 PM:

Hi Pat and all,

Joe's Shahsavan khorjin face is definitely pile, not sumak. You may not be surprised to learn that mine, its mate, is also pile. The colors and wool quality are almost unsurpassed.


Posted by Bertram Frauenknecht on 02-04-2003 05:15 AM:

Pile Shasavan

Hi everybody,
It is really amazing to see how long a fairy tale stays with educated people. Why should a huge confederacy as the Shasavan not have produced pile rugs. On my last visit to their dwellings I was sitting on pile rugs that they claimed to be made by them. That was 1980.
And those rugs looked as poor as the makers. ''
There are so many so-called Caucasian rugs that don't fit in any of the normal criteria and which are attributed to South-East Caucasus. The question who lived there can easily be answered.
I bet Wendel is smiling now.:-)
All the best


Posted by Mike Tschebull on 02-04-2003 08:52 AM:

I suppose, in the end, it’s harmless enough to insist that “Shahsavan” wove “pile rugs”. Nevertheless, let me have a go at it. First, it’s better to define the weavers in question as “Azarbayjani nomads”, so there’s no confusion over settled vs. nomadic peoples. “Shahsavan” is so imprecise.

Then, it’s instructive to delve into statistics a little bit, which will quickly show that the idea of “huge” numbers of Shahsavan is, indeed, a fairy tale. Settled villagers in Azarbayjan - and in the Caucasus - hugely outnumbered nomads in the 19th century, and were the source of most weaving, especially pile weaving,

Next, a good conversation with a competent anthropologist and/or some field work will reveal enough about the technology of nomadism so that the idea that Azarbayjani nomads were ever more than fringe weavers of pile objects is shown to be nonsense.

It is of course frustrating to be unable to accurately attribute rural pile weaving.

Posted by Steve Price on 02-04-2003 09:10 AM:

Hi Mike,

Could you expand on some of this? Specifically,

1. "Azerbaijani nomad" is a term I've rarely seen used as a descriptor. Is it more precise than "Shahsevan", and if so, in what way?
2. How many Shahsevan tribespeople were there in the 19th century? The same number might seem huge to one person and trivial to another. If there were enough of them to crank out lots of flatweaves, wouldn't that be enough to crank out lots of pileweaves?
3. What is the "nomad technology" argument that makes it highly unlikely that Shahsevan did pile weaving? Does the same apply to Turkmen, Belouch, Yoruk, etc., nomads?


Steve Price

Posted by Bertram Frauenknecht on 02-04-2003 09:35 AM:

Hi Mike,
I agree that much weaving was done by villagers. I didn't say anything about nomads, for me Shasavans are not necessarily nomads. This tiny nomadic group we still have are only a small portion of the Shasavan. A settled nomad still after generations would consider himself part of the group, even if he hates the nomads for a number of reasons. (Fieldwork in Turkey)
Their Khans were telling them what they are, as it happened in every feudal society. Just look into Germany 17th c.
Fieldwork done 150 years after their structure broke down simply cannot reveal much (or anything at all) of their life then. Only witnesses from those times can.
What is still the same (compare Turkmen nomadic life), a wealthy nomad (being a successful entrepreneur) had and has today several wives. Through that they make space for weaving which adds to their wealth. His poor neighbour has to sell the wool he produces, for nobody has time to do something with it.
I believe that our ideas about Shasavan has little to do with the people who lived in Azerbaijan in the 17th and 18th century and not too much more with the ones in the 19th, when they went through their distruction.
The poor fellows that Tapper researched in the 1960's were far away from their ancestors.
I remember quite well how impressed I was by these wild nomads I visited first in 1972. What a world opened to my eyes.
Just that I didn't really understand anything. I knew nothing about their history. Things have changed.

best regards


Posted by Bertram Frauenknecht on 02-04-2003 09:59 AM:

Just a word about statistics,
In 189o Hahn, a teacher at the 'highschool' in Tiflis writes about:
" the 'tatares' who come from the southern Moghan, use to settle in the outskirts of the city with piles of rugs and weavings for sale.
They speak a language which is a mixture of Persian and Turkish"
He numbers them to 900 000 people.
Hard to believe!


Posted by Bertram Frauenknecht on 02-04-2003 10:17 AM:

I missed completely the nonsense part that nomads don't weave or weave little. Those poor nomads were often not so poor. They loved to be independant. That's why many became nomads again after being forcefully settled for many years.
I seriously doubt that an anthropologist understands much of nomadic life. We had some strange examples during Nazi times and you all know what happened. I bet you mean an ethnologist.
And again, fieldwork today, oops...

Posted by Rick Paine on 02-04-2003 04:38 PM:

I think we have a little European/N. American jargon problem here. In Europe, an "anthropologist" is one who studies human physical characteristics. This person would be labeled a "physical" or "biological" anthropologist in North America. In North America, "anthropologist" refers to a whole range --including physical anthropologists, archaeologists, ethnographers, and ethnologists (the last two often grouped here as cultural anthropologists). It is pretty typical in N. America to use "anthropologist" to describe an ethnographer.

Really fascinating stuff in this Forum. Thanks.


Posted by Mike Tschebull on 02-04-2003 07:05 PM:

One problem with the term “Shahsavan”, as Herr Frauenknecht makes clear, is that to some, it implies settled as well as nomadic people. “Shahsavan” means “nomads” to me and to most other specialists in this area; there is and has long been a rather sharp distinction between nomads and settled people in Azarbayjan. Another problem is that there are/were Azarbayani nomads not belonging to the “Shahsavan”.

Headcount: An 1870 estimate is 12,000 Shahsavan families. Tapper estimates 60, 000 - 80,000 people. The number rises and falls over time, depending on settlement policies in Iran, border disputes, winter pasturage available, etc.

The number mentioned by Herr F. from Herr Hahn - 900,000 - seems wildly off the mark. A well-regarded Historian, Muriel Atkin, estimated that the entire population in the Caucasus at the beginning of the 19th century at a quarter of a million.

Nomad technology: I don’t think it’s a secret that the biggest problem when sleeping on the ground in cold weather is keeping the ground chill away. The kind of rugs Bertram’s talking about won’t do the job, won’t provide a warm mattress for high altitude nomads - and are too heavy to haul around. Lighter, warmer and cheaper to make are felts. There is plenty of information on Shahsavan use of felts to sleep on, often backed by jajim or gilim. To carry this train of thought a bit further, there are pile rugs used by transhumants and nomads, but they are not the same as most Caucasian and Azarbayjani pile rugs. Good examples of pile rugs used by nomads are the ivory and dark brown “sleeping fleeces” woven at one point by the Beni Ourain in the Middle Atlas. Maybe six knots to the sq. inch, lots of thin wefts and warps, light weight, very long pile, very warm, very much like a sheep skin.

As an aside, Ali Hassouri told me that when he was a child growing up in Bostanbad (a town in eastern Azarbayjan), he sat on coarse, long-piled goathair gabbeh-like floor rugs. Villagers in the region still sit on locally woven rugs if they can afford them. The rugs used domestically in eastern Azarbayjan have a standard format.

“Wild nomads”: Yes, sitting on a mountaintop at about 2500 meters, looking out at Shahsavan ahlechik arrayed on slopes above the cloud line is impressive. Actually, there is a powerful connection between today’s mountaintop Shahsavan and their ancestors that is hard to miss - and the view alone is well worth the trip. (See the Frauenknecht article in a 1981(?) Hali about visiting Azarbayjani nomads.)

Earlier fieldwork: The problem is, there really isn’t any to speak of. That’s why Richard Tapper’s large body of work is so important. Yes, he’s an anthropologist. You have a bone to pick with how he defines himself, take it up with him. He’s quite approachable. And yes, he would best be described as a cultural anthropologist.

Rich nomads: Sure there are/were rich nomads and they weave/wove (or had woven) high quality textiles. Witness the amount of silk in nomad woven artifacts.

Posted by Steve Price on 02-04-2003 08:53 PM:

Hi Mike,

Thanks for the clarifications. One more question: your argument against Shahsevan weaving of pile seems to focus on the properties of rugs as places to sleep. But pile can be used in lots of things besides sleeping rugs, easily exemplified in so much of the output of Turkmen, Belouch and many other nomadic tribal peoples. Is there some reason why the Shahsevan could not or did not make pile stuff? Or have I missed the point of the debate altogether?


Steve Price

Posted by Mike Tschebull on 02-04-2003 09:18 PM:

Not only places to sleep, but places to sit.

It seems that the Azarbayjani nomads specialized above all other west Asian nomads in the use of "sumak" for transport bags. Why is that? Clearly, the best Shahsavan transport bags - best designs, best color - are sumak; many/most of the pile bags - like Wendel's - are derivative, and may be village weaving.

It must be kept in mind that a lot of pile-woven bags were woven for the bazaar, and ultimately, Western consumption.

To paraphrase John Cage, it troubles me to see how much Turkman weaving is pile.

Posted by Steve Price on 02-04-2003 09:55 PM:

Hi Mike,

You say, troubles me to see how much Turkman weaving is pile. I don't want to put words in your mouth, but is it correct to interpret this as meaning that you are skeptical about the general notion that most Turkmen pile bags and trappings were made for use within the tribal community, at least until 1875 or so?

Steve Price

Posted by R. John Howe on 02-05-2003 06:29 AM:

Mike -

And to add to Steve's question, "And if so, where are all the Turkmen flatwoven items that are presumably the things really "made for use?"


R. John Howe

Posted by Mike Tschebull on 02-05-2003 07:40 AM:

John, Steve - These are indeed good questions for Turkmen collectors to ponder. Compare the corpus of Turkmen weaving with the variety of wool and goathair woven textiles from, say, Morocco and, well, Fars in Iran, and explain the lack of structural variety.

Posted by Bertram Frauenknecht on 02-05-2003 08:20 AM:

Hi everybody, hi Mike (or should I say Herr Tschebull?),
I doubt that Shasavan means nomad to any specialist. Even Tapper would disagree with you. And they are not only in Azerbaijan, but spread all over Persia since Nadir Shah distributed them, as they had become so numerous and lazy that they did not help against the Afghans in 1722. That was the end of the Safavids.
For example (fieldwork!) there is a rather big group in the Hashtrud area and further east, all villagers who called themselve still Shasavan in the seventies when I visited several of their villages.
Their production then was still slightly related to what we are really talking about, those wonderful Soumac weaves from the 19th century or earlier and their beautiful rugs.
A tribe is always also partly settled. Or how could we explain that Soraya had a German mother and her father a University degree
from Europe (Bachtiar). The palaces of the leaders are still existing.
I know that it is beautiful up there on the Savalan range. In summer! When it has 110 degrees in the valleys and the ground is breaking up. In 2500m it is still quite warm.
I visited them last in early March in the southern Moghan, this tiny piece that is left to them which is the reason that there are only few left. It was cold! Still I remember this little kid playing on the ground half naked.
Your head count from 1870 means exactly this small part of the Moghan. Don't you know that till 1828 they were occupying the whole fertile ground of the Moghan steppe? Spreading up the Kura , moving into Southern Shirvan? The Khans of Baku, Shusha and Gendje were with the Shasavan confederacy off and on.
We are talking of a group of people who were obliged to keep an army of ca 20000 under arms. Who were even in the thirtie's of the 20th c. able to block the connection between Täbriz and Tehran for months.
These wonderful bags they produced were not for use. They were money! And not for export, as most were made before export played any role.
I think the biggest problem is that we simply cannot put our finger down and say this rug is Shasavan, as we can do with Gashgai for example. The reason? There high culture ended in the middle of the 19th c.
All the travellers from the end of the 19th call them robbers, bandits, dangerous.
By the way, I consider Hahns estimate also far too high. My point was that nomads from the southern Moghan speaking a dialect that I know (?) were selling their rugs at the market in Tiflis in 1890 and I doubt that they came from Persia. It is a fact that nomads, if they were contollable, were considered as important meat producers and whatever else they came up with. That's why I call them entrepreneurs. The Russians needed food too.
And last, have you ever carried one of those felts covering the tents? I rather carry five nice carpets. As more beautiful they are as lighter they get.


Posted by Mike Tschebull on 02-05-2003 09:36 AM:

>Herr F. is making a near perfect argument for the use of the term "Azarbayjani nomads".

>With regard to the numbers of "Shahsavan", to paraphrase Sgt. Joe Friday, just stick to the facts.

>Roof felts are of course heavy, but that's why on a typical ahlechik, the felts are made of 27 triangular pieces, sewn together to make three overlapping shaped covers. The roof cover is the difference between life and death for these nomads. Lots have died from exposure. The felt roof covers are a good example of this "technology of nomadism". Heavy? Sure, but great insulators, which is my point regarding their use as ground covers.

Posted by Bertram Frauenknecht on 02-05-2003 10:20 AM:

Hi Mike,
Azerbaijani nomads have been by large in the 18th and a good part of the 19th century groups that were part of the Shasavan conferacy. Please read Tapper more carefully.
Your point was that they did not make pile rugs for the weight.
And now you agree that their felts are very heavy.
Do you know why many people and especially their flock died?
The southern moghan, where they had to go to is only for a small part plain and fertile (there is a huge irrigation program now for
agriculture). In the adjecent hills there the temperature is about 2-3 degrees celsius less(ca 4-6°F) than in the plain. That proved to be enough.
Again, a visit in the 90's does not say ANYTHING about these people 150 years ago.
Claiming they never made pile rugs is simply absurd. You have no argument against it. They were mostly Turkmen by origine. Not to forget Kurds. And their tradition was rugs, rugs and again rugs.

Another question for you. Why is it that from approx. the middle of the 19th c. there were plenty of weavers to produce a large number of Herizes, Baghsheichs, Serapis?

A highly educated Professor in Tiflis might not have been able to count right, why should a Sgt. Friday (maybe it was already Saturday and he wanted to enjoy his weekend )

With nomadic greetings

Posted by Filiberto Boncompagni on 02-05-2003 10:57 AM:

I think Bertram is right.
"It is a fact that nomads, if they were controllable, were considered as important meat producers and whatever else they came up with. That's why I call them entrepreneurs."

That was what they were. They produced meat, wool and textiles. Part of their production was for their own consumption, the rest was for trade!
I think we should abandon the romantic idea that "tribal" textiles were made exclusively for the tribe’s own use.
That doesn’t mean they were produced for the western market - I bet there has been ALWAYS a market for this stuff and the trade is as old as the nomadic weaving tradition.


Posted by M. Wendorf on 02-05-2003 11:50 AM:

Their tradition

Dear Folks:

I think their "tradition" first and second was simple structures and flatweaves, but I also think they wove knotted pile rugs.

I think the "they" is anyone who calls or thinks of themself as Shahsavan. I have heard/known people who called themselves Shahsavan. I have never known or heard of anyone who called themself an Azerbayjani nomad and I think the distinction Mike makes is either too sharp or too fine.

One anecdote. Some years ago I happened into a small, dingy rug shop in Middleburg, Virginia. Striking up a conversation with the seemingly Persian owner he learned that I was interested in old weavings from northwestern Iran. At this point he volunteered that he was not really Persian, but "a Shahsavan man." Is he Shahsavan, Persian, Iranian, American? Well, he thinks of himself as Shahsavan. He then proceeded to show me several old 20th century pile village rugs that he said were Shahsavan. To him these were clearly unmistakeable. When I told him that there is a debate about whether the Shahsavan knotted pile rugs or not, he simply shook his head in disbelief.

Best, michael wendorf

Posted by Mike Tschebull on 02-05-2003 02:25 PM:

Filiberto - There's no disputing what you write there, but what you write hasn't been in dispute, so there's nothing to be "right" about.

If you accept Herr F's definition of "Shahsavan", you render it practically useless, as it will describe villagers and nomads in many areas of western Iran w/o discrimination. That's why "Azarbayjani nomads" makes sense. Of course, nomads in western Iran wouldn't call themselves "Azarbayjani nomads". It's an anthropoligical term - hey, cultural anthropological term.

Read Tapper for definitions.

I wouldn't put too much stock in those Virginian Shahsavans.

Posted by M. Wendorf on 02-05-2003 02:49 PM:

if a then b?


(1) Shahsavan necessarily refers to or means an Azerbayjani nomad even though not all Azerbayjani nomads are Shahsavan.

(2) People who think of themselves as Shahsavan are not now and never have been Shahsavan if they no longer are Azerbayjani nomads and regardless of any family history during which their ancestors were or may have been Azerbayjani nomads.

(3) People who may have been Azerbayjani nomads and Shahsavan at one time but are now settled are no longer Shahsavan because they became settled.

(4) People who consider themselves to be Shahsavan and weave knotted pile carpets are not Shahsavan because they weave knotted pile carpets and/or are settled.

Are any of these four statements incorrect as applied to your argument regarding the weaving of knotted pile carpets by the Shahsavan is concerned?

Best, michael wendorf

Posted by R. John Howe on 02-05-2003 03:41 PM:

Mike -

Actually, it was Michael Wendorf who spoke last.

On the other hand, Murray Eiland, who often sounds like he's from Missouri with regard to evidence in the rug world, WOULD accept an indication by a person claiming to be Shahsavan that he/she was that and that the pieces he/she indicated were Shahsavan were that.

I think Murray's acceptance is based on the liklihood that most folks in this part of the world would not claim to be a member of a given tribal group lightly.

There may be some similar designations that might be avoided by some ("Sart" seems often to carry a prejorative meaning that might make some avoid it.)

What's your own experience in this regard? Do some folks you have talked to in the field give seemingly incorrect indications about the tribal groups to which they belong?


R. John Howe

Posted by Mike Tschebull on 02-05-2003 04:44 PM:

In Qaradagh, where I've had some experience, nomads I've met don't call themselves "Shahsavan", but rather by the name of the individual tribe they're part of. For example, I've spent time with Moghanlu families, and that's how they identify themselves. And by God, that's where the Tapper maps show they lived a hundred years ago. Census data shows that the Moghanlu lived in present day Armenia two hundred years ago.

Asking questions in the field is tricky. Ask things the wrong way, you get the wrong answer. Every professional in the field runs up against this. Ask the weaver if she is "Shahsavan", and she'll say yes, if only to a) please you, or b) to get rid of you. I addressed the asking of questions in the field at the last ICOC and my paper should appear in the next OCTS. I understand it should appear shortly. (What did the cat say as he placed his tail across the tracks before the 5:15 came through? It won't be long now.)

We ruggies play fast and loose with facts/reliable field information vs. conjecture; we should be more careful.

Posted by Wendel_Swan on 02-05-2003 05:13 PM:

To begin, I had exactly the same experience in Middleburg as did Michael. The pieces that I saw were comparatively late (20th Century) but shared some designs and motifs that we would recognize from what we collectively call Shahsavan flatweaves. I would not have identified any as Shahsavan myself had I been doing the attributions. But this man was very positive and showed me articles about his family and their connections to Persia and the Shahsavan.

Depending upon the era, the Shahsavan have been either more or less politically, militarily or strategically important and, during briefer spans, either more or less wealthy.

One cannot overlook the relationship between the nomads and the villages. Nomadic pastoralists do not live in isolation. As important entrepreneurs, as Bertram correctly points out, they had to sell their herds (or flocks) to and in the villages. Likewise, they needed supplies that they could only purchase or barter for in the villages.

During difficult economic times, a nomadic individual or family might become settled, for pastoralism was as much an occupation as a lifestyle.

As both Bertram and Filiberto have alluded to, trying to determine whether an individual piece was produced by a weaver who was at that moment sedentary or nomadic is virtually impossible. While room size carpets would have been produced in villages or towns, so could have horse trappings, khorjin and mafrash.

Jijims and felts are objects likely made by nomads while nomadic for their own use, but a weaver would not spend all her time on those projects. However, I understand that some felts are (or were) even produced in villages. If the idea is to weave things which could be bartered or sold, then a khorjin produced by a nomadic Shahsavan is unlikely to be much different from one produced by a sedentary Shahsavan.


Posted by Mike Tschebull on 02-05-2003 06:35 PM:

>Felts were usually made for the nomads by villagers who visited summer camps.

>I think jajim and gilim were made by villagers in Azarbayjan in much greater numbers than by nomads, if only because of the larger number of village weavers. There is plenty of evidence of (old) village flatweaves, both warp and weft-faced, in houses. I've seen, especially in Qarajeh, killer jajim, tucked away because they were made by great grandmama.

>There are many Azarbayjani pile kennereh, pushti, and kelleh which can be clearly identified as coming from specific villages (or groups of villages) which have been in existance for two hundred years. Diligent (and time consuming) field research could probably identify more villages as sources of pile rugs.

>It seems unlikely that villagers would weave mafrash, as they pile their bedding in the corner of a room and cover it with a jajim. A mafrash is used to pack away nomad bedding during the day, and serves as a backrest on the inside wall of a kume or ahlachik. I've never seen mafrash in a village house, but there may have been some at one point earlier in the 20th century, left over from the nomad life. Other items, like camel trappings and packbands, would only have been woven by nomads.

>I've never seen any sumak bags in village houses in eastern Azarbayjan, and have assumed they were woven only by nomads. The large number of sumak mafrash seems to butress the idea that sumak is a nomad weave.

Posted by Filiberto Boncompagni on 02-06-2003 04:51 AM:


I wasn’t disputing anything you said, and I didn’t want to intervene in the Shahsavan diatribe.

I only wanted to stress that MOST LIKELY tribal weaving has ALWAYS had two destinations: one for family consumption, the other for the "bazaar". The weaving for export to the West is just a more recent development of the latter.
I used the word "tribal" instead of "nomad" because even if the nomads are settled, they still consider themselves as belonging to the tribe. The Middle East - I can see it especially here in Jordan - is still largely a tribal society.

Let’s remember, too, that the boundary between settled and nomad was never clear-cut. There is an history of forced settlement of tribes by the rulers followed by rebellions and return to "nomadism". Not to mention that some tribes conduct also a semi-nomadic (seasonal transhumance) way of life.


Posted by Bertram Frauenknecht on 02-06-2003 04:56 AM:

the discussion is about pile and flatweave or only flatweave amongst the Shasavan.
We know a lot about nomads and settled nomads, but also very little for sure, as serious research (more or less) is only done in the last 20 - 30 years. Apart for very few people who know everything, most of us go with the fact that their is little written about nomadic production in former times.
We know, no nomadic tribe in the countries of concern produced only flat or pile (except for the Shasavan as Mike claims).
All these tribes continue a healthy life with nearly unbroken traditions, except now they drive a truck or a motorcycle and produce mostly for a market. Many of them are more or less settled or partly settled, as the herds have to be moved around for food.
They all have their trade marks, visible and easily recognizable by most of us.
The Shasavan basically stopped existing in about 1850 to 1860. By far most of the research by ethnographers is done a lot later and does not convey anything about nomadic production. Tapper in his book 'Pasture and Politics' never mentions a word about textiles of any sort.
But we can see rugs that show clearly designs we know very well from Soumac bags and nobody doubts them to be Shasavan.
Often these rugs have a different weave than any of the known and accepted (again more or less) categories.
Would it be possible that everyone who owns a 'Shasavan' rug
comes up with structure analysis and we compare the results and maybe define what a 'Shasavan' rug is?
I invite everybody to send images of front and back together with an anlysis to me. I'll categorize the material and come back with the result.
I put up here two pile bags which I am convinced to be Shasavan.
I hope my computer knowhow allows me to do that.

Hope it works.


Posted by R. John Howe on 02-06-2003 06:52 AM:

Herr Frauenknecht (we Americans do not really understand the difference between "vous" and "tu" but I will try)

We have examined some pile pieces here on Turkotek, thought by some to be Shahsavan.

In Salon 68 in our archives, Daniel Daniel Deschuyteneer led a discussion of the glorious "Italian Rug."

Here is the lead image of it.

Daniel's initial essay included a technical analysis of this rug.

And earlier, in a salon on a rug morning at the TM by Wendel, he presented another "Shahsavan" pile piece which he indicated had "distinctive structural features."

Here is the link to that salon

And here is a detail of the rug.

In a still earlier salon, Wendel offered a "yellow ground rug."

Here is his lead image of it.

Again, he provided a technical analysis.

And finally, in an early salon that I hosted, this piece occurred.

I think this too is a "Shahsavan" pile piece that Wendel owns.

There are undoubtedly others that we've treated from time to time, but I wanted you to see that there is a fairly accessible cache of pile weavings thought by some to be Shahsavan on which you could collect the data you propose.


R. John Howe

Posted by Steve Price on 02-06-2003 07:31 AM:

Hi People,

It seems to me that much of the debate is hindered by the fact that the term, "Shahsavan", means different things in different contexts. The Virginia dealer who refers to himself as Shahsavan is correct, just as I would be if I referred to myself as Russian (my father was born there). But what if we began talking about Russian traditions, using my immediate family and experiences as the basis? I think we'd get bogged down pretty fast.

The NW Iran/Azerbaijan distinction is similar. We are accustomed to thinking of those as two different places because they are two different places today. But until fairly late in the 19th century, Azerbaijan included NW Iran, and Tabriz was the capitol of Azerbaijan. Would anyone around here call a Tabriz carpet woven in 1850, a Caucasian carpet? It is as much a Caucasian carpet as the guy who owns the shop in Virginia is a Shahsavan, or as much as I am a Russian.

If the debate about whether Shahsavan wove pile stuff in the period that's interesting to most rug collectors - the 19th century - is to get anywhere, I think the combatants have to agree on who they mean when they say "Shahsavan" in that period, and how they identify a Shahsavan product. Defining it as something that must be flatwoven is pretty unsatisfactory, since it assumes the result of the debate rather than pavng a pathway toward it. Any argument can be resolved by simply defining the outcome in the premises. It works every time, but doesn't advance our understanding.

Finally, I point out in passing that Tanavoli's beautiful book, Shahsavan, appears to include practically every pastoral group between Armenia and Bijar as Shahsavan.


Steve Price

Posted by Bertram Frauenknecht on 02-06-2003 08:07 AM:

Hi Steve,
thank you. I always thought somebody must come up with Tanavoli's wonderful book.
To add to your Russian origin I have something too.
I am Frankonian(grew up here)- that's the tribe of Charlemagne.
Living in Bavaria, that's the State of a tribe which became stronger through the path of history. Living in Germany (exists only since 1871), which was occupied by Roman legionnaires,
Huns, French, and who knows. Am I now a frankonian German or
a roman Bavarian with Frankonian origin?
Oh, I forgot, I have French Hugenot and Jewish ancestors. That makes me a ( forget the Frankonian, my grandfathers moved here), Roman-French Bavarian Jew with a German passport
I should be wearing Lederhosen with a black hat.
So much for tribal origine.

Have a good day

Posted by Lloyd_Kannenberg on 02-06-2003 08:28 AM:

Hello everyone -

Perhaps someone can help me out. Plate 14 in the current NERS on-line prayer rug exhibit - my rug - shows a Shahsavan beetle (the accompanying pop-up is the same as Plate 71 of Wertime's "Sumak Bags") and is a pile rug. But is it a

SHAHSAVAN pile rug?

Lloyd Kannenberg

Note: I've added the image of the rug for convenience, and will remove it promptly if Lloyd Kannenberg or any representative of the New England Rug Society objects to it being displayed here. Steve Price

Posted by Mike Tschebull on 02-06-2003 05:45 PM:

German tribes

My wife is also from Franken, but her forbears came from the Sudetengau, which is now part of the Czech Republic, and was formerly part of the Hapsburg (Austrian) Empire. It's hard to pin a label on her, too.

Compared to the US, the German-speaking part of Europe is really quite tribal, with many citizens able to cite the Germanic tribe from which they decend.

Posted by Filiberto Boncompagni on 02-07-2003 02:25 AM:

I’m glad to see you barbarians still remember your origin!

But the tribal way of life is not a mere reminiscence: It’s a social organization alive and kicking.

The Bedouin tribal structure here is composed by clans (or quawm) headed by elders. A number of these clans make up a tribe, o or qabila headed by the sheikh. The tribes have enough power to oppose to the central government, as it happened recently in the village of Ma’an when the Police tried to capture the alleged killers of that American diplomat…
This structure still exists among the Bedouins of the Arabic Peninsula.
I do not know what kind of "tribality" is left in Caucasus/Persia, but surely it’s not only about labels. It’s about family and blood ties.

Now, changing of subject, I hope somebody could answer to Mr. Kannenberg…


Posted by Bertram Frauenknecht on 02-07-2003 06:33 AM:

Hello you all,
there is no need to differentiate between you and you ( or vous and tu or Du and Sie), let's just have it the american way.
The rug of Mr. Kannenberg or may I say Lloyd looks like a late version of a Shasavan rug. Border and field design can be found on bags. The border drawing is known from Anatolia and from reverse soumacs. The colors are chemical, at least what I can see on my monitor. This means the piece is made some time after 1890. The outer border I've seen on a number of late Caucasians.
As I'm stuck at home (too much snow) I don't have acces to my books, but I'm sure somebody will be able to show this border.
Look in Fokker, Gregorian or Kerimov. The weave looks like a possible Shasavan weave, maybe made by people who stayed in Russia after the closure of the border.
My guess southern Shirvan.
There must be several different structures, as there are a number of different tribes (or families) who formed the confederacy.


Posted by Tracy Davis on 02-07-2003 09:47 AM:

Filiberto, is the reason you can refer collectively to "you barbarians" because your people were ruling the world from Rome when the rest of us were running around in loincloths and string skirts? You're such a snob.

I've been following this discussion with interest, as it illustrates the problems we have with scholarship. For example, the beetle device we all "know" is Shahsavan..... do we really know that? Is there a piece with 100% known provenance (collected in the field from the weaver or her immediate family, verifiably dated), that has this device? And given the fact of design migration, and the use of talismanic symbols, why are we so sure that this beetle device means the piece was woven by a Shahsavan?

I wonder if this isn't an example of how theories are advanced, develop a critical mass of acceptance, and become ruggie canon and cited as evidence from that point forward.

I'm not really trying to make the case that the above is happening in this discussion; I'm just wondering how much we really *know* that conforms to currently accepted standards of scholarship in the soft sciences.

Posted by Steve Price on 02-07-2003 10:09 AM:

Hi Tracy,

I believe you point to the most serious obstacle in trying to assess what we know of rugs in a scholarly way - epistemology (how do we know what we think we know?). I believe that the entire process of attribution is, ultimately, founded on similarity in designs, motifs and layouts, and hosted a Salon on my take on structure-based attribution awhile ago, at


Steve Price

Posted by Mike Tschebull on 02-07-2003 10:14 AM:

Lloyd's rug

Tracy's right that we have no real basis for assigning that "bug" sumak design to any specific group of weavers. It has even been speculated by reasonably responsible Iranian dealers with field experience that the very fine "bug" and "cloud collar" sumak bags were woven under some type of workshop ("factory") regime. So, trying to assign a rug such as Lloyd's to specific settled nomads, whatever they call themselves, has a very strong chance of being wrong. Trying to sort out Caucasian rugs as to origin is very difficult due to the lack of information available.

Isn't it better not to guess, especially when the chance of being wrong is huge? Guessing wrong, creating "facts", means the right answer is still out there, and you are not aware of it.

Creating a structural definition for "Shahsavan" pile rugs would put more bogus "facts" out there.

Posted by Filiberto Boncompagni on 02-07-2003 11:40 AM:

Exactly, Tracy.

Posted by Bertram Frauenknecht on 02-07-2003 12:44 PM:

In 1981 I published a Shasavan mafrash panel in Hali (Vol IV, No 2
page 26) with the design in question. It was bought from a reasonably responsible Iranian dealer who at the time was one of the main sources for high end Shasavan pieces amongst other stuff. (John Wertime knows him well too) At the time nobody had any doubt about pile material from those people.
Bogus facts seem to be all over the carpet world. Or does any one have any proof that Kazaks were made in the places they are named for?
I think what we need is some sort of working definition like Ushak
or Zeychur or 'Shasavan' for pieces that fit into certain categories and where we can say with high probability that they are what we call them. Very often it will be hear-say as it is simply impossible to find a 100% proof. But at least at the end we all have an idea what somebody means when he uses a certain attribution.
I place Lloyd's rug through looking at the structure, not only the design. And the structure is similar to pieces that I'd call Shasavan. Aother "expert" calls the piece Anatolian.
My limited knowledge only allows me to guess. That is for me still better than insisting on an assertion that is lacking every probability.
As I said earlier, I've been in Shasavan country several times and spoke with these people when they had not yet been a tourist attraction. And I've studied the map while I studied their history.

Posted by Tracy Davis on 02-07-2003 01:16 PM:

Okay, Steve, I went back and read your old Salon re: the foundation of structural attribution. I see your point.... but I'm not convinced that structural attributions developed out of palette- or design-based attributions. When Bogolubov et al were running around plundering the Turkmen and amassing respectable collections, they were about as close to the source as possible, and their labels and attributions should be taken seriously. It's not a big stretch from there to noticing that the Tekke pieces share structural commonalities, etc.

Maybe it's a little more dicey relying on bazaar info from southern Iran for initial labels and attributions ... but maybe not. Rug dealers couldn't *all* have been ignoramuses (ignoramusi?) who gave labels based solely on design or colors. Sometimes they actually knew something. The problem is, among other things, sorting out the wheat from the chaff.

Because apparently the human species (or maybe it's just westerners?) has a need to label things *before* they can define them, or refine the definition, I propose that we come up with a good label that we can give to weavings with nomadic characteristics that don't fit well into a category based on structural characteristics..... like, gee, maybe "Kurdish"?

Posted by Steve Price on 02-07-2003 01:40 PM:

Hi Tracy,

You're quite right about taking attributions seriously when they were made by people who were at the source a century ago - but the Bogolyubov and Rickmers collections are the rare exceptions. If they were the norm, life would be simpler.

As for attributions originating in the bazaars (which is where most probably did originate), I don't think it implies that the dealers were idiots. The reality is that collectors generally want an attribution of time and place, and usually of function as well if the item isn't a carpet, and tend to patronize dealers who provide them with attribution. That is, the dealer who gives his customer a confident-sounding attribution may do so, not out of ignorance, but to promote his wares. The attributions that wind up being published do take on an authority (just browse our archives and see how often attributions are made on the basis of similarity in appearance to some picture in a book or magazine, with the assumption that the published attribution is correct). And many (most?) of the books about rugs were written by dealers. Some, of course, were/are very well informed. We have at least two such people debating the matter at hand right here (for anyone who doesn't know it, Bertram and Mike are both dealers, each has written books and articles, and each has done field work among tribal peoples in western Asia).

I'm not offended that you don't buy into my thinking about the epistemology of rug attribution. I know that I represent a fairly small school of thought on that subject, and it's OK. I think multiple views are valuable - they keep us from getting too firmly convinced of the correctness of our misconceptions.


Steve Price

Posted by M. Wendorf on 02-07-2003 01:56 PM:

attribution and labels

Dear Folks:

The Kurdish label is already taken so "weavings with nomadic characteristics that don't fit well into a category based on structural characteristics" will have to be called something else.

I also think that color or palette plays an important role in attribution. There may be a role for design too, I would just put it last. That is after structure, wool and after color - including palette, value, saturation etc.

Now, please carry on.

Best, michael

Posted by Bertram Frauenknecht on 02-07-2003 02:35 PM:

How about 'BOSCO'
Based On Structural Characteristics.
The last O just added to make it sound nicer

Great idea, Michael


Posted by Tracy Davis on 02-07-2003 03:17 PM:

Based On Structural Characteristics Or...

Wasn't there an old joke going around about "Bilmem" rugs? Turns out that's what the dealer said when asked what certain rugs were called ... "bilmem" ("I don't know")

Michael, I was trying to be humorous with the "Kurdish" suggestion, but there's a grain of truth in there. It seems to be the preferred label given by dealers when they're really not sure what something is.

Posted by M. Wendorf on 02-07-2003 04:02 PM:


Dear Tracy:

I like the Bilmem label.

There may be a grain of truth to the use of the Kurdish label by some dealers. But that grain is limited to the fact that some dealers use it rather than Bilmem. There is no grain of truth to the implication that Kurdish rugs have neither any structural nor other characteristics that give them identity. This is a myth that should not be perpetuated.

Mike T. got it half right, we should not play fast and loose with the facts or create bogus facts. Bertram got the other half, we should work as hard as we can to create a plausible framework with which to study and understand rugs tempered by the humility of knowing we can never know it all. Bilmem.

Bertram: Bosc would be a pear. Maybe Bosco will stick.

Regards, michael

Posted by Mike Tschebull on 02-07-2003 06:04 PM:


Most dealers have bought rugs and flatweaves in the bazaars in big cities, be it Tehran, Istanbul, or Bukhara, to pick a few. It takes less time, is less dangerous, costs less. The attribution is less dependable, as the piece in question has gone through more hands.

Spending extended time in remote areas of countries where "rugs" are woven is expensive, time consuming, sometimes a bit dangerous, hard to access, not guaranteed to be financially lucrative, and often stressful. That's why few even make the attempt to do it. I've never seen tourists in Qaradagh, and the locals gawk at Tehranis, let alone Feranghi. I once saw a movie crew at about 3000 meters, though, with the director's assistant dressed in tribal costume while holding a clipboard and a cellphone.

Posted by Bertram Frauenknecht on 02-08-2003 10:11 AM:

it might be sad, but I know of two organizers here in Germany.
One leaflet I remember, it said approx.:
visit to the Shasavan nomads, sit in their tents, a day with real nomads.
And these trips cost a lot of money too.

May I also bring back to your mind the article of Tanavoli in Hali about Shasavan pile pieces. I can give more info about it tomorrow.
Goodbye nomadic world

Have a nice weekend


Posted by Mike Tschebull on 02-08-2003 04:43 PM:

mtn tops

If movie crews can get up on mtn tops, why not those legendery German tourists? However, no tourists are gonna go very far into nomad land.

I note with interest that the ad uses "Shahsavan" and "nomads" in the same breath.

Caucasian rugs used to be called to be called "Cabistans", Turkmen rugs, "Bokharas". Could "Shahsavan rugs" be next?

Posted by Richard Tomlinson on 02-09-2003 02:58 AM:


interesting thread. seeing as we are talking about structure and dyes, i have 2 relatively simple questions for the experts;

1. my understanding is that cotton warps and wefts usually indicate a later dating (usually 20C) BUT i have read here and there that shahsavan weavings (and i am referring to smaller pieces here - not rugs) from the 19C ?often? have cotton warps and wefts. TRUE or FALSE?

2. did the shahsavan (during the 19C) EVER use cochineal?


richard tomlinson

Posted by Bertram Frauenknecht on 02-09-2003 05:03 PM:

it is a few years ago that a dealer from Baku offered me two flatwoven bag faces ( mixture of simple flatweave- kilim type- and brocade) with cotton warps. I don't remember the wefts. He claimed that these pieces are called Shasavan, although he could not combine much with it. He said that they came from people in the southern Shirvan. They looked 19th c., no artificial dyes.
And I am sure they used cochenille, not the Spanish brand, but from a louse that was available in the area.
The article of Professor Tanavoli is in Hali 45, page30.
Since these ads are done for tourists, they certainly stay at the lowest level, Mike.
Cabistans were a sign for bilmem. That a group of inventive dealers came up with nice names is not the fault of the Shasavan group. They were not around to say, 'hi, these are our rugs'

Posted by Mike Tschebull on 02-09-2003 05:32 PM:


>Some what seem to be old mafrash from the area below Sarab have cotton warps. Cotton pattern weft is not uncommon in old sumak bags, nor is cotton ground weft. Some of those may be Baghdadi Shahsavan work, but some with cotton ground weft are from Moghan. I have a Sarab rug with cotton wefts and a clear woven-in date the equivalent of 1840, so one can safely assume cotton was available to nearby nomads by that date. Nomads are said in the past to have come down to Sarab to trade.

>Fred Mushkat had a very finely woven "Shahsavan" packband with cochineal as the only red dye in his bands exhibition at ACOR 6