Joel Greifinger
November 18th, 2013 12:21 AM

It's a Kurd, It's a Flame, It's...Sauj Bulagh?
A couple of fairly recent rug additions have revved up a quandary that I've had for a while: what designates the boundaries of the widely-assigned category, Sauj Bulagh?

In 2009, Rippon-Boswell had this description in their catalogue:
“Sauj Bulag” is now an established generic term for Kurdish carpets made around Lake Urmia; the name derives from a town where such pieces, made in the surrounding area, were sold in the bazaar... Sauj Bulags are considered beautiful Kurdish pieces due to their brilliant colours (and often) the corroded brown ground of the field."

William Eagleton writes, "The illusive Sauj Bulaq (pronounced 'Sa Blah') rugs of the late 19th and 20th century, usually with reddish wefts and deep pile colours, should be placed in this (i.e. "Western Mountain") category since they were woven in the mountains near Sauj Bulaq, modern Mahabad. However, rug books give such a variety of conflicting data on the structure and designs of Sauj Bulaq rugs..."

At an exhibit of Kurdish weaving in 2000 where six Sauj Bulaghs were on display, the description of these read: "United by a common palette and secondary border treatments, the field patterns are all different." and a leading New York dealer, acknowledging the wide range of field designs on rugs they designate as Sauj Bulagh comments that nonetheless, "the particular range of colors – autumnal oranges with soft blues, greens, and aubergine is now recognized a distinctive Kurdish palette of the Sauj Bulag region in Northwest Iran."

With all of the variation, there are elements that are common to all of these rugs. Like other Kurdish rugs, they are woven with the symmetrical knot. The knotting is generally coarse to moderate and the back of the rug is generally flat or flat-ish. While the foundation of some of the earliest rugs of the type are cotton, from the beginning of the 19th century on they are wool. Wefts are often (but not always) a light red color through all or some of the rug.

With all of the variety of field motifs, there are nonetheless a few that appear characteristic of many rugs designated as Sauj Bulagh. Most identified with this group is the so-called "flaming palmette". However, stepped diamond, "tuning fork", floral lattice and a lozenge design that Burns calls ashlik (as opposed to ashik?) are also clearly associated.

Probably the most indicative are two secondary border treatments: particular versions of medachyl (running dog) and vine meander designs.

Now, the first of my additions just happens to have both types of secondary borders. Combined with the corroded brown in the field, light red wefts, a flat back and ashlik lozenge, this is surely what is widely termed Sauj Bulagh:

Design-wise, this rug has a great deal in common with Burns's "Ashlik" Sa'uj Bulagh rug (#49 in Antique Rugs of Kurdistan). In terms of range and intensity of color, well that's another story. The Burns rug has fifteen colors, mine a none-too-shabby twelve (gold, white, medium purple, turquoise, gray, orange, two reds, two browns and two blues). Here's the Burns:

And here's one with ashlik posted on Turkotek years back by Bob Kent. Notice the multiple vine meander borders:

Flaming palmettes are probably the most widely recognized indicator that a rug should be called Sauj Bulagh. Probably the best known of these is this Meyer-Muller Collection rug. If you look close, you can see the secondary vine meander on a blue ground:

A rare instance where you get the combination of flaming palmettes and both the running dog and vine meander borders, is in this rug:

These are some other flaming palmette models. The first has a common main border for the type:

While the corroded brown field is a strong Sauj Bulagh indicator, not all of the rugs that get this designation (by auction houses, widely recognized dealers or rug book authors) have brown as the background field color. Some have red, a few yellow. One larger group has blue, as with these:

Which brings me to my second recent addition. I call it "Wild Thing". It has many of the elements of rugs widely called Sauj Bulagh. The palette, with fourteen very saturated colors, is certainly of the type. The pile is long, very lustrous and extremely silky. It has flaming palmettes (wildly drawn, but quite recognizable) going in both directions. Both the 'exuberant' drawing and the bi-directional motifs are seen on some other Sauj Bulagh rugs. Some of the wefts are light red, but others are brown and still others are green. Neither the major or minor borders are particularly characteristic, though the major border is a funkier rendition of the motif found on both my other rug and the Burns rug.

Sauj Bulagh?

And here's a rug that was published in the Chicago Rug Society's Mideast Meets Midwest catalogue as possibly Kordi, that I now wonder if we'd call Sauj Bulagh. Notice it has the secondary vine meander border and the same motif in the major border as a number of the rugs above. The back is flat and it has thirteen colors:

What do you think? Are the last two rugs Sauj Bulagh, too?
Is all of this merely another instance of 'market category-creep' because "Sauj Bulagh" fetches higher prices than plain old "Northwest Persian Kurdish"?

Joel Greifinger

Marvin Amstey
November 18th, 2013 02:04 AM

'evening Joel,
I can't answer your question, but I admire your summary of the rug type. Thanks.

Rich Larkin
November 19th, 2013 12:01 AM

Hi Joel,

I'm with Marvin. Your summary of this topic is admirable, and I have no problem with it as such. However, I wonder where the underlying and possibly varied criteria came from that seemed abruptly to coalesce now and again into the elusive category. I would love a chance to get all those rugs you posted together in one pile and decide whether they all belonged under the same rubric. Furthermore, are all the "genuine" Sauj Bulagh rugs supposed to have a familial resemblance anyway? I'm referring to wool, palette, weave and texture, more than design. I think the dealers of the earlier part of the 20th century, and perhaps the late 19th, had a certain type of rug in mind when they used the term, and the question is, are the rugs that have lately carried the label of the type? I used to encounter a type of rug with long but disciplined pile (not so much shaggy), a lush and meaty handle, and very saturated colors that would have the Sauj Bulagh label on them. Maybe those dealers were making it up too.

A brief note on the pronunciation of the name. It has been a topic of interest for me for years. I put the "gh" consonant on the word above because the original spelling utilized a consonant (that does not exist in English) that is pronounced like a Parisian "r," but more voiced. I think we got into this subject a while back, and I don't intend to slog through all that again. However, I will mention that I asked an Iranian friend (and rug dealer) how it was written and pronounced. I was focusing on that final consonant, but to my surprise, he pronounced the first part of the place, "sovaj," as you would pronounce the word, "savage," if the first vowel were a short "o." "Sovaj Bulagh," he said. "Are you kidding " I said? "Nope," he answered.


P. S.: BTW, I'm not about to tell Eagleton how to pronounce Kurdish place names. Regional variation, I'm sure.

oel Greifinger
November 19th, 2013 04:29 AM

Hi Rich and Marvin,

As the Eagleton quote indicates, the earlier rug literature is all over the place when it comes to describing what was termed Sauj Bulagh (or Souj-Bulak or Suj Bulak). The authors from the first half of the twentieth century that specifically mention Sauj Bulagh rugs (e.g. Mumford, Dilley, Hawley, Edwards and later Jacobsen) provide, in sum, a rather incoherent description. There are a few common features cited: the wool is of very good quality; the colors are "deep", principally red, blue and brown. Hawley, whose description is the more elaborated adds "minor quantities of green, yellow and ivory." He adds that the weave at back is "slightly coarse grain" and the texture is "moderately loose" with pile of "medium length". Says it all, no?

The overall assessment varies from Mumford: "While strong and serviceable, the Souj-Bulak rugs are far from maintaining the standard of the first-class Kurd product." To Jacobsen: "Few rugs are more delightful than the old Suj Bulaks." Were they all referring to the same rugs? Who knows? 

At some point, the Sauj Bulagh discourse clearly shifted to focus on some of the criteria I mentioned in my post. While I don't have evidence for this, I imagine the publication of examples with structural details that were described as Sauj Bulagh in the earliest publications devoted to Kurdish weaving, Discoveries From Kurdish Looms, (1983) and even more Eagleton's An Introduction to Kurdish Rugs (1988) may have shaped the subsequent perception of the category. Here is the Herati pattern "Saujbulagh" from Discoveries listed as having two shots of red wefts, a loose handle and brown corrosion in the field. The colors are listed as: red, ivory, light green, light blue, orangish red, gold, brown and dark blue.

The rugs that Eagleton published bring us even closer to the contemporary prototypes. For the first time we see the addition of the secondary "running dog" border and a "tuning fork" field design in one and a variant of the vine meander that I described and a "flaming palmette" motif in the other. The palette on the latter contains the range of colors that have come to be associated (red, dark blue, burnt orange, white, medium blue, green, brown, blue-green) with the exception of purple. Interestingly, the purple has recently come to be seen as a marker for older examples.

Some of the most interesting discussion of the Sauj Bulagh category has taken place in the earlier years of Turkotek. A number of Salons devoted to Kurdish weaving took up some of these questions. Some of the highlights are: http://www.turkotek.com/salon_00007/salon.html, http://www.turkotek.com/salon_00007/messages/12.html, http://www.turkotek.com/salon_00092/salon.html

In the discussion of Daniel Deschuyteneer's 1998 Salon, Michael Wendorf wrote this useful short description, "Generally, Sauj Bulagh rugs have a flat back, symmetrical knots, ivory warps and red or coral wefts although I have seen some examples with yellow and even green wefting. Most telling is a corrosive ground, usually brown, that creates an embossed effect when the rug is in otherwise good condition. And the color, these rugs are all about color and the best are painterly in their use of it. Also, they almost all have some ivory in their design and many have the same minor border done on two different ground colors. The knots are very rounded and the wool is very thick staple. Persianate designs predominate. For example, the Herati pattern. This may suggest the rugs were woven by kurds for the local market. The rugs do not seem to be inherently "Kurdish" except for deeply saturated colors, lots of flowers and shiny or glossy wool." He further identifies the "running dog" (medachyl) minor border as the most common on these rugs.

Burns, in Antique Rugs of Kurdistan, argues for a continuous tradition of Sauj Bulagh rugs produced by Kurds going back to the seventeenth century or earlier in the form of "garden carpets". Of the nineteenth century tribal rugs that emerge from this tradition, he writes, "The tribal rugs - small in scale, long in pile, heavy in handle, and highly suitable for use in tents - were generally woven with traditional Kurdish designs rather than the patterns of the urban carpets. They frequently have two minor hooked borders, referred to in rug literature as the "running dog" or wave design...These tribal pieces have all-wool construction, red-dyed wefts, and a corrosive brown ground color." This emphasis on "traditional Kurdish designs" seems to contrast with Wendorf's comment on the predominance of Persianate ones.

Rich, you asked:

I think the dealers of the earlier part of the 20th century, and perhaps the late 19th, had a certain type of rug in mind when they used the term, and the question is, are the rugs that have lately carried the label of the type?
Given the conflicting accounts that the early twentieth century rug writers (who generally got it from the dealers) gave of Sauj Bulagh rugs, I'm not sure what "certain type" they could have had in mind.  What does seem clear, is that in the last twenty years or so, some rugs that seem to display some family resemblance based on structure, palette, wool quality, handle, field and border designs and particularly brown field corrosion have become established as a recognized (and valued) market category.


Joel Greifinger
November 20th, 2013 01:02 AM


In the previous post, I neglected to cite the most detailed Turkotek discussion of the Sauj Bulagh rug type, in which you were an active participant.


There, in response to your question:

The rubric, Sauj Bulagh, is one I have never understood very well in the context of rugs. I realize it is a town or village in Kurdistan, and I have read many general descriptions (not especially consistent) of the rugs supposedly produced there, but I do not recognize any particular rug type to be "Sauj Bulagh," as I do regarding, say, Bijar. Can you point me to sources where the type is well described or illustrated?
Michael Wendorf wrote:

There really should not be any confusion about the type of rug we are referring to when we call something Sauj Bulagh. The tendency to label this group of rugs as Sauj Bulagh as opposed to Mosul or something else seems to have its origins with a single carpet illustrated by A. U. Pope in 1917. The carpet published in Pope's "Catalogue of the Mrs. Phoebe A. Hearst Loan Collection" San Francisco 1917 pp 101 - 02 is described as having "deep resonant tones" and a "gamut of deep blues, purple, , glowing muffled red and soft yellow that was never surpassed as sheer colour at any time." Consistent with this description of color a relatively rare group of carpets began to emerge, but with little written about them until 1993 when Alberto Levi publihed an article in Issue 70 of Hali and documented several examples that fit the color description and also with consistent structural details. As it happened, a number of others including James Burns of Seattle, Brian Morehouse in Los Angeles, Michael Rothberg in San Francisco and myself were all aware of and looking for examples at the same time Levi was working on his article. Indeed an example from The Burns collection had created a certain buzz when it graced the cover of Hali Issue 62 in April 1992. It remains in the Burns collection and is plate 49 in his book Antique Rugs of Kurdistan. This particular rug is also distinctive because although it is otherwise structurally and color consistent with the group, it has a design consisting of ashik devices. Examples with more tribal and non-floral designs are much rarer than those in the more persianate, floral designs. My collection has only two such examples and the ashik device and its varients is one type and the so-called shikak device another.

These structural details are as follows. 2 ply ivory or ivory and tan wool warps; 2 red dyed wool wefts between each row of symmetrically knotted wool pile; a corrosive brown field that often creates an etched or embossed effect with the field design. The backs are flat. Long staple and very glossy wool with heavy, floppy handles. The best of these rugs have as many as 15 deeply saturated colors and designs that are typically floral and curvilinear.
The Burns rug on the cover of Hali #62 is the one in my initial post that he labels "Ashlik".

This type with shikaks:

and perhaps some with these common motifs on Kurdish pieces:

are among the types that Burns referred to as "traditional Kurdish designs".

In Levi's 1993 article in Hali that apparently contributed substantially to the contemporary use of Sauj Bulagh as an "established generic term"(Rippon-Boswell) he illustrated this rug as an example of the type. The major and minor borders look like they connect up with the ones on my "Wild Thing", no?


Patrick Weiler
November 23rd, 2013 06:54 PM

long winded reply
Here is my detailed, elaborate, extensive and informed addition to this complex issue:
Burns spells it Sa'uj Bulagh.

November 23rd, 2013 10:56 PM

Market Creep
Dear Joel,

I believe the first of the two rugs you ask about, the "Wild Thing" is most likely Iraqi Or Mosel Kurd in origin. Perhaps it was traded around Mosul - Kurds, of course would say that Mosul is in Kurdistan. It would not fall under any Sauj Bulagh definition or category known to me. I have seen a handful of these pieces over the past 20 years and a few have had banded warps. While there are some design similarities to one group of rugs possibly coming from around Sauj Bulagh, the more reliable details are structural and in this case we are dealing with very distinct weaving groups.

In my opinion, the Chicago rug you ask about would also not be considered a Sauj Bulagh. Although the warps and selvedges are generally consistent with the group, the wefts are not. Still, no one knows for certain. No rugs have been made or traded in or around Mahabad in many years. Eagleton found no evidence of weaving or anyone who could remember weaving done there when he visited. The rug on the previous page of the catalogue with the avshan pattern would fit the group and is a nice pair to consider next to the rug Levi used in his article.

The years have brought more pieces and more confusion with what you gently call market creep. A group of possibly tribal pieces such as the "this type with shikaks" above, which I think is one of mine by the way, with the distinctive latch hooks devices would probably fall in the group and is most probably Kurdish. That device is one that I and others consider distinctively Kurdish although there is a group of mafrash panels with such devices that seem Shahsavan. Confusing.

There is no evidence beyond the persianate patterning/design of some/many of these rugs to relate them to the 17th century garden carpets. It is wishful thinking without any evidence and you can infer that I would love to find the evidence. What Burns is propagating in his book is the theory articulated by Levi in Hali 70. Yes, it is possible, even likely that some Kurdish weavers may have been engaged with weaving the garden carpets. Does that mean there was a design continuum? I think you would need to see a continuous line of production and some structural connection. This really cannot be confirmed. Sorry to say.

Well, I hope that perhaps I have answered your questions at least to some extent.

With best wishes, Michael Wendorf

Joel Greifinger
November 24th, 2013 04:58 AM

Hi Pat,

As always, your contribution, though in this case a bit terse, cuts to the very heart of the issue.

As I wrote earlier, reading through the early commentators that mention Sauj Bulagh (or even Sa'uj Bulagh), you get the impression that they are not probably referring to the same set of rugs. It's hard to imagine the underwhelmed attitude of Mumford and the poetic effusiveness of Arthur Upham Pope being focussed on the same variety of rug. As much as there seems a virtual consensus of early rug writers that Sauj Bulagh rugs have, as G. Griffen Lewis says, the "best, soft" nap and a palette of "dark reds, blues and browns, also yellow and white", there is little more specificity, as Michael Wendorf suggests, until Pope presents Phoebe Hearst's Sauj Bulagh rug.

After a burst of explication of the design as reflective of the attempt to express infinity by the "Oriental" mind, Pope introduces the aspect that has become the central claim of subsequent Sauj Bulaghs: "Splendid as is the design, the color of this rug is its chief glory. Such deep reds, blues, and green have never been surpassed. To stand close up under this rug when the light is good is to discover a satisfaction and exhiliration in sheer color that few could have imagined possible. Nor is it fair to consider design and color apart. Indeed here, if anywhere, they 'blend confederate to one golden end - beauty.'"

Interestingly, Pope goes on to say that this example is different from the Sauj Bulaghs of his time ("not agreeing entirely with present-day examples"). It appears he is harkening back and appropriating the label for this type tied to some past models. This particular example seems to be in a line of rugs with a harshang pattern on a blue ground. The colors of the field (for which I only have a b&w photo) are "deep blue, red, gold, light blue, dark green and lavender." I think the Sauj Bulagh fragment gives a good indication. Like the fragment, the Hearst rug ties Sauj Bulagh to this version of the "running dog" border, an association that helps establish the way the category is currently constituted:

Meanwhile, back in the present: the Sauj (sorry Pat, Sa'uj :laughing_2:) Bulagh label has also become associated with a type of panel with the shikak motif. I haven't seen any rationale given for this attribution:

beyond the identification of the shikak as a traditional Kurdish motif that is used in rugs that would now be identified as Sauj Bulagh:

In the marketplace, including the major auction houses, a wide range of rugs, including those conforming more narrowly to mainstream Persianate design (as with this herati patterned piece listed at Sotheby) are labelled as Sauj Bulagh:

Categories change. Is this one useful?
And, even if we can't solve that question, doesn't anyone out there have some cool Kurdish rugs that they think might fit the Sauj Bulagh rubric that they could post to satisfy some of the desire out here to see more interesting and beautiful rugs appear on our screens?


Joel Greifinger
November 24th, 2013 03:39 PM


I hope that perhaps I have answered your questions at least to some extent.
Hi Michael,

Great to hear from you. I became aware of your post just as I was submitting the one above, so that doesn't reflect your helpful input.

Unfortunately, I have handled very few rugs from the areas of Kurdistan that are in Iraq, other than Jaff weavings from areas south and east of Mosul. This limits me to published examples for comparison, and there possibilities run almost dry. None of the Dizai, Girdi or Herki rugs pictured in Eagleton's book show much resemblance to my "Wild Thing" in palette or in their end treatments. Here's a picture of its twined heading cord:

This rug that Eagleton says may be from north of Erbil (that has similar, multiple overcast banded selvedges) seems to come closest:

More than the other features, it's the extremely diverse palette on the "Wild Thing" that doesn't seem a comfortable fit.

Would you say that the first rug I posted fits your criteria as Sauj Bulagh? It's the one with the ashik (ashlik?) motifs in the field, the "animal head" main border and "running dog" and vine meander minor borders:

Some years back, when I first started getting interested in rugs, and Kurdish rugs in particular, one of my first postings on Turkotek was a battered little Herati-pattern bagface. On that occasion, you graciously answered my questions and then, with a clear sense of amusement, commented upon the effort I was putting into that rather modest object. Now, some years on, I'm still at it. And, it's partially your fault.


Joel Greifinger
November 28th, 2013 07:48 PM


the "Wild Thing" is most likely Iraqi Or Mosel Kurd in origin. Perhaps it was traded around Mosul
Like many other rug designations, the term 'Mosul' has had more than its share of uses by different commentators at various points in time. It has fallen into disuse recently after some mid-century rug authors dismissed it as merely a misnomer for Hamadan area rugs (Edwards - rugs from villages around Hamadan "were once commonly called Mosuls...which has as much to do with Hamadan as it does with Hoboken." Jacobsen calls Mosuls, "the very poorest of all Hamadan rugs."

Other pictures and descriptions of Kurdish rugs from areas now in Iraqi Kurdistan don't offer much better sense of what a 'Mosul' might have been a century ago. The rugs that Eagleton published from the Dizai, Girdi and Herki tribes and from the area north of Mosul all partake of a somewhat dull palette, dominated by browns and rust reds. From Eagleton's dating, these are almost all from the 1920's-1950's. Another possible source, the Kurdish Textile Museum in Erbil (which specializes in the weaving of that area) seems to own few older pieces or ones that don't appear largely dyed with synthetics: http://www.kurdishtextilemuseum.com/

However, Michael's suggestion of Mosul as the origin of the "Wild Thing" becomes more persuasive when older sources are consulted. Both Mumford and Lewis provide descriptions of what the former terms "the Mosul fabric" that are quite different from what later authors seem to be writing about. The common features that all cite are: long, very glossy wool, coarsely woven with heavy handle, a range of very saturated colors with a particularly strong emphasis on yellows ("a Mosul mark"), with reds, blues, greens and white commonly included; the field devices are quite diverse, with Persian, Caucasian as well as Kurdish influences and barber-pole stripes. Langton and Lewis both mention a characteristic end finish that Lewis describes this way: "At each end a narrow selvage with one or more stripes of colored yarn running through, usually red and blue. As a rule one of these selvages is turned over and hemmed."

Could this be the double-twined heading cord?

While the diverse designs on these earlier Mosuls are "wrought roughly" but with "honesty" (Mumford), Lewis's commentary on the example in The Practical Book of Oriental Rugs (4e, 1913) reflects the esteem that some of these rugs attracted: "The superb lustre of the wool, comparable only to silk plush, and the softness of the long pile - features which have made famous the rugs from this region - are here conspicuous. The color scheme of the piece is remarkable for its harmony. The dominant color is wine, introduced in many beautifully blended shades, and relieved with correspondingly soft tones of old blue and green, the whole illuminated with well-arranged bands of white."

I think the characterizations of the pre-twentieth century Mosuls fits the "Wild Thing" pretty well.


Rich Larkin
December 2nd, 2013 01:03 AM

Hi Joel,

About Mosul (Mossul, etc.), Heinrich Jacoby said, "A misleading trade description for all imaginable sorts of Kurdish Namasés and Sedjadés [i. e., size references] from the neighbourhood of Hamadan. The name was doubtless adopted because such carpets in old times passed through Mossul." How to Know Oriental Carpets and Rugs (English edition, George Allen & Unwin Ltd. [1952]. It's a bit anomalous to write Mosul off as a legitimate label on the grounds that it is applied to rugs from the greater Hamadan weaving area, as the latter term is itself a huge catchall. Edwards says several hundred villages contribute to the inventory, and it may reach a thousand if the category is taken broadly. Furthermore, the name "Hamadan" is applied to rugs that passed through that city. So, which town did the rugs pass through? The two places are a good distance apart. This is the kind of loose talk that gives rug research a bad name.

Even so, I know what Jacoby was getting at. My own experience is that a wide variety of heavy, coarse rugs, the majority probably featuring single-weft construction, and perhaps all-wool construction, too, would attract the Mosul label. I think it would be more accurate to say that type of rug was likely to carry either the Hamadan or the Mosul label, depending on the proclivities and prejudices of the dealer writer, or auctioneer in question. The wilder and woolier models were likely to come off as Mosul. It's up for grabs whether they came from that region, or whether they represented a particular production that once tended to pass through Mosul. I suspect that some of the types traditionally attributed to Mosul were based on some level of knowledge of the attributer, and others, especially more recently, have been based on that venerable, essential tool of rug research, the guess. :cool:

My experience over the years is that the term "Mosul" has been applied widely to a great range of distinct rug types. The same can be said of Sauj Bulagh, and there is probably a good deal of overlap, too. Both terms are often the last refuge of the desperate. But I think you are right that the the two terms are usually applied to different types of presumably Kurdish rugs.

BTW, your link to the Kurdish Textile Museum is interesting. They are putting lots of stuff in the Erbil Plain. Sounds good to me, as long as no one confuses it with the gerbil plain.


Joel Greifinger
December 5th, 2013 12:58 AM


the term "Mosul" has been applied widely to a great range of distinct rug types. The same can be said of Sauj Bulagh...the two terms are usually applied to different types of presumably Kurdish rugs.
Hi Rich,

And, as we've seen, different sets at different times and by different authors.

For another historical take on the categories, I consulted Werner Grote-Hasenbalg. In his 1922 collection, he has an array of Northwest Persian rugs he labels as Kurdish. In addition to the Bijars and Sennehs (and one Gerus), he shows two attributed to Sautschbulag and two as nomadic rugs he terms Mossul. First the Sauj Bulagh:

and now the Mosuls:


Both terms are often the last refuge of the desperate.


Rich Larkin
December 5th, 2013 08:39 PM

Hi Joel,

Here's what the quirky but always entertaining Jacoby had to say about S in How to Know Oriental Carpets and Rugs.:


"Souchbulagh. An Adshemi-Turkish word meaning “cold spring.” There is no justification for pronouncing it “Sautschboulach” because some people so pronounce it. [Huh? Ed.] This small city lies in northern Kurdistan, on the Persian side, to the south of Lake Urmia and near the Turkish frontier. There is good pasture-land all around it, with great flocks of fat-tailed sheep, whose wool is of exceptional quality. Souchbulagh is the bazaar where neighbouring Kurdish tribes buy their requirements and sell their products. It is the seat of the highest ecclesiastical authorities of Kurdistan at whose head is the Sheikh-ul-Islam, whose place in Islam is somewhat like that of an archbishop of the Christian Churches.

The city’s inhabitants, but to an even greater extent the sedentary and nomadic Kurds of the neighbourhood, have for centuries been making carpets whose colour, workmanship and wool are equally attractive. They are easily distinguished from those made in the larger cities by their originality. Aniline colours have never yet been found in them, and the author has always been struck with admiration for the glorious vegetable dyes in these carpets, indigo (for the most part locally produced) and madder-red from ripened roots, buckthorn-yellow, Djefti-brown, etc. In addition to a violet-blue with brownish tinge, a wonderful dark green is often found, which must be a mixture of indigo and buckthorn, or possible of indigo and Isperek. The general colour-effect is rather dark, as a great deal of Surmey, dark green and dark brown is used, but this darkness is not unattractive, since the exceptional luster of the wool gives brilliance to the colours. Only Keleďs, so far as is known, are produced in this neighbourhood."
He defines "keleď" as a sort of double-wide runner that is typically about three times length to width. In truth, he is much given to assigning archaic rug size terms (zaronim, dozar, etc.) to the products of various venues which I am seldom able to affirm with confidence based on my own experience. For what that's worth. In any case, he does seem to be describing a particular kind of rug which is in accord with my old time sense of the rugs from that source. The variety of types you've posted seems broader. If S was the market town for Kurds from the western sector of the mountains lying easterly of the town, one would expect some variety in the collective weavings that showed up there, quite likely along the lines of your posts. One might speculate that Jacoby was describing a fabric woven in or very near the town itself, but that proposition would seem to run counter to Michael Wendorf's comments regarding his own understanding and that of Eagleton.


Alex Wolfson
December 10th, 2013 08:48 PM

Hi Joel,

Thank you for a very interesting thread.

In this case it really is difficult to make attributions based on pictures. The structure of what I would term a Sauj Bulagh rug is so distinctive as to leave little doubt, and is described well by Michael Wendorf.

Your 7th image (the one after the Meyer-Mueller rug) has all the structural and colour characteristics of the type, including the selvedges. Some of the other rugs you show, although sharing many design elements, plainly have different structural characteristics.

Where precisely all these weavings were made, and how they relate to each other geographically, is of course a matter of speculation. I don't believe we have a known provenance for any of them, although I would love to be corrected. That would be the best starting point.

Benjamin Tholen
December 11th, 2013 10:05 PM

Hi Joel,

Thanks for that interesting thread.
As you might remember I had a rug in discussion here a short time ago, wich might or might not also fall into the sauchbulag categorie.
It seems to me that we have to make a distinction between the actual place sautchbulag and the rugs made/sold there on the one hand and a specific group of rugs called sauj bulag on the other, wich all share certain elements of design as well as structure and may originate in that area so nicely decribed by jacobi, but may not be the only kind of rugs originating from the area?

I suppose that not all rugs made and sold around saujbulagh will necessarily share all the typical attributes of the given group, yet it may still be proper to call them sauj bulagh in terms of the area of their origin?

My rug is a good example. It shares the field design of many rugs in the group, has the typical background color of corroded browns, very soft wool and general structure seems in line, but weft color, some of the pile colors, and the style of border are not identical with other rugs of the group.
Therefore I would conclude that it may very well have been woven in the area, but possibly at a later point in time. It may belong to the area but not exactly to the group. Should we call it sauj bulagh then or not?
I would count it in as an late example of the group maybe, but others may disagree.
What I am interested in now is the question how we should judge, in general, rugs that share certain but not all elements of a group. Certainly some aspects should be more important than others but wich are these? Design? Structure? What aspects of structure?
I suppose that here too, it will depend on the kind of weaving we are looking at to say wich attributes should be concidered crucial?

For example weft color. While it seems resonable to me that in a quasi industry like setting of commercial workshops a standard yarn with standard color may have been used for all rugs woven in a certain period, I always wonderd what should keep a kurdish nomad weaver from using any yarn of any color available? Tradition? In weft color?

On the other hand I would expect that certain colors and woolquality were bound to local tradition and a timeframe and so on.

So, taken the given category of saujbulagh, wich attributes would we concider as hallmarks, wich ones might be secondary?

After all I tend to have the impression that its alway a complex mixture of elements and the way they are combined, something that may more "accurately" be guessed, based on intuition, rather then counted by checking a list of elements?

Kind regards

image von Benjamin Tholen auf Flickr

Marla Mallett
December 13th, 2013 11:08 PM

I have some comments on Benjamin's question about important structural features, but I'll put them on a separate thread so as not to divert attention from the specific topic here.


Joel Greifinger
December 15th, 2013 12:13 AM

Distinctive structure?

Alex Wolfson wrote:
The structure of what I would term a Sauj Bulagh rug is so distinctive as to leave little doubt
Hi Folks,

My journey through the history of the "Sauj Bulagh" designation and even its current usage by those with the greatest impact on its denotation (i.e. rug departments at auction houses, high-profile dealers, rug book authors, etc.) leads me to a different conclusion. In the spirit of Marla's comments on structural features, a review of these (and some other elements) cited by Michael Wendorf and others is in order:
1 - flat back
2 - symmetrical knots between 45-80 ksi
3 - ivory warps
4 - two or more shoots of red or coral wefts (though sometimes yellow or green)
5 - rounded knots
6 - thick staple, glossy wool
7 - corrosive brown pile in the field
8 - rounded overcast selvages not divided into color bands
9 - a wide array of clear, saturated colors often including green, yellow, white and purple
and, as for design:
10 - Persianate designs predominate, but there are a lot of exceptions using shikak, ashik and other common Kurdish motifs

In her "Overlooked Structural Features" thread, Marla lists these five:
1. Crossed wefts.
2. Devices used to straighten crooked weavings.
3. Devices used by weavers to better articulate certain kinds of designs.
4. Distinctive end finishes.
5. Distinctive selvages.

While a number of these occur frequently in Kurdish weaving (end finishes and selvages are usually noted, crossed wefts certainly less so) they don't seem to provide any basis for the differentiation of a Sauj Bulagh type from other weaving from the Persian regions of Kurdistan, as far as I can tell. While overcast selvages are often cited, it is striking (but hardly surprising) that a large sample of published Sauj Bulagh rugs are listed as not still having their original selvages. And the most notable thing about their end finishes is the conspicuous lack of the sorts of distinctive end finishes so commonly found on many other varieties of Kurdish rugs. Here again, these may have once been there, but have not survived the years.

Rather than concluding that there is a distinctive structure, I think what defines Sauj Bulagh is fairly broad family resemblance mixed in with a jumble of historical usages and under the influence of evolving market forces. In other words, the usual story. I agree with Benjamin about the "complex mixture of elements", but don't think that it leads us to any je ne sais quoi that is more "accurate", with or without scare quotes.


Your 7th image has all the structural and colour characteristics of the type, including the selvedges. Some of the other rugs you show, although sharing many design elements, plainly have different structural characteristics.
Alex - Could you specify which rugs and what characteristics?


Patrick Weiler
December 15th, 2013 06:39 AM

Watch Out

You mentioned this structural feature:
"two or more shoots of red or coral wefts"
Marla has noted in her book Woven Structures, that the correct term is "shot".
As in "drinking a shot" or "getting shot".
If she were there with you, she might smack you with a shuttle.
Carry on.

Patrick Weiler
I have searched my bunker and find no Sauj Bulagh rugs, so my commentary is limited to polite observations of little relevance.

Joel Greifinger
December 15th, 2013 03:59 PM


If she were there with you, she might smack you with a shuttle.
Thanks, Pat,

As one of Marla's avid (but less than able) students, I try hard when I write anything about structure to keep her strictures (terminological, as well as substantive) uppermost in mind. I must have been so focused on not producing a real zinger (like describing the "wefting") 
that I let a "shoot" slip through.

In my (feeble) defense, I was quoting (without suitable citation, of course:devil: ) elements of the summary of an earlier discussion of Sauj Bulagh rugs in the Turkotek archives. And, since my rationalization is already getting a bit unseemly, I'll leave the original author's name out of this. :angelic:

No Sauj Bulagh in the bunker? Aw, Patrick!


Filiberto Boncompagni
December 15th, 2013 06:37 PM

Hi Joel,


As one of Marla's avid (but less than able) students, I try hard when I write anything about structure to keep her strictures (terminological, as well as substantive) uppermost in mind
You could suggest Marla to rename the next edition of her book to “Woven strictures”.



Benjamin Tholen
December 16th, 2013 02:18 AM

As the rug I added has what looks like a figure 8 eight overcast of two warp bundels, that would fit a kurdish but not a saujbulagh attribution I suppose?
An interesting Feature regarding the shots is that almost any color one can find in the pile has also been used as shot ocassionally....

Best benjamin

Joel Greifinger
December 16th, 2013 04:35 AM

Shots of many colors

An interesting Feature regarding the shots is that almost any color one can find in the pile has also been used as shot occasionally....
Hi Benjamin,

As Marla mentions in the "Overlooked Structural Features" thread, it's not unusual for Kurdish weavers to use any suitable left over pile yarns as wefts. The "Wild Thing" rug in this thread has red, green and brown wefts, in just about equal proportions.


Joel Greifinger
December 19th, 2013 02:31 AM

Hi Benjamin,

Earlier, you wrote:

After all I tend to have the impression that its alway a complex mixture of elements and the way they are combined, something that may more "accurately" be guessed, based on intuition
Any aspiration to accuracy aside, I must admit that when I saw the picture of this Kurdish rug online (with no further information, structural or otherwise) my immediate intuition was that it is close kin with the "Wild Thing".

I will file them together in my recently acquired and perhaps idiosyncratic category, "Maybe Mosul: Late-nineteenth century criteria". :cool:


Patrick Weiler
December 19th, 2013 09:41 PM

Megri Christmas?

That rug looks a bit like the traditional Makri or Megri style from Anatolia.

Patrick Weiler

Benjamin Tholen
December 20th, 2013 12:30 AM


Originally Posted by Patrick Weiler (Post 16351)

That rug looks a bit like the traditional Makri or Megri style from Anatolia.

Patrick Weiler

Indeed. This may lead to think about crossborder connections between kurdosh tribes in iran and irak. Again


Joel Greifinger
December 20th, 2013 04:01 AM

Cross which border?

This may lead to think about crossborder connections between kurdosh tribes in iran and irak. Again
Hi Benjamin,

While the idea was once widely propagated that so-called Makri (or Megri) rugs such as those with the two-panel design that Patrick is referring to, came from the island of Rhodes, it is now established that they were woven in the vicinity of what is now Fethiye in Southwestern Turkey. To my knowledge, there has not historically been any Kurdish presence in this area, either on the island or the mainland. There was, however, a large Greek population prior to the population transfers, though I don't know how any of this bears on your speculation about relationships between east Anatolian Kurdish weaving and Kurdish production in Iraq and Iran.


Alex Wolfson
December 20th, 2013 05:50 AM

Hi Joel,

I have not looked back onto the thread for some time, but here are some examples that I hope will shed light on the questions you were asking.

The first is a rug that sold at Christies a couple of years back. It is similar to the black and white image from Eagleton, although probably a generation or so younger based on the relative stiffness of the drawing.

It has the red wefts, meaty handle, corroded browns and rich palette already mentioned, but two additional features stand out:

- thick red selvedges bound in a figure-of-eight loop
- outer guard border is formed by a line of single colour with another line of single colour between it and the selvedge.

Interestingly this rug retains its end finish, though I have not seen enough others to tell whether this is typical.

Here is a detail from another rug, with similar features:

These two are what I would call typically Sauj Bulagh. Although the designs differ considerably, there is enough commonality to group them together.

And here is a detail from a rug with a divergent structure:

It has a looser handle, with softer wool, and plain coloured wefts and selvedge. Taking this together with the design and colouring I would classify it as being closer to the 'Mosul' group.

Benjamin Tholen
December 20th, 2013 01:31 PM

Dear Joel,

I was just, on a much more general level, assuming you suggested that the "megri style" rug and your "wild thing" might have been woven in the same area/by the same people, so that in this case a designtransfer happened from anatolia to iran via kurdish people, who you assume to be the weavers of both rugs?

I dont know if my impression is right, that kurdish weavers showed a relatively open attitude towards external design sources, but reinterpreting them in a way that is very expressive and destinctively their own, therby spreading designs from iran to anatolia and vice versa?

In this context I have noticed that many authors try to seperate designelements of urban origin in nomadic weavings from other elements they concider "authentic" tribal in origin, therby suggesting the influence of Urban motivs in nomadic carpets is a form of cultural decay and a weaving showing these influences is less authentic and probably more commercial.
While I dont want to question the relevance of tracing back sources of designs, I think the conclusions they jump on (urban origin=commercial) are wrongly underestemating the playfullness of the weavers?
After all to me its not the design but the way that its articulated, that speaks of the circumstances of its production and therefore should be our main focus if trying to determine "authencity" of a weaving?
Or is the ikea-cotton-cloth with tribal print more authentic than a qhasqai pictorial rug?

Regaring Playfullness: while it may very well be, that there were very strict traditions as to what has to be woven in what way with certain tribal Communities and for certain kinds of textiles, do we have to assume it was always this way?
How free was, for example a kurdish weaver, to adobt designs of her liking (and likewise techniques) she encoutnered in lets say weavings seen on a bazaar?
Could this be the story of the rug you posted? The weaver of your wild thing, as productive as fassbinder in his best years, after finishing it, gets a peek on a megri rug at a bazaar, and says to herself: If I cant do this...! ?

Looking at Gabbeh rugs, wich are mostly, as I am aware of, a pretty late phenomenon, they feature all kinds of structural variation. I have a very long piled gabbeh, with very soft wool, employing symmetric as well as assymetric knots in the same rug and while the wool seems handspun, the selvedges (original?) are overcast with what looks like stripes of pinkish cotton first aid bandages???

Now while obviously learned techniques are not easyly replaced by new ones, there may very well be much more variation than we assume or want to aknoledge in our attempt to classify weavings and relate structures to origins in time and space?

Kind regards

Marla Mallett
December 20th, 2013 07:41 PM

Let me say just a bit about why technical/structural features tend to remain constant within the work of groups of village and tribal rug weavers, while design elements can vary widely.

It's not just a matter of sticking with familiar processes...though continuing to use techniques learned from early in one's weaving career is a large part of the matter. But more important is the fact that very few of these weavers work alone...Most rugs are woven by two or more women working together. Several family members may participate in processing the materials. Thus it is crucial that everyone be "on the same page"--that everyone involved stick with the same standards, the same identical processes. Materials must be spun and plied to the same precise standards, and weavers working side-by-side must handle the yarns (both for knotting and wefts) in exactly the same fashion, or we would be guaranteed badly crafted, badly lop-sided rugs. I've seen instances in which four or five family members or friends have participated from time to time in the weaving of a single rug. That common practice exerts a very strong pressure on individuals to honor their "community standards" exactly. This is why structures and weave balance can be reliable guides to a piece's origin--much more reliable than motifs--which indeed can be easily "picked up in the marketplace."

Of course we encounter rugs that include two different knot types. it is very common for weavers who use asymmetrical knots throughout their pieces to place symmetrical knots at the edges. These are more secure, and they don't hang over the edge on one side and leave the foundation on the other side exposed, as do asymmetrical knots. But to "see a different structure in the marketplace" and try it out??? That might have some validity when it comes to some of the more unusual flatweaves...but knotted pile?


Marla Mallett
December 23rd, 2013 02:48 AM

People continue to ask what I mean by WEAVE BALANCE. So here's part of what I've written previously on this:

"Along with dramatic differences in designs, dyes and fibers, subtle variations in weave balance help us to separate knotted rugs into distinctive groups. A satisfactory weave balance is not easily achieved. It requires experimentation. A perfect harmony must exist between the yarn sizes of warp, weft, and knot, as well as several construction features. The kind of material, its preparation by carding or combing, the amount of spin, and the kind of ply all affect the loft, density, and elasticity of each structural component. These must be coordinated. The final variables are the warp spacing, the configuration of the knot, the amount of ease that wefts are allowed, and the amount of beating they are given. If a single one of these factors is altered, the critical balance can be disrupted. Rather than tamper with a good thing, village and nomadic weavers tend to replicate the distinctive weave balance gradually worked out by people before them."

Problems resulting from poor weave balance may indicate that certain groups of peasant weavers do not have lengthy histories of pile rug production. "Poor craftsmanship" may suggest disruption or immaturity in a tradition, and not just shoddy work by a single individual.


Benjamin Tholen
December 24th, 2013 01:51 PM

Dear Marla and all,

Thanks for the explanations on weaving as a community work and how technical standarts must be matched to achieve a satisfying result. I once had a very large quchan kurd in typical hawzi design, the right half of it was woven pretty disciplined, the left a complete mess. Probably the work of weavers on diffrent level of experience? Why didnt the more experienced weaver correct the less experienced one but seemingly accepted the lower standart of the other weaver? Could we concider such weavings as part of education? Like a "gesellenstück", in wich failure is tolerated?
Was sharing work more common on larger rugs than smaller ones?

I was always drawn to such pieces and my main intrest in rugs and almost any art ist the delicate balance between success and failure, symmetrie and assymetrie in shapes and colors, the thin line between a shape falling completley apart on the one side, and dying in stiffness on the other, with an artwork being alive as something inbetween, maybe a constant struggle between these forces. Enough order for us to recognise a pattern but enough chaos so that we have to do (at least subconcious) work in our heads to "make sense", take it inside...

Backt to the sauj bulagh issue:
Despite the rug having been in discussion before, I am still stuck with the piece I presented a few posts ago and the question In how far it belongs ore relates to the sauj bulag group?
I have just now taken some close up pictures of the pile in daylight, that I would like to share, to discuss some features.

We allready discussed the question of natural dyes in this rug and came to the assumption that a mixture of natural and chemical dyes has been used in this rug, wich would make it younger than others of the group, probalby around 1900.

Now I wonder if wool quality isnt an underestimated aspect determining the impression that colors make and beyond this a good hint on determining the origin of a piece?
I assume that wool in village and nomadic production and even smaller workshops was always of local origin and rarely imported from areas far away?

Now what makes this rug interesing in this context is that large parts of it have been rewoven with yarns of inferior dye quality and - more striking - much lower quality of wool. I would assume that on this kind of dry, dead wool, any quality of color would look bad, while the soft, superglossy wool used in the original would ad live to any dye, especially if the yarn was dyed manually?

Here come some examples

In the last picture one can see what the original brown ground looked like before corrosion (almost golden, to the left of the aubergine ground shape) in opposition to the repairwork in brown, that is dull and dry (right above it in the picture.)

Now there is one other thing I found while taking the pictures, strange I never noticed before despite all the time I spend looking at the rug...

(the bright green is repairwork as well as the dark brown of course)

Usually I am sceptical of giving to much importance to the appearance of a cross shape in a rug, as the cross seems to be to universial as a motiv, but this one is really strange:
Basically all other motivs in the rug seem to be either part of the "harshang plus flame" pattern, or random filler Motivs of probably tribal origin, but all of them appear several times. Now exactly in the "goldene Schnitt" (beginning of top third, central position), this cross is placed. Singularily within in the whole rug and with stong resemblance of a cristian cross?
Is this usual with kurdish weavers? Are there any similar devices in other examples of the sauj bulagh group? Could it hint on an armenian influence?

Kind regards and a merry xmas to everyone...
(The cross counting as my attribution to the seasons textile ;) thread )

Steve Price
December 24th, 2013 03:38 PM

Hi Benjamin

I wonder if that cross is original to the weaving or was added later. The knots look like they're oriented along the lines of the cross rather than all being tied around adjacent wefts.

Steve Price

Benjamin Tholen
December 24th, 2013 05:14 PM

Dear steve,

An interesting possibility, though type of wool and colors seem absolutely identical with the rest of the rug. (Apart from the bright green repair of course.)
I have a problem understanding the technical terms here, shouldnt a knot be tied around warps rather than wefts? Sorry for my misunderstanding here...
I could try to look at the back at some time after xmas to get a better impressin maybe?

Kind regards

Steve Price
December 24th, 2013 05:17 PM

Hi Benjamin

My error. Knots are tied around adjacent warps, of course.

Steve Price

Joel Greifinger
December 24th, 2013 10:48 PM

Hi Alex,

I want to follow up on your posting of those three rugs and your comments. The first rug was the one that sold at Christie's. I did not see that rug in person and would not have recognized it as the same rug that appeared in their catalogue photos. Yet another reminder of how dramatically different colors in the same piece can look in different phtographs. Here's the catalogue picture next to yours:


You wrote
It has the red wefts, meaty handle, corroded browns and rich palette already mentioned, but two additional features stand out:

- thick red selvedges bound in a figure-of-eight loop
- outer guard border is formed by a line of single colour with another line of single colour between it and the selvedge.
Did you mean to imply that those lines in a single color are diagnostic for a Sauj Bulagh attribution, what you in an earlier post called a "distinctive" feature? The details from the following example show "barber pole" lines on either side of the main border. You say that it is nonetheless "typically Sauj Bulagh". Can we assume that it shares the red wefts and the meaty handle you describe in the first rug? It also has a repeated minor border motif on different background colors, a design feature that has sometimes been proposed as common in the Sauj Bulagh group.

Is there any possibility that you could post a full photo of this second rug? :thumbsup:


Marla Mallett
December 24th, 2013 11:47 PM


In my experience visiting village and nomad weavers, just about all weavings except for small bags are woven by two or more weavers--NOT just really large rugs. Even on fairly narrow rugs, they normally work side by side, two or three girls or women crammed together, arm to arm in a very tight space. The women enjoy this sociable time together, away from less pleasant household chores. Also with just one loom per household, it means that the rugs can be completed in much shorter periods of time--thus producing more income faster. Josephine Powell used to tell me that sometimes friends or neighbors barter their weaving time--one woman helping the other, then when that rug's finished, reversing the process to switch to the other's rug.

As for inept weaving on part of a rug: that's a natural result of the educational process for novices. If "errors" in the patterning occur, that's often not noticed until after the fact. I have never seen or heard of anyone backing up to correct any knotting errors. Nor do mothers seem to criticize a daughter's work--at least I've not witnessed that. Women have, instead, laughingly pointed out to me strange idiosyncrasies in their rugs as parts that were woven by their youngsters. Occasionally I've encountered situations with young girls working on their own, always a small bag or yastik. But this is rare, as families seem reluctant to have a young girl tie up their loom. This seems to be allowed only during periods when the women are busy with other chores--perhaps times when they are preoccupied with cheese-making or caring for new lambs. Or perhaps, as mentioned above, when they've agreed to help a neighbor with her rug.


Benjamin Tholen
December 25th, 2013 04:38 PM


I suppose the same can be expected for flatweaves?


Who could have added such a cross to an existing rug and for what purpose? Have you encountered this phenomenon in other rugs previously?

Kind regards

Steve Price
December 25th, 2013 05:34 PM

Hi Benjamin

I've never seen anything similar. That cross sure looks like a Christian symbol (most "crosses" don't). Who put it on, and why? Who knows? Not me.


Steve Price

Marla Mallett
December 25th, 2013 06:47 PM


You've asked if the same production factors apply to the flatweaves. Well, yes and no. First, there are now so few nomads and semi-nomads left that production of the flatweaves is nearly at an end, at least in the places I know about. The typical situation among recently-settled nomads is that a few of the older women still weave an occasional kilim or brocaded piece, while their daughters have learned to weave pile carpets, since that's what they can sell. Thus the traditional chain of mother-to-daughter teaching of weaving skills is broken. Since kilims and bags were accoutrements of the nomadic life style rather than commercial items, like pile carpets, the production of them was not typically accompanied by time pressures to produce as much as possible to increase the family income.


Joel Greifinger
December 25th, 2013 07:26 PM

Kurdish Nestorian Christians?

That cross sure looks like a Christian symbol (most "crosses" don't). Who put it on, and why? Who knows? Not me.
Hi Steve and Benjamin,

Well, me neither.

But, for what it's worth there have been Kurdish Christians, mostly Nestorians, living in Kurdistan for a long time. Whether they wove rugs, I don't know. However, John Joseph in his The Modern Assysians of the Middle East wrote that these Nestorians were "scarcely distinguishable from the Muslim neighbors who were interspersed among but not intermingled with them." He points out their "uniformity of custom" in all but religious practices, including "industry". Perhaps this included weaving.

Then again, as Steve suggests, perhaps this was the work later of a restorer.

"Who knows? Not me?"


Marla Mallett
December 26th, 2013 12:01 AM

Someone has just asked me why there are so many knotted-pile saddlebags and bag faces available to collectors, if, as I've insisted, nomads wove primarily tapestry, brocaded, soumak and other flatweave pieces???

As I mentioned in a post above, when nomads gradually settled in most places across the Middle East, they gradually abandoned their flatweaves and turned to pile weaving, because knotted rugs were such a good source of family income. But even though they gradually lost those flat-weave skills, villagers still kept horses or donkeys, and thus had a practical use for saddlebags. The result: knotted-pile saddlebags became common village products. To me, they are normally among the most interesting of all pile pieces, as they were most often non-commercial, ethnographic weavings. Since they were pieces usually made for the family's own use, and normally made by a single individual, weavers had more freedom to be experimental and creative if they wished.


Alex Wolfson
December 26th, 2013 12:12 AM

Hi Joel,

The second rug also has a heavy handle, red wefts and the distinctively looped selvedges. I took those pictures while visiting a private collection, and do not have a picture of the whole carpet. The design as I recollect is a field of tribal and geometricized floral motifs arranged in a pattern that recalls earlier 'vase' carpets.

The stripes of single colour forming the outside of the outer guard border is a feature I have noticed on many rugs sharing the same structure, which is why I wanted to add that observation. As you also note, these rugs often have a different background colour for each minor border - but it doesn't appear to be the rule.


I see no reason why the cross motif would have been added later - it does indeed look like a Nestorian cross. Nestorian (or Assyrian) communities were living in this part of Kurdistan, as one can read in this report from 1915 (Viscount Grey, The Refugees in the Caucasus):

For two months---June and July, 1915---the Armenians of Van enjoyed an autonomous national government under Russian protection. But in the last days of July the Ottoman armies on this front received strong reinforcements, and were able once more to take the offensive. The Russian troops began to fall back from Van on the 30th July, and practically the entire Armenian population of the Vilayet accompanied them in their retirement.

The retreat was unexpected. The refugees had few conveyances and hardly any provisions ; and, though their rear was protected against the descents of the Kurds by the heroic fighting of the Cossacks and the Armenian Volunteers, the suffering and mortality, during their flight over almost trackless mountains, was appalling.

At Etchmiadzin and Erivan, across the Russian frontier, the Armenian refugees were joined by the stream of Nestorian fugitives from Urmia, and the total number of Christian exiles in the Caucasus rose to over a hundred and eighty thousand.

A brief history of the Nestorians from the Encyclopedia Britannica:

Nestorius had been anathematized at Ephesus in 431 for denouncing the use of the title Theotokos (“God-Bearer”) for the Blessed Virgin, insisting that this compromised the reality of Christ’s human nature. When supporters of Nestorius gathered at the theological school of Edessa, it was closed by imperial order in 489, and a vigorous Nestorian remnant migrated to Persia.

The Persian Church’s intellectual centre then became the new school in Nisibis, which carried on the venerable traditions of Edessa. By the end of the 5th century there were seven metropolitan provinces in Persia and several bishoprics in Arabia and India. The church survived a period of schism (c. 521–c. 537/539) and persecution (540–545) through the leadership of the patriarch Mar Aba I (reigned 540–552), a convert from Zoroastrianism, and also through the renewal of monasticism by Abraham of Kashkar (501–586), the founder of the monastery on Mount Izala, near Nisibis.

After the Arab conquest of Persia (637), the Caliphate recognized the Church of the East as a millet, or separate religious community, and granted it legal protection. Nestorian scholars played a prominent role in the formation of Arab culture, and patriarchs occasionally gained influence with rulers. For more than three centuries the church prospered under the Caliphate, but it became worldly and lost leadership in the cultural sphere. By the end of the 10th century there were 15 metropolitan provinces in the Caliphate and 5 abroad, including India and China. Nestorians also spread to Egypt, where Monophysite Christianity acknowledged only one nature in Christ. In China a Nestorian community flourished from the 7th to the 10th century. In Central Asia certain Tatar tribes were almost entirely converted, Christian expansion reaching almost to Lake Baikal in eastern Siberia. Western travelers to the Mongol realm found Nestorian Christians well-established there, even at the court of the Great Khan, though they commented on the ignorance and superstition of the clergy. When during the 14th century the Church of the East was virtually exterminated by the raids of the Turkic leader Timur, Nestorian communities lingered on in a few towns in Iraq but were concentrated mainly in Kurdistan, between the Tigris River and Lakes Van and Urmia, partly in Turkey and partly in Iran.

Happy Christmas to you all!

Benjamin Tholen
December 26th, 2013 05:11 PM

Dear Alex and all,

Thanks for the highly interesting information on nestorian kurds and the dramatic, sad fate they shared with the armenians. Initially I was thinking of an armenian influence, But the explicit mentioning of nestorian kurds settling around lake urmia would be in line with the area sauj bulag rugs are supposed to come from, if I recall proberly and could provide a reasonable explanation for the cross featured in this rug?

Among all the rugs diplaying the harshang pattern in the sauj bulag group there are some quite dramatic diffrences in the style of drawing, with the stiffer designs usually being assumed to be less old.

This rug seems to contradict this assumption, as judging by the use of some chemical colors, it probably is a turn of the century product, yet the style of drawing is not only very free, but also extremly similar to these two rugs wich are both assumed to be of early to mid 19th century origin:

Interesting enough the white ground version is attributed to shirvan, the yellow ground to sauj bulag.
The rug in discussion, being much wider, repreats the central motiv axis of the other two and is so similar in the way the motivs are articulated, that I would a) assume that the first two have been woven by the same people and b) that the one in discussion must be a direct "offspring" of them?
...left aside the highly interesting nestorian cross, that I cant find in the other two, nor any sauj bulag rug I could find...

Steve, I took a picture of the back and hope it helps to settle the question of wether the cross was added at a later point or not...

Finally, I am not sure now what kind of selvedge structure would be expected in a sauj bulag rug. Alex mentions a red figure eight overcast, wich would be in line with this rug, but other carectristics have been claimed to be typical of saujbulag rugs before...

Kind regards

Benjamin Tholen
December 30th, 2013 05:17 PM

Just found a highly interesting work on nestorian communities in the area between lake urmia and lake van online. It describes them as very war like people, often in conflict with the muslim kurdish people surrounding them.
While the sedentary parts of the tribe are mentioned to have settled in nw iran, the nomadic tribes occupied the mountain areas of the hakkari region between the lakes.
This could be an explanation of the nw persian drawing but more anatolian colors found in this rug?
Anyhow a higly interesting article about the nestorians, including the role women took in the armed conflicts. Besides the authors seem to see nestorians and kurds as different people...

You can download it here http://www.ccsenet.org/journal/index...load/9872/8059

Kind regards

Horst Nitz
January 22nd, 2014 11:41 PM

Hi Benjamin,

thank you for that interesting article. I've downloaded it and am looking forward to look into it on one of my next train rides.

Best wishes,


Joel Greifinger
January 25th, 2014 03:19 AM

To bring things sort of full circle to the beginning of this thread, here's a rug that recently sold at auction:

It has two shots of thin red wefts on ivory wool warps, overcast selvages, and a saturated palette of yellow, white, blues and greens on a field of corroded dark brown. The wool (where it it isn't quite worn) is lustrous, the back is flat and the handle is heavy and floppy.

The field is a variant of the design termed "tuning fork" in the Levi article in Hali #70 that seems to have sparked the brief flurry labeling such rugs as "proto-Kurdish" (cf. Christie's auctions and others in the late 1990's) despite objections to the logic of such a designation. This is the rug from that article:

and this is a probably older example from the Burns collection:

A discussion of these rugs, and the controversy over their design antecedents, can be found in the Turkotek archives in the salon, "The 'Tuning Fork' Kurdish Rugs'": http://www.turkotek.com/salon_00092/salon.html

The combination of the "tuning fork" motif and the half shikak filler motifs combines a floral Persianate form with a characteristic Kurdish motif. All parties seem to agree that these rugs originated around Sauj Bulagh.

The auction house sold the rug as "Southwest Caucasian".


Horst Nitz
February 7th, 2014 07:30 PM

The elusive Nestorian rug
Hi Benjamin and all,

it is an interesting article you have dug out in the internet, the first one coming to my knowledge by a Turkish author; perhaps he is the only one still. That it is published in a Canadian journal is telling.

Robbery, blood feud and gunpowder have been mentioned by the early authors before, who had visited the Mountain Nestorians in the 19th century. However, since this is about all the author tells us, the picture that emerges is rather incomplete and somewhat skewed. He excuses the Nestorians for having become what they were due to their contact with the surrounding Kurds, but this tells more about the ambivalence of the author for having picked up the theme than that it does fair justice to either party mentioned.

To fill in just a little of what might be of interest to you, the Mountain Nestorians claim to be descendants of the lost tribes of Israel who were led into captivity by the Assyrians; they were among the first Christians, lived in the most secluded valleys in the High Kurdish Taurus and roamed the highest pastures. They were reknown for their conservatism and pitied for their poorness. This was until they were evicted at the beginning of the 20th century by Kurdish irregular troops with modern weapons to whom they were no match. Their Patriarch resided near Julamerk, the old name of the present day town that lend its name to the province: Hakkari - and the Nestorians wove rugs! In a parallel thread started by Pierre, I have been discussing a classical rug that either stems from a Nestorian context or has Nestorian roots: http://www.turkotek.com/VB37/showthread.php?t=2251

Best wishes,


Joel Greifinger
March 24th, 2014 03:15 AM

Hi all,

A rug that just sold at auction raises for me the same difficulties in assigning an attribution of its origin that were raised by the rugs I was asking about when I began this thread. This one measures 8’ 4” x 4’ 6” and has, like those, lustrous soft silky feeling pile. It is woven on ivory wool warps and has two (or sometimes more) shots of light red weft. The end is finished in a band of oblique interlacing. It has eight colors. There is no corrosion in the browns.

Here it is:

And, for convenient comparison, here are the rugs I initially asked about. Spurred by Michael Wendorf’s suggestion, it seemed reasonable to conclude that the first, the “Wild Thing” was what early 20th century rug commentators called ‘Mosul’ in describing some 19th century Kurdish rugs from that area of what is now Iraq.

The other, when it was published in the “Mideast Meets Midwest” catalogue, was described as “Iranian Kurdish, possibly from Quchan.”

The auction house that sold this most recent addition described it as “Kushan Kurd Rug. Northwest Persia” (which strikes me as sort of like wanting to have it both ways, since Quchan is in Khorasan in the north east of Iran). And, while the ashik main border is certainly one mainstay on Kordi rugs (Stanzer shows three), it is used on other Kurdish rugs as well, including this one that Brian MacDonald attributed to Sauj Bulaq:

Any thoughts about whether further specificity is possible on this one?


March 27th, 2014 04:33 AM

There is no corrosion in the browns
Dear Joel,

My opinion is that you have answered your own question with the observation that "there is no corrosion in the browns" of the rug with the latch hooked lozenges that just sold at auction. Although I have not handled the rug, I have handled numerous other examples from the group. I do not believe they originate from the area of Sauj Bulaq, I think they originate from Central Kurdistan and could best be called Jaf area production. In the hand, they have a different feel from the rugs typically associated with Sauj Bulag - notably heavier and with more extensive use of blue as a ground color. They also lack corrosive browns and some other colors and color combinations that are relatively stable in Sauj Bulaq area production so far as we understand it. It is difficult to attribute much to the presence of ashik type devices, they are found in the weavings of many areas. It also appears that the rug in question may have been reduced in with on one side with the removal of a guard border. That guard border which does remain on one side and is fragmentary on the other is often found in Jaf rugs.

I do not recall the Chicago rug with the memling devices very well, but I do not see it as coming from Quchan. I would think it's origin is more likely to be Eastern Kurdistan. What I do vaguely recall is that it had fairly think and glossy wool, not attributes I think of with Quchan pieces.

I also do not know the MacDonald piece but it would appear to be closer to the first rug and therefore from Central Kurdistan. Stunning colors or not, the palette and structure from what i can see do not appear to be that of the rugs that I think of as being Sauj Bulaq pieces. Please recall that just 20 years ago no one would have given the origin of Sauj Bulag much thought when considering these pieces. That designation was reserved for such a small group of rugs that many people did not even know the term. It has now nearly become a term tossed about without real meaning or consistency.

I have great and growing respect for your appreciation of and enthusiasm for rugs. I hope you will continue to study and write about them.

Best wishes, Michael

Joel Greifinger
March 29th, 2014 06:40 PM

Hi Michael,

As always, thanks for your input. The latest example (with the large latch-hooked lozenges) certainly fits many of the characteristics of Jaf production with its pile of thick, glossy wool and consequent heavy handle. The red wefts and flat back are features common to a range of Kurdish groups. The particular combination of this field design and ashik main border on a blue ground pops up in other (putatively) Jaf rugs, like this one:

I agree that the rug labeled Sauj Bulaq in MacDonald is likely from this group, as well.


It also appears that the rug in question may have been reduced in with on one side with the removal of a guard border. That guard border which does remain on one side and is fragmentary on the other is often found in Jaf rugs.
Since the selvages have been replaced, it's quite possible that the guard border has been removed from the left side of the rug. On the other hand, the weaver(s) clearly made an adjustment on the placement of the minor borders at the point where the incipient guard border abruptly disappears on that side. It may have been sacrificed to maintaining the efficacy of the overall design.


I do not recall the Chicago rug with the memling devices very well, but I do not see it as coming from Quchan. I would think it's origin is more likely to be Eastern Kurdistan. What I do vaguely recall is that it had fairly think(sic) and glossy wool, not attributes I think of with Quchan pieces.
Not generally. But then there are always the interesting exceptions. This dated Kordi rug (the date converts to 1893 in the Gregorian calendar) has a mass of thick, glossy wool.

While I haven't had the pleasure to handle it, the "Chicago rug with the memling devices" doesn't strike me as coming from Khorasan, either. If we eliminate Sauj Bulagh, what area of Eastern Kurdistan might you have in mind?


April 10th, 2014 02:45 PM

A pastiche
Hi Joel,

Sorry it takes me so long to reply to your posts. I would be guessing as to the origin of Chicago rug with memling guls. I should probably have suggested or guessed northern Kurdistan rather than eastern Kurdistan - at least as those terms struggle to define regions of historic Kurdistan. Based on color, drawing and what little else is recorded about the rug it would seem more likely than not to come from some area in the triangle formed by Lakes Urmia, Sevan and Van. But this is nothing more than a guess.

That dated Kordi rug is very interesting not just for the wool but also the remarkable combination of design elements. The piece is handsome with nice dyes but really struggles for an identity. It has design elements from multiple weaving traditions. I suppose one might also see this piece as a reflection of its time and place. No tradition is static.

Best wishes, Michael

Joel Greifinger
April 12th, 2014 12:17 AM

Pastiche or syncretic?

That dated Kordi rug is very interesting not just for the wool but also the remarkable combination of design elements. The piece is handsome with nice dyes but really struggles for an identity. It has design elements from multiple weaving traditions. I suppose one might also see this piece as a reflection of its time and place.
Hi Michael,

I'm intrigued and a bit confused by your characterization of the dated Kordi I posted. While much of my exposure is influenced by Stanzer's study which has only a couple of pieces that date before the 20th century, I tend to think of the Khorasan Kurdish pile weaving tradition as particularly syncretic in terms of design elements and their combination. The juxtaposition of Turkmen, Baluch and Kurdish motifs while utilizing their own characteristic palette (of brick red, a purple/brown red, blues, white and prominent green and yellow) seems prototypical.

If this is what you mean by pastiche, are there extant earlier Kordi pile rugs that predate such influences? The earliest Kordi rug that Burns shows ("circa 1800") has rows of tauk noska guls that appear quite directly drawn from the Turkmen tradition. Was the "struggle for identity" already reflected in that choice?

Here are three that form a timeline from the mid-19th to the mid-20th century. All are organized around large hauzi motifs that appear frequently in Kordi pile rugs with a wide array of smaller motifs both within the hauzi and distributed throughout the field.

The first is Burns' plate 99 that he dates to the "mid-19th century" :

The next is a somewhat more complete picture of the Kordi dated 1893:

The third Stanzer attributes to Baqjirg, just north of Bojnurd and dates as "circa 1935:


Sorry it takes me so long to reply to your posts.
Thanks for staying in the discussion.


Patrick Weiler
April 12th, 2014 02:36 AM


Have you read the Hali issue #59 article by Ulrich Alheim, Kordi Without Tears, in which he lays out his identification criteria?
He mentions that many Kordi rugs entered the market after two bad winters in Khorasan in the early 1970s. In addition to format/sizes, he discusses materials, structure composition and designs.
Wefts, he says, are almost always dyed - usually red - but green, blue, purple etc can be found. Structural features include symmetric knots, no warp depression, 2 to 6 weft shoots, knot density of 19-59 per square inch. They are flexible, soft, thick and heavy but not stiff as in northwest Persian Kurdish rugs.
The end finishes include "pattern wefting technique", "transversely braided warp ends" two-strand twining etc.
He says many were used as partitions or hangings and therefore the corner warps can be used as cords for hanging.
Borders are often B-A-B, with a major A border and two flanking identical borders on either side. This, along with his mention of lattices in the main field with one or two designs, correlate with the Kordi rug I believe you are discussing with Michael:

Speaking of syncretic, he says Kordi weavings have a larger range of motifs than other comparable weaving people. Having come in contact with Caucasian, Anatolian, Turkmen, Afshar and Baluch weavings they absorbed their motifs and designs. They exclusively use the Water Tank design, but also Memling guls, Turkmen guls, hooked diamonds and stars.

Patrick Weiler

April 12th, 2014 03:47 AM

A rose by another name...
Hello Joel and Patrick,

Patrick can you check your Hali reference? I paged the entire Hali 59 and was not able to find this article by Ulrich Alheim though I do recall there being an article or note like that in Hali sometime in the past.

The structure of the so-called Chicago rug is given as symmetrical knots 7 x 7. Warp listed as tan and white wool Z2S. Weft brown and tan wool Z spun singles, No. of shoots: 2. Selvedge of 3 cabled warps. Weft twining at bottom.

I would add this: although no special knowledge or experience of so-called Kordi/Quchan rugs would ever be claimed by me I have seen and handled my share. They seem to have a distinctive back in most cases and dyed and particularly red wefts are common. I do not recall the Chicago rug well but it did not strike me as coming from this group. If Alheim is right, and so far as I know he has handled more than his share of these rugs - but mostly 20th century examples - then the Chicago rug does not have the structure he references. Most significantly dyed wefts. Some of the other structural characteristics he references are so general as to be unhelpful - such as number of weft shoots and knot density. Finally, so far as I know, Mr. Alheim is not known as someone who has studied or handled 19th century non-Kordi/Quchan rugs. Although that may be up for debate, I have only met him once many years ago.

As for my use of the term pastiche and Joel's view of these Kordi rugs as being particularly syncretic in terms of design elements and their combination I would say that Joel is an empathetic student of these pieces and I am a cynical old man. I wish I could share that empathy for these pieces but they have never resonated with me.

Keep your minds open and keep the dialog moving...

Best wishes, Michael

Frank Martin Diehr
April 12th, 2014 05:25 PM

Ahlheim in Hali
Hi all,

Ulrich's "Kordi without tears" was in Hali 111, and another related article "Seen but not Kurd", in #97. He has been studying, collecting and writing about Kordi/Kurdish rugs extensively for many years now.



Joel Greifinger
April 12th, 2014 07:25 PM

Settle for salmagundi?

As for my use of the term pastiche and Joel's view of these Kordi rugs as being particularly syncretic in terms of design elements and their combination... they have never resonated with me.
Hi Michael,

Thanks for the clarification. Perhaps rather than pastiche (with its connotation of imitation), we might settle on describing the 19th and 20th century Quchan/Kordi pile weaving style as a distinctive salmagundi or potpourri. ;)

Here's the "circa 1800" Kordi rug in Burns (plate 100):

and this one in Stanzer from the second quarter of the 20th century:


I am a cynical old man.
Well, perhaps  figuratively (though I have my doubts). I think I've got about a half a decade on you chronologically.


April 13th, 2014 03:04 AM

Salmagundi it is, but hold the anchovies
Hello Joel, Patrick and Frank:

Thank you to Frank for the citations to Hali issues 111 and 97. But Patrick was not completely off, the article Kordi Without Tears by Mr. Alheim begins on page 59 of Hali Issue 111. Moreover, Hali Issue 59 contains a beautiful image of the Meyer-Muller Flame palmette rug on page 143. Joel has supplied an image of that rug early in this thread.

Joel is always stirring the pot but we can certainly agree that when describing 19th and 20th century Ouchan/Kordi pile weaving the style might best be described as a distinctive salmagundi or potpourri.

In Kordi Without Tears Mr. Alheim does state that:

"The range of motifs employed by Kordi weavers is considerably wider than those used by other comparable weaving peoples. Their diversity suggests that weavers were encouraged to vary traditional designs, draw upon alien patterns or even invent motifs. Designs are therefore fairly soft criteria in attributing weavings to the Kurds of Khorasan.
It is not uncommon to find motifs familiar from Anatolian, Caucasian, Turkmen, Afshar and Baluch rugs...."

Two designs are referenced as only being found on Kordi weavings. The first Alheim calls the pagorbe or cat's foot (Stanzer p.83). He refers only to the Stanzer citation (a book I do not own) and a rug with many motifs so I am not certain what he is referring to. The second is the hauzi or water basin motif which Joel has supplied several examples of on April 11.

Alheim further discusses Format and Function, Materials and Color, Structure and Composition. He describes warps of wool or goat hair or a mixture of the two, in much later rugs sometimes cotton. All natural shades from ivory to almost black can be found and quite often a salt and pepper mixture is observed. Dyed warps he states "are never used." Wefts are usually of wool, rarely goat or camel hair, and "are almost always dyed. Red predominates, especially in early pieces, but green, blue, purple, yellow, orange, brown or black wefts are found. Weft color changes are often seen." Structure is recorded thus: "warps of Kordi rugs are invariably two strands, Z spun, and S plied (Z2S) and are always on one level, never packed.The number of weft shoots between each row of knots may vary between two and six in a single piece. Wefts are often a single strand, Z-spun."

I find this information to be marginally helpful and include for anyone who might be following and without access to Hali 111.

I find Mr. Alheim's observations in Hali Issue 97 "Seen But Not Kurd" which challenges the 8th ICOC catalogue for not attributing a single piece in the catalogue as Kurdish when in fact several examples obviously are. I agree with him that several of the rugs referenced are Kurdish including a flame palmette rug that Joel has also included in this thread. See plate 111 in Atlantic Collections. But Mr. Alheim compounds the problem by attributing this piece and others as Quchan Kurd relying mostly on the very "soft criteria" such as design that he discusses as unreliable in Hali 111. These pieces are not Quchan Kurd pieces and I have very personal knowledge of two of them and know other examples of two others equally well. Yes, this is my best educated guess but I am confident that most people who care about these pieces would agree with me. I will leave it at that and also ignore why an ICOC catalogue could also be misattribute such distinguishable pieces.

I would offer one more unsolicited opinion. Joel posts two rugs - Burns' plate 100 and another from Stanzer with tauk noska like guls. The circa 1800 Kordi piece is fabulous in the flesh with a real presence. However, there is no basis to my knowledge for attributing the rug to circa 1800 except that it would seem older than any other rug in the group. Does that make it circa 1800 or 5 - 6 generations older than the Stanzer rug? Personally I doubt it very much. Everyone has their opinion but there are precious few facts.

I believe Joel has exhausted my knowledge of, interest in and appreciation for these Quchan Kordi rugs. Please do carry on...

Best wishes, Michael

Patrick Weiler
April 14th, 2014 01:56 AM


Here is a picture of the rug with what "Alheim calls the pagorbe or cat's foot (Stanzer p.83)". I was similarly confused when reading the Alheim article, because he didn't say where to find it in the first rug in the article, which was one similar to the two Turkmen-gul rugs Joel posted below. It appears in the center of the guls on Alheim's rug.

Thank you for clearing up the confusion regarding Hali 59 vs Hali 111.
I could say it was an error which occurred while I was editing my post, but more likely it was because I had to hurry out the door to visit our famous annual Skagit Valley Tulip Festival about 75 miles from home.
To add to the salmagundi - a word I was previously not familiar with - which Joel referred to, here is a photo I took at one of the many tulip farms there yesterday. A veritable carpet of flowers:

This may be the first Turkotek thread to include references to both food and flowers.

Patrick Weiler

April 15th, 2014 10:55 AM

couldn't help but notice...
hello all,

great info on rug design and tradition. keep up the good work!
i came across this thread after running a search online for "julamerk" (Hakkari province).

regretably, however, and with regards to the designation of "Nestorians", as a descendant of the Assyrians now living in diaspora, I would like to put things in place:

there has been much misinformation circulating by both scholars and enthusiasts (regretably) online regarding the community. Much of what is known on the Assyrians of Sth-East Turkey, is in fact, due to missionary activity. So scholars today reference nearly all work upon the journals of missionaries 19th-20th centuries. please note the following. any works not citing the Hakkari Mountain missionaries are near propaganda or theory-based, and thus;

1) the title and designation: "Nestorian" is REJECTED by Assyrians today,
2) the title "Nestorian" is used mainly by Roman-Catholic Uniates to designate Hakkari Assyrians since they are of Orthodox, Middle-Eastern hegemony,
3) Assyrians themselves DO NOT claim to have descended from any of the lost tribes of Israel and reject the popular theories of A. Grant (1841) which deludes the western world - without any scientific evidence - into accepting the theory that Assyrians are of Hebrew stock,
4) Today's Assyrians CLAIM and designate their descent as per Genesis 10:11, and are sons of Shem;
5) based on the above, Assyrians are therefore a SEMITIC, unique race - non-aryan, non- Indo-European, and non-Arab race.
6) Assyrians are not Kurdish, nor are the designated Orthodox Assyrians, aka: "Nestorians" Kurdish (as per above),
7) Assyrians now are campaigning to renew their homeland in North Iraq.

on the "cross" i have noticed similarities between early iconography and some of the patterns found here on the images shown. interesting indeed!

good work all.



Steve Price
April 15th, 2014 12:15 PM

Hi Daniel

Thanks for your most informative post. I'm glad you stumbled upon us.

One favor - when you post, please overwrite the word "unregistered" in the user name field with your full name, That way, it will display as the author of the post. Alternatively, if you register and are logged in when you post, that will happen automatically.


Steve Price

Patrick Weiler
April 15th, 2014 08:24 PM

Cat got your rug?
Here is another "pagorbe" or cat's foot

rug, from Burns Antique Rugs of Kurdistan, plate 55, which he notes is from Zanjan in the Eastern Kurdistan Region that includes Sa'uj Bulagh and Senneh. This is one of only two designs which Alheim confidently attributes to Kurdish weavers. It consists of a small bush or a four-petal flower and two leaves on a short stem. Burns says the border is a Kurdish characteristic.

And in another likely instance of mislabeling a Sa'uj Bulagh weaving, here is a page from Hali 150. This rug has characteristics of the Kurdish Sa'uj Bulagh type, with corroded ground and flaming palmettes.

As you can see, it is described as East Caucasian.

Patrick Weiler

Joel Greifinger
April 16th, 2014 03:21 AM

Other hauzis?

In Kordi Without Tears Mr. Alheim does state that:
Two designs are referenced as only being found on Kordi weavings... The second is the hauzi or water basin motif...
As Jenny Housego pointed out in Tribal Rugs (pl. 137) the Kordi hauzi motif is clearly related to a motif common on east Anatolian Kurdish rugs. In his An Introduction to Kurdish Rugs, Eagleton pictures a number of these from Malatya (pl. 88-91). Here's an example from Bruggemann and Bohmer where the middle medallion well demonstrates the affinity:


Alex Wolfson
April 21st, 2014 02:06 PM


With regard to the second image from your last post, the given designation looks to be correct. There is no more than a superficial resemblance to Sauj Bulagh rugs. Neither the borders, format or pile correspond to the type.


I did not quite understand: Are there Uniate and Orthodox Assyrians?
Is the Assyrian rite essentially the pre-Justinian eastern Christian rite, as followed by Nestorius.

Joel Greifinger
May 5th, 2014 04:16 AM

As a postscript to this long thread, let me note that soon after I posted the rug in post #48 that had just been sold at auction, I discovered that it had been published in the 4th (current) edition of Eilands' Oriental Carpets: A Complete Guide (plate #75, p.104).

There the authors, with their characteristic epistemological modesty (a rare trait in rug circles), describe it as "Kurdish Village Rug, late nineteenth century". They continue "its use of color changes from row to row makes an interesting composition. The wool is particularly lustrous."



May 23rd, 2014 05:55 PM

A related rug
Gentle Readers,

The rug Joel recently posted in this thread is very similar to a rug in my own collection. This rug measures approximately 3' 8" x 6' 10" which is smaller than the rug in post #48 and seen again in the previous post. The size of the rug posted by Joel is given as 4' 6" x 8' 4". The top of this rug as it would have been on the loom is finished with oblique interlacing or plaiting and a single row of two-color, two-span twining. My recollection is that Marla Mallett has pointed out that such interlacing or plaiting is finger-woven after a rug is removed from the loom.

On the other end or bottom of the rug there is again a single row of two color red and blue two span twining as well as a simple weft-float pattern. The selvedges appear to consist of two sets of interlaced paired warps. They have been wrapped probably later.

Warps are ivory wool and wefts are also wool but varied in color from very soft red that is almost tawny to brown. There are least two weft shots and sometimes three.

The above image is the bottom right corner of the rug. A small area of repair is seen at the selvedge. There is another small area of repair on the bottom left side.

There is corrosion to the very dark brown wool knotted pile that is the ground between the serrated edges of the variously colored geometric devices each containing a double wrench like device. The corrosion appears deeper in areas with the light red wefts than in those areas with brown wefts. I am not certain if this is an illusion, wear related or some chemical reaction among different wools since my understanding has been that brown corrosion results from a chemical reaction with the mordant. I recall Joel stating that the rug he posted was not corroded. Curious difference that I cannot explain.

I do not consider this rug to be a Sauj Bulagh piece. Once again, my opinion is that this is a piece from Central Kurdistan and perhaps can be called a member of the Jaf type. Interesting to me is the fact that my rug was originally sourced in Syria from a suburb north of Damascus. Since then it has travelled to Beirut and then Japan before landing with me. I have no idea how it got to Syria. The knotting is much coarser than the Sauj Bulagh rugs I know, colors are very good but also different from the bright reds and true purples I associate with Sauj Bulagh. Wefts here are not all red and much lighter in tone where they are used. The back of this rug feels much more coarse and is ribbed, not flat. Wool quality is high but not as soft as most Sauj Bulagh pieces.

I hope you enjoy the images of this rug, it really shines in good light.

Best wishes, Michael

Filiberto Boncompagni
May 24th, 2014 09:06 AM

Hi Michael,

Your photos were way too large for the pages of our forum, obliging the viewer to scroll a lot, so I took the liberty to upload them to our server after a robust resizing – even though I left them quite large for our general standards.

The zigzag lower end finish is interesting. I don’t recall having ever seen that on Kurdish rugs.


May 25th, 2014 02:29 PM

Hello Filiberto,

I hope the sun is shining as usual in Cyprus. Thank you for correcting the format of my photos. Please feel free to do what is appropriate should the problem persist.

The weft float zig zag pattern on the bottom of my rug is seen with some frequency on Central Kurdistan rugs oftentimes examples labelled Jaf. In fact, in my collection is also a long rug with hexagons and the XXX motifs of the type discussed in the related thread to this started by Patrick called Kurdish or Knot. That rug also contains the same weft float zig zag end finish. I have other examples in the bag format that would also be labelled Jaf with this end finish. Occasionally you can see this weft float technique used the followed by a weft brocade pattern - this typically in bags as the weaver transitions to a weft faced plain weave for the back. I consider this weft float end finish another marker that helps one feel more secure in making the attribution to Central Kurdistan, perhaps Jaf. To be even more precise - an end finish with twining followed by the weft float patterning.

I do not recall seeing any Sauj Bulagh rugs with this end finish.

Thank you again for your help with my previous post.

Best wishes, Michael

Marla Mallett
May 25th, 2014 04:53 PM

Hi Folks,

I’m not going to add another unsubstantiated guess at the origin of any rugs on this thread. Without documentation, I suppose the best anyone can logically do is create loose rug design or structural “groups” to locate within vague geographic areas, labelled with equally vague and questionable names. In doing even this, however, it makes sense to not group pieces with only superficial similarities. The last piece posted (in Post number 65) would seem to fit this description—being only superficially similar to pieces posted earlier—for several reasons. But to focus on one aspect with broader applications: This rug provides an almost perfect example of how structural and design features can combine to give us an important clue to design origins and migration. It involves an aesthetic matter that has not, as far as I can remember, ever been discussed on Ttek in any detail.

What I’m talking about is the artisan’s LACK of concern for a satisfactory relationship between the structure and the motifs. It seems obvious to me that the design in this piece was NOT part of this weaver’s heritage--her family’s or her neighbors’--that instead it was an attempt to copy motifs from elsewhere. I say this because the SCALE and character of the motifs and the weave balance are so out of whack. The last, close-up detail photo in Post 65 shows this clearly. The warp sett, along with the coarse scale of the weft elements and knotting would be far better suited to a simpler or larger scale design. Instead, this weaver’s heavy, coarse weave balance has failed to allow her to clearly articulate the small details she attempted. The dominant narrow, multiple sharp-angled outlines, for example, are not suited to her wide warp sett. Nor are the small border motifs. Instead the results are clumsy, weak and awkward—too fuzzy, blurry and crowded for the sharp-edged patterning she attempted. In common parlance, folks tend to just describe products with this lack of clarity as not “well drawn,” without realizing precisely what they mean. It is NOT a matter of a woven design being “primitive,” as plenty of primitive tribal designs are indeed coarse, but still perfectly balanced, strong, and effective. Nor is it a case of inept weaving.

To sum up, I think it logical to assume that the particular design concepts displayed here were very likely foreign to this weaver and originated elsewhere, within a different group, within a different weaving tradition. Examples of poorly coordinated weave balance and design are of course not rare in the rug world, since designs have moved about so frequently. It’s just fortunate that we have examples on this thread featuring the same field motif to allow for easy comparisons. If not for these examples, we might conclude that the design concept in the last rug was the product of a relatively new weaving venture in which inherent problems had not yet been resolved—and the weave balance and designing refined.


Patrick Weiler
May 25th, 2014 07:11 PM

Superficial, out of whack, unsubstantiated.

Right up my alley, "clumsy, weak and awkward—too fuzzy, blurry and crowded...LACK of concern...inept... poorly coordinated".
If I had been the weaver, I would be embarrassed by your critique, but as a fan of Kurdish weaving it is quite instructive. For one, I have never approached a weaving from this point of view - to be able to see how the design is inarticulate due to its unfamiliarity to the weaver and her inability to put the round peg of the design in the square hole of her structural technique as it were.
As for that weft float border, I have a late, clumsy, coarse bag with a design that usually has the central panel twice as large as the top and bottom panels. It also has the weft-float zig-zag border above and below the closure tabs:

It is fairly consistent at the top, but the lower row transforms from the zig-zag into an X-O-X-O pattern right about the middle:

Later in its life, someone added holes in each closure tab, as though they didn't understand how the closure was supposed to work.

Based on your review of Michael's piece, can one reasonably consider this piece as one whose weaver was removed from the tribal milieu of the original design? Or is it just a late example with the design degenerated from the original?

Primitively yours,

Patrick Weiler

Marla Mallett
May 25th, 2014 08:12 PM


Yes…This rug appears to have been produced by a weaver quite “removed from the tribal milieu of the original design” as you have said. It does NOT show the kind of aesthetic degeneration that occurs naturally with time. A group’s carefully refined WEAVE BALANCE simply is not altered so drastically, certainly not while continuing to produce the same patterning. This weaver’s work in using her group’s proprietary weave balance could well have produced quite satisfactory results with other designs…just not this obviously borrowed kind of imagery.


Patrick Weiler
May 25th, 2014 08:49 PM

Thank you for explaining your view of this phenomenon, of designs being transported away from their origin and becoming clumsily executed when woven in a structure not suitable for accurate articulation and reproduction.
And my bag certainly deserves to be "displayed" where i keep it, on the seat of my recliner.
There is also another late version of this "spanner" design (due to its similarity to what we call an open-end wrench) in a Shiraz rug of mine, as the central design in the guls of a 3-medallion carpet.
I hope someone posts some pictures of more attractive rugs to this post. It would be a shame if the final picture was the "rear-end" to this thread.
Patrick Weiler

Marla Mallett
May 25th, 2014 10:09 PM

Since we’re on the subject of design transfers, a couple more things…There’s a huge difference between the natural evolution or transfer of an “aesthetic concept” and the mere copying of motifs from outside a group’s milieu.

One additional aesthetic element in the Post 65 rug which suggests to me that this rug was almost certainly produced outside the circles in which the imagery evolved is the color usage. NOT just the palette. Many of the “related” pieces on this thread display clear systems of color organization and distribution, with careful concern given to color usage that is sophisticated and subtle—an integral element in the designing. In the Post 65 rug, bright kindergarten colors are merely splashed about at random-- every color in the crayon box. There is no concern for relationships between the parts. The frequent color changes between adjoining motifs and their ever varying outline colors contributes to a scrambled, cluttered appearance and actually exacerbates the problems of unsatisfactory design scale/weave balance relationships. There is a stark contrast indeed between the sensitive color usage in the rug of Post 64 and that in Post 65. How can we possibly consider these “very similar”? In the first, the color distribution and usage are a critical organizing element in the design concept; such concerns are missing in the last rug. This kind of aesthetic deterioration is often a strong clue to a design’s origin and migration. If the basic, coherent aesthetic concept is missing, what of significance is left? Just shiny wool?

Another factor to consider is that in thoughtfully conceived rug designs, the relationships between borders and field are complimentary. The attitude and care with which these parts are combined is as important as the actual choice of motifs. And possibly more critical for us than the mere cataloging of motifs. Again, let’s compare the rugs in post 64 and 65. In one, the borders and field are complimentary—in character, color usage and contrasting scale/intricacy; in the other rug, the borders simply add clutter. Again, this is a sign of motifs merely being copied, while the overall aesthetic concept has been ignored.


May 26th, 2014 12:48 AM

Superficial, weak, clumsy and awkward with bright kindergarten colors
Dear Readers,

Marla Mallett has posted what I consider to be unduly harsh and extremely negative comments about the rug from my collection that I provided images of in post 65. I am not going to defend the rug, it can and does defend itself. This is a discussion forum for people interested in rugs to discuss them. This is the second consecutive time, separated by a nearly a year, in which a rug image I have posted has come under harsh criticism from Ms. Mallett whose knowledge of woven structures is well respected by me. Words and tone matter in an internet based discussion forum and when comments as unrelentingly harsh, negative and ultimately judgmental as those written by Ms. Mallett here and the last time I dared to post images of a rug from my collection are made it certainly inhibits, at the least, me and other collectors from posting images of and commenting on other rugs.

As most of you know, I care about Kurdish weavings and have collected, sought to study and understand them since the early 1990s. I have handled many examples and have engaged in many gratifying conversations with rug enthusiasts. Oftentimes there have been disagreements over the merits of a piece or idea but always with some element of constructive criticism and a goal at better understanding a group of pieces. It's a small collector circle that seems to get smaller with the passing of time. I am lucky to count many of the other collectors as friends sometimes for decades. I also have collected a fairly substantial number of Sauj Bulagh rugs including a piece related to one posted previously by Joel Greifinger in from the exhibition book Discoveries From Kurdish Looms that set me on the path to collecting these rugs. I believe that images of that and other rugs would add to the discussion, but under the circumstances why would I or anyone else bother? Given the reaction of Ms. Mallett to the last images I posted, I hesitated to post images of the rug in post 65. Obviously, I disagree that the similarities with some of the other rugs posted is superficial. But even if they are superficial there are positive ways to discuss, compare and contrast the details that advance our understanding and appreciation for all weavings.

Although I have not spoken with Ms. Mallett in perhaps 10 or more years, I have considered and still consider her a valuable resource on matters of structure. In the past she has been quite generous with her time and knowledge in helping me to understand entire groups of weavings. I remain grateful. I am not, however, grateful for the out of hand dismissal of the rug in post 65 on all levels - structural, scale/weave balance, colors and aesthetic concept or deterioration. According to Ms. Mallett all that is left of this rug is shiny wool. Shiny wool? Really, I hope every fan of this discussion forum will make their own aesthetic judgement of the rugs with in this thread. I will simply state this - in our small world - Ms. Mallett is far from the last word on rug aesthetics. What appear to her as bright, kindergarten colors that are merely splashed about at random -- every color in the crayon box might be seen by others as joyful and inspired. I will leave it to each viewer to decide for him/herself with best wishes for creating your aesthetic whether you enjoy the weave balance or imbalance. Personally, I intend to listen to Mozart and look at some brightly splashed colors. Bach just isn't my jam.

All best wishes, michael

Marla Mallett
May 26th, 2014 03:22 AM


I sincerely apologize if you read my comments as a personal attack. In the context of the pieces presented on this discussion thread, your rug simply presented a superb example of a kind of aesthetic/structural problem that has not been discussed previously. It’s impossible to discuss this kind of problem without a concrete example. It just happened to be yours. If you notice, in my remarks, I even referred to the rug as “that in Post 65”, rather than ever saying “Michael’s rug.”

Actually, I was anticipating that someone might now reference the Burns rug at the beginning of this thread, because it raises some similar aesthetic and design transfer issues. Perhaps if I had talked about yours and his together you might not have been so offended. Michael…it’s NOT personal! The comments had only to do with the transfer of motifs and design concepts and the relationship of these actions to formulating useful “groups” of rugs—the subject on which you contributed several remarks.


PS. Michael, one more thing: You have explained why you were offended by my comments, but have not countered any of my specific arguments about the inappropriateness of the motif for the weave balance in your rug and what that says about the design transfer and illogical grouping of these pieces.

You also seemed to miss the point that corruption of a basic underlying aesthetic concept has significance when we try to understand the relationship between the various rugs on this thread. THAT, after all, was my reason for discussing various aesthetic issues--issues which, of necessity, must deal with specifics.

Richard Tomlinson
May 26th, 2014 03:32 PM

hi all

colour arrangement has always fascinated me - i think too much emphasis always seems to be placed on simply 'colour', without enough attention to the aesthetic qualities derived from correct colour arrangement.

finding a balance between 'stiff' colour combination and 'kindergarten colors that are merely splashed about at random' is a difficult thing.

rug 65 exudes a splendid array of wonderful colours, but i would have to agree that the arrangement is rather shambolic.

i would agree with marla that colour distribution in #64 is more pleasing to the eye. it is however, difficult to ascertain how good the individual colours are.

i had never though of colour arrangement in the way marla talks of so this thread is illuminating to say the least!

roll on !!!


richard tomlinson

Marla Mallett
May 26th, 2014 06:51 PM

Hi Richard and all,

Thanks for your comments. Yes, there can be a thin line between color usage that is intuitive, loose and carefree, and results that are chaotic. Or, on the other hand, between excellent compositions and results that are stiff and dull. The best artisans are normally well aware of the extent to which color distribution can organize the elements in a design. There are SO MANY different ways in which this is done. A creative artist may use this element of designing to create tension in a piece, to add focus, to create a soft harmony, to unite diverse elements, or to add variety within otherwise tedious repetition. Only an unthinking, insensitive individual plops colors down at random or plugs them into repetitive pigeonholes without carefully considering their relationships and distribution. The same is true of the arrangement of VALUES—the placement of dark and light elements.

I think it is a mistake to believe that the more colors the better. An excessively large number of colors in a rug do NOT guarantee a superior aesthetic result. Instead, they typically increase difficulties for the designer trying to produce a coherent work, and increase the likelihood of aesthetic problems. When the prime design dictum “Economy of Means” is ignored, excessive, non-essential elements dilute the expression with unnecessary diversions. This important design consideration applies to color usage as well as to every other aspect of composition.

When there is no clear hierarchy of values in a rug’s design, and merely lots of unrelated strong colors tossed about, the resulting chaos is unsettling. When no neutral tones are present to serve as foils for especially vibrant hues, their effectiveness is compromised--even while producing sensory overload. There is a huge difference between lively, vibrant color, and garishness. Much too often in the rug world folks talk of “good” colors and extol the merits of “fully saturated” colors, without realizing that tints and shades (the light and dark versions of each hue) along with intermediate hues, are extremely important in composing with color. The failure to understand the realities of color usage explains, in large part, the limited success of some modern day natural-dye rug “revival” projects. When an expert dye master produces wool yarns in only intense primary hues for the use of co-op weavers, and the more subtle light, dark, neutral and intermediate tones of traditional palettes are unavailable to those weavers, it’s hardly a wonder that warehouses full of the resulting brash, garish products are left without buyers. Unfortunately, no one seems to diagnose the problem.

Ah…so much for my soap box. I needed an opportunity to expound on this one!


Joel Greifinger
May 26th, 2014 10:28 PM

Seemingly similar
Hi Marla,

Since I'm hoping to coax you to further elaborate your thoughts on the relationships between color, weave balance and design transfer, I'm tossing some pictures of another Kurdish rug with the same field design and many colors into the mix. I would have rather straightforwardly presented this as quite similar to the rug that Michael posted in Post #65 (except for some obvious condition issues), but I am wondering how you would assess their similarities and contrasts.

As you wrote earlier, discussing aesthetic issues entails "dealing with specifics" and discussing rugs without discussing aesthetic issues seems to me to miss a great deal of the point. Thus, any commentary on this rug about what you think does and doesn't work and why is most welcome.


Marla Mallett
May 27th, 2014 03:48 AM

Hi Joel and all,

It is a truly interesting example of the serrated medallion/lattice field rug that you’ve now posted. Feeling a little gun-shy, I’ll see if I can deal with the issues without referencing Michael’s rug. We can make interesting comparisons with others on the thread.

First, let’s look at the specific question of design/structure problems or accommodations, and then later the more complicated matter of overall aesthetic concepts and design transfers.

We need to compare your photo detail with that in the detail from post 48 (below). In my opinion, that rug (also shown in Post 64) seems among the “purest” and least degenerate examples on this thread. Anyway, it’s convenient to have a detail of it. The repetitive field motif is a standard serrated medallion—actually a version of a slit-tapestry motif that’s been turned on its side so as to be more easily knotted. Each outline is formed with a series of verticals and diagonals. (Tapestry versions use horizontals and diagonals.) You can see most clearly in the red outline that each serration was constructed with just two knots diagonally aligned between two verticals. The combination of the warp spacing, the interval between knotted rows, the character of the materials, the coarseness of the knotting and the length of the pile, needed to be perfectly balanced to allow crisp articulation of this size motif. MOST of the rugs on this thread (with this serrated medallion/lattice field) most likely follow this kind of motif construction, though without detail photos, we can’t be sure.

In your piece, Joel, we see a quite startling accommodation. Perhaps realizing that she might have difficulty articulating this same form clearly in her weave, this woman has adopted another practice entirely. Her solution gives only the “illusion” of serrations. Instead, each medallion in your detail is formed by a thin, straight diagonal outline intersected by thin vertical lines. Single knots of contrasting colors have been placed in the intervals where knots normally forming the serration diagonals would have been. She has also crowded the motifs so that their points are nearly touching. It’s an ingenious solution, an arresting feature substituted for what might have been a series of fuzzy, awkward, less than perfect serrations.

Without handling your rug, I cannot say with any confidence exactly what about the weave balance in this rug might have made this design device preferable to ordinary but perhaps difficult serrations. I would feel fairly certain, however, that the rug was NOT made by weavers who were used to weaving this design in the conventional manner.

Since this is getting long, and my comments on the broader topic of overall design concepts and transfers are likely to be longer still, I’ll make that a separate post, maybe tomorrow.

Thanks, Joel, for introducing this example.


May 27th, 2014 06:43 AM

In Defense of Kurdish Aesthetics
Dear Readers,

One person's soapbox is another person's diatribe.

With 5 posts plus edits from Ms. Mallett it's hard to keep up with all the effort to dismiss as merely shiny wool what is after all just a rug. I do apologize for posting images of a rug that if it bears similarities to other rugs in the thread does so only superficially in her judgment. I will observe that, to the best of my knowledge, Ms. Mallett has never handled this rug in the flesh and the judgments rendered are all based on photos and my descriptions of the end and side finishes. But all of that does little to advance the discussion.

There are some real nuggets in all of this, however, that do present a real opportunity to advance discussion.

First among these is the statement in what must be post 70 - May 25 at 7:12pm regarding "a group's carefully refined WEAVE BALANCE simply is not altered so drastically ... this weaver's work using her group's proprietary weave balance could well have produced quite satisfactory results with other designs ... just not this obviously borrowed kind of imagery." This is breaking news to me - that weaving groups have proprietary weave balances and certainly merits documentation.

Second, the weaver of the rug in post 65 was produced by a weaver quite removed from the tribal milieu of the original design. Perhaps but then what exactly was the tribal milieu of the original design? If asked I would have suggested that the primary design is pan-Persian which is my way of saying one can find rugs with this design as part of an overall pattern or as secondary elements in a variety of rugs woven in a variety of places and that the tribal milieu of the original design would be difficult to even guesstimate. But i would also say that the weaver of the rug Joel posted (post 64) images of is likewise woven by a weaver quite removed from the tribal milieu of the original design.

Turning to aesthetic concepts and design transfers. Both I and the weaver of the rug i posted images of are also accused of corruption of a basic underlying aesthetic concept of color, color placement and distribution. "Only an unthinking, insensitive individual plugs colors down at random or plugs them into pigeonholes without carefully considering their relationships and distribution." The results are condemned as chaotic and garish. Perhaps. I cannot state what went on in the minds of these weavers. However, I can state that there is another perspective quite outside the milieu of proprietary weave balances and careful consideration of color, color relationships and their distribution. In 1992 Mehrdad R. Izady, then at Harvard, published his book "The Kurds, A Concise Handbook." This book was influential in collector circles and certainly with me. In chapter 10 beginning on page 237 of that book Professor Izady described the Culture and Arts of the Kurds. Citing the excavations at Shanidar and Solecki's excavation reports, "Shanidar: The First Flower People" (1971) Izady remarks on the preoccupation of local Kurdish workers with wild flowers as referenced in the excavation reports. Solecki wrote "They attach wild flowers to everything and wear riotously colored costumes." Izady then states: "This national preoccupation with flowers, or rather their colors, is one of the strongest traits of the Kurdish national culture, consistent in every niche and corner of the land and in every segment of the society - more uniform and pervasive than any other single national characteristic or trait." Later in discussing local costumes the term used is "exaggerated colorfulness." p. 256.

One can argue and opine endlessly on aesthetic concepts, design transfers and weave balances. One can even deconstruct pieces, trivialize and toss them in the trash. Personally I am rather saddened by this reality since I enjoy looking at the rug in post 65 and seeing a field of Kurdish wild flowers.

Joel, I would also have considered your latest rug as similar to the rug in post 65 except that I recognize the weave of the rug and I would think it comes from a group closer to the Bijar area. The minor borders are also consistent with this in my experience.

All best wishes, Michael

Dinie Gootjes
May 27th, 2014 11:59 PM

Kurdish cooperation?
Hi Michael,

I hope you don't mind me saying something about your rug.


There is corrosion to the very dark brown wool knotted pile that is the ground between the serrated edges of the variously colored geometric devices each containing a double wrench like device. The corrosion appears deeper in areas with the light red wefts than in those areas with brown wefts.
Looking at the picture, I think the area with the deeper corrosion is roughly the top half of your rug. And not only are the wefts and the amount of corrosion different, the design elements are more elongated there, and the colours change.

To me the bottom part truly looks like a field of colourful flowers, but with several same and similar elements placed on diagonal lines, giving the design a certain balance, all held together by the frequent use of a closely related yellow and light green in the serrated outlining. I love it.

In the top half, not only are the elements irregularly elongated, the colours are used differently, and the light green is cooler. There is one horizontal band with three different blues as background, the top row is also dark, the cheerful balance of the bottom half is missing in my opinion.

Taking this all together: different wefts, different browns, different weave balance (??? Thicker wefts or thicker wool???), different use of colours, different greens, does this not look like a rug started by one weaver, and finished by another? The second one less experienced or unfamiliar with the design, or just less able to work with an all-over design using so many colours. This of course is all based on a couple of pictures. Does this look plausible "in the wool"?

To save everyone some scrolling up and down, here is the rug once more.


Rich Larkin
May 28th, 2014 04:46 AM

Hi Folks,

We can be certain about one thing for every rug ever woven: It came from a finite set of circumstances. The tantalizing and elusive part is, we don't know what they were for just about all of them, though the cursed denizens of this site apparently can't resist trying to know them.  Some are quite astute in going about the quest, such as the observant Dinie in the prior post.

Speaking as one who has been looking at and thinking about rugs for many years (not especially profoundly, much the pity ), I find Marla's notions as outlined in this thread to be novel and intriguing. I'm not sure I'm convinced on all fronts. My first reaction to Michael's rug was that it is boffo, and I still think so. I may well be irretrievably brainwashed into finding excellence anywhere there are saturated colors of many hues deposited in very shiny wool. However that is a side issue, the main one being whether one can, properly informed, assess the ambient circumstances of a particular weaver in the manner Marla has attempted to do here. I'd like to hear more about it, and from as many well-informed participants as possible. It really doesn't have to be touchy. It's just opinion.

Marla says,


Yes, there can be a thin line between color usage that is intuitive, loose and carefree, and results that are chaotic. Or, on the other hand, between excellent compositions and results that are stiff and dull. The best artisans are normally well aware of the extent to which color distribution can organize the elements in a design. There are SO MANY different ways in which this is done. A creative artist may use this element of designing to create tension in a piece, to add focus, to create a soft harmony, to unite diverse elements, or to add variety within otherwise tedious repetition.
Of course, it is likely that most weavers are not gifted creative artists, though the idea is persuasive that many if not most discrete weaving communities with a long tradition will have collectively brought their traditional designs into a balance with their weaving decisions and practices. It is very interesting to consider that one might be able to discern where this standard hasn't been upheld, and from that, achieve some understanding of the circumstances of the weaver as she sat at the loom. My greatest fear here personally is that Marla's got it right, but the whole thing will be over my philistine head. I'd like to hear more about it anyway.


Patrick Weiler
May 28th, 2014 07:29 PM

Obscure references

You have asked "whether one can, properly informed, assess the ambient circumstances of a particular weaver".

To which I can only reply:
"Well, here's another fine mess you've gotten me into."
Laurel and Hardy, 1930.

"There's something happening here. What it is ain't exactly clear."

Buffalo Springfield, For What It's Worth, 1967

James Burns, in the book Antique Rugs of Kurdistan, delves into what he calls Central Kurdistan. He mentions that it is the smallest in size of all the regions of Kurdistan and "the most exposed to outside forces".
He identified the main tribes of weavers as Jaf, Diza'i and Herki, but "There were also hundreds of smaller Kurdish groups that produced weavings."
To complicate your search even more, "The vast numbers of nomads who coveted and benefited from Central Kurdistan's relatively warm winter climate, with its rich winter pastureland, caused the cultural traditions of Central Kurdistan to become intermingled with those of Eastern, Northern, and Southern Kurdistan."
The area not only included all the major Kurdish dialects, but also all the different religions practiced by Kurds were there, including the largest community of Kurdish Jews in Kurdistan - outnumbering even Christians in the region.
All this complicates narrowing the focus of Michael's rug to one specific tribe, the Jaf, as he suspects. Burns shows a rug from Eastern Turkestan, plate 49 (shown as rug #2 in Joel's post #1), that is Sa'uj Bulagh, with this "spanner" design which Burns calls ashik, but the guls are more flat than tall. And he notes the rosette and bar border as being a Jaf signature design.
So the border design of the Jaf migrated to Eastern Kurdistan and the spanner/ashik migrated to Central Kurdistan to land on Michael's rug. My inclination is Jaf also, mostly due to the colors which Burns points out "Jaf weavers utilized the finest wool dyed with rich, bright, harmonizing colors."
It also could have been woven by any of the Kurdish tribes who migrated to Central Kurdistan, picked up the design and produced this Boffo rug.
The Jaf, according to Burns, mostly wove bags and kilims because large carpets would have hindered migrations and the larger rugs they did weave were used by them rather than being sold. All leading to a likelihood that this may be a Jaf rug based on a Sa'uj Bulagh pattern and woven for personal use. This might explain the phenomenon Marla has proffered, the Structure/Motif Disconnect.

Michael noted the widespread distribution of this motif. I have a Shiraz rug with it, (See post 15 of Salon 25) and here is a Baluch version:

I will open a separate thread devoted to just this motif and see what other regions pop up.
And....Salon 25 popped up! Thanks, Joel. Your memory is longer than mine. So, to top off the Baluch version, here is a little soumak piece with the sole medallion including S figures surrounding the spanner:

Patrick Weiler

Joel Greifinger
May 28th, 2014 08:40 PM

Been there, done that

I will open a separate thread devoted to just this motif and see what other regions pop up.
Hi Patrick,




May 28th, 2014 08:56 PM

Different hands
Dear Dinie in particular and Readers,

Of course, I welcome Dinie saying something about my rug as I welcome, and we all welcome, discussion of these pieces from all perspectives. If I was not open to comment, why post the pictures? Moreover, it has been my experience that no one person or perspective has all the answers. My understanding and appreciation is enhanced through collaboration and consideration of the information and insights of others. I would like to believe that this is why collectors get together and why forums such as Turkotek exist. Certainly this also extends to the important insights and perspective of Marla Mallett who I respect not only to be a serious weaver and scholar of weaving and woven structures but also a generous and passionate person.

Dinie's thoughtful and observant post raises the question whether it is plausible that this rug was woven by more than one person. I have asked myself the same question. The best answer that I can give based on in the wool handling is; yes, it seems plausible. I will try to make some photos of the back when the sun comes back - assuming it does at some point - but I can tell you that there are places where the weave looks and feels different. But it is hard to tell whether this is just the eye playing tricks with the color changes or not and really there are about six color changes and places where the weave seems to change. I will try to enumerate them so that Dinie and anyone else who wishes can look at the areas of change and judge what other results might follow. When I first saw the rug, i did not put much stock in the wefts changing - there could many mundane reasons for this. But then as I started to water the flowers my thinking started to ask the same questions Dinie has asked. Among the questions raised in my mind was whether this could have been a mother/daughter project.

So in lieu of more photos I would say that the first weft color change occurs at the first incomplete lozenge descending from under the border in the middle of the diamonds that are in each lozenge. Visually there appears to be a line where the wool changes in the photo and you can even discern two knots of ivory pile on the left side in these lozenges but this is not the place where weft color changes. The change is higher just at the center of the lozenges. The weft color changes from from to light red, almost tawny.

The second place where weft color changes is near the bottom of the second row of complete lozenges (or near top of third row, if you prefer). If you look for the two blue/green ashik like devices within the inner border (I mean the inner of the two main borders) that are woven on red this change occurs almost exactly in the center of the ashiks. If you look carefully to the red frame that separates the ashik inner border from the field on the right side you can see two knots in light green (may appear ivory of your screen) that mark the exact spot of the change in weft. Here i can also state that you can almost feel the change in weave with your hand from the back as if woven with different hands. Here the wefts seem mixed light red and brown.

Also two points to consider are, first, that the red frame that separates the borders from the field is not continuous around the entire rug. Second, in the complete lozenge on the right of the third row of complete lozenges) there is a color change in the the bottom arm of the vertical wrench. It happens only here. And also the lozenges appear elongated.

Color change #3 is just below the place where the color change in one wrench arm appears. Using the ashik devices again, it goes from the green ashiks on purple and across the top of the square boxes at middle of each lozenge. You can also see that here on the right the red frame between field and the ashik border returns. Your other reference here is the green lozenge with the two ivory wrenches. The wefts in this area are light red.

Color change #4 is at the bottom of this same row lozenges or better near top of next row of lozenges since the color change is just at the bottom of the top wrench arms. The weft color here is light brown. Lighter brown than elsewhere. Here the weave looks and feels different again - I would say cleaner and smoother but this only lasts about 20 weft rows. It goes back to light red at color change #5 and then color change #6 is the last row of complete lozenges and beyond - back to brown.

I cannot explain it but i would agree with Dinie that it is plasible that more than one person wove this rug. And about 2/3s of the way up the rug there appears to be hands at work different from those below and above. Probably this raises more questions than answers but these things sometimes reveal themselves only slowly.

Thank you for your interest, Michael

Marla Mallett
May 28th, 2014 09:51 PM

Assorted Aesthetic Issues Continued
Hi folks,

It is of course possible and reasonable to approach rug studies from various perspectives. For me, cataloging and taxonomy is pretty boring, and especially pointless when there is insufficient documentation to allow credible attributions. I realize that others find these puzzles intriguing, however. My interest lies more in determining what attitudes are responsible for producing a work of art, instead of just adequate floor covering.

I’d like to demonstrate how these studies can overlap. Artistic products can reflect either a group aesthetic, or those of individuals. Or both. Thus when deciphering attitudes toward design construction and artistic expression, not only is connoisseurship involved, but comparisons should assist thoughtful individuals engaged in speculative cataloging.

This is not the place for outlining the basic tenets of Aesthetics 101. I would like to merely consider some different approaches to design construction revealed by specific rugs on this thread.


First, let’s clear out the underlying clutter.

NICE WOOL, of fine quality, properly selected and processed, is not an aesthetic concern. It is a BASIC REQUIREMENT for a satisfactory woven rug.

COMPETENT CRAFTSMANSHIP is not a debatable aesthetic feature. It is also a BASIC REQUIREMENT for a satisfactory piece. This requires both a proper, carefully-refined weave structure and competent execution. It requires the use of design material suited to a community’s traditional weave. If the kinds and sizes of motifs included cannot be well articulated and presented effectively because the weave is either too coarse or too fine, a knotted-pile piece can hardly be satisfactory. This was my major concern with the Wendorf rug in Post 65 that I thought needed discussing. Handling such a piece is not necessary to recognize the problem. Fine quality, sensuous materials may temporarily distract us, but they do not compensate for either inferior craftsmanship or weak designing. Whether dealing with knotted pile or one of the several distinctive tribal flatweaves, no credible critic or connoisseur of woven art can ignore the requirements of each structure and each different “weave balance.” To cite an extreme example: if one cannot understand why designing must be different for delicate Senneh rugs and robust Kazak weaves, further critical analyses are probably useless.


Specific MOTIFS used by any weaver have less to do with a group’s aesthetic than the ATTITUDE she displays toward the imagery and the use of those motifs. The SOURCE of the motifs may be irrelevant. This is especially true with figures as simple as those featured in this thread’s serrated medallion/lattice rugs. If a person saw one of these details in a rug in the market and decided to use it herself, she did not need a model or cartoon to copy. She was most likely to just use the simple idea in her own manner—in a way compatible with her community’s standards. In other words, it is a matter of STYLE that makes works from different groups or weavers distinctive and identifiable, not particular motifs.

For the sake of simplicity, I’ll focus my comments on the serrated medallion/lattice rugs on this thread, and try to deal with them as specifically as possible. We are limited, of course, by having only small photos with uneven color rendition. We should, however, be able to recognize a few significantly varying aspects of design—and thus different aesthetic attitudes. Unfortunately, since this thread is now so long, I suppose I don’t dare copy the long string of appropriate photos to accompany my remarks. I realize this will necessitate bothersome scrolling back and forth.


It is immediately obvious that the several different weavers using the common serrated medallion/lattice field motif had quite different ideas about appropriate border systems. The character of differing border/field relationships may sometimes be more important than the actual motifs chosen.

The border system in the Post 64 rug is unassuming and complementary: two simple zig-zags enclose a border of small ashik motifs that reflect the forms in the field. A strong, coherent expression results, with no distractions. The field remains the clear focus. In related examples, pairs of simple borders also enclose a slightly wider and more complex band. This repetition--the pairing of borders--creates a simple, compatible and unobtrusive “frame.”

Joel’s rug in Post Number 77, also displays a pair of identical borders. Made of complex, gracefully undulating designs, they also enclose an ashik border that reflects both the characteristics and palette of the field design. Though this was a difficult combination, by minimizing the number of diverse elements while adding contrasting complementary, repetitive forms, a quiet and refined elegance was achieved. As in other excellent examples, significant VALUE contrasts, (light or dark contrasting colors) are important in separating and enclosing both field and borders. It is a masterful achievement.

In the MacDonald piece of Post 49, the borders seem too “busy” for the field. It is surprising to realize that these are constructed with the same sequence of motifs as in Joel’s Post 77 rug. They simply add nothing here but a sense of overcrowding.

The last rug shown in Post 48 has borders that seem to overpower a field severely reduced in size. The three borders of identical width leave one wishing for more interesting proportions, more reasoned color usage and a bit more simplicity. The differing colors of the first and third borders that carry the same motifs are jarring in this instance. Though the field is attractive, the borders are an unwelcome diversion.

Another rather odd border solution: The second rug in Post Number 5 simply displays three borders identical except for a slight color change in the center. It’s a bland and fortunately rare approach, though the borders’ intricacy contrasts in an interesting way with the bolder field pattern.

In some examples, the “framing” function of the border systems seems especially weak, with features so unrelated or confused that they compete with the fields for our attention. In the Burns rug of Post Number 1, three totally unrelated borders show no connection with the field design; nor do they complement it. These border elements are awkward in their proportions, emphasis, color and execution. Jarring indeed. Joel’s first rug in his first post also displays three different borders, each unrelated to the field design, but at least these are compatible in their coloration. The inner border provides a suitable separation between the other border elements and the field.

This matter of composition--of the overall layout and relationship between borders and field--thus varies widely among these many rugs that seem at first glance to have “similar” fields. The attitudes of the designers toward the total image and the style of the total composition differ enough to suggest that the rugs were not produced in close proximity or by a single group.

(Continued on the next post)

Marla Mallett
May 28th, 2014 09:53 PM

Assorted Aesthetic Issues Continued

Again limiting my comments to differences among the serrated medallion/lattice rugs shown in this thread, let’s look at the sometimes subtle but distinctive differences in how the weavers approached the distribution of color.

Let’s start again with the Rug in Post 64. Color usage in this one is extremely restrained. Bands of identically colored medallions in a limited palette create a calm, serene effect. Colored outlining is restrained and subtle, consistent across each row of motifs. Value contrasts are restrained: whites are limited to carefully placed interior motifs, and light touches reappear as small balancing accents elsewhere. Field colors are repeated in the ashik border, their vertical sequences providing variety, though they are carefully and symmetrically arranged from one side to the other.

The second rug in Joel’s Post Number 5 also has serrated medallions arranged in bands, but the arrangement is not so obvious. Instead of continuing the rows with half medallions at the sides, this weaver instead filled these spaces with small checked motifs. Thus we tend to see the field design as groups of two and three medallions. White (?) motifs in the upper center are balanced by pairs of white interior motifs in a couple of other places. Though not subdued, colored outlines are consistently the same on adjacent medallions.

Bob Kent’s rug in Joel’s first post comes close to dissolving the “bands” of motifs, by giving the field an almost all-over blue tonality. This weaver has created interest primarily by varying the character and color placement of the interior motifs, altering those seemingly at random. In a few places she has lined up similarly colored interior motifs, while in other areas, the changes are startling. She apparently was not bothered at all by an erratic distribution of the strongly contrasting whites.

Joel’s ninth rug on his first post also comes close to abandoning the idea of bands of medallions in the field. The colors are muted, but a close look reveals that pairs of similarly colored medallions alternate with sets of three—each of these consisting of pairs separated by a slightly different model. It’s a very carefully constructed arrangement. In this rug the various medallion outlines have disappeared, as well as background spaces between the medallions, to be replaced by consistent light blue (?) diagonal serrated bands. This forms a bold lattice throughout the rug. In fact, the emphasis here has shifted from the serrated medallions to the lattice itself. The overall effect is one of extreme and careful regularity. Any possible tedium is relieved by color variations within the small interior motifs, although the colors within each pair are the same.

The MacDonald piece of post 48 likewise shows colored medallions arranged in pairs—in MOST places. The regimentation is broken by occasional isolated, completely different examples that pop up randomly. Color is arranged in the borders quite erratically.

With the MacDonald piece in Post 50, we see a much different attitude toward color organization. Each serrated motif is colored separately, the varying hues bouncing around seemingly at random. There seems to be no organizing principle, except for the desire to create a collection of separate entities. Outlining colors vary, changing from one medallion to the next, contributing to a haphazard and overly “busy” effect. Strong value contrasts are used carelessly, with the white medallion and partial medallion at the lower end producing a jarring note.

The Burns rug in Joel’s first post displays a similar attitude toward color organization, though the severely compressed motifs and strong colors create a different effect. It is an assemblage of totally separate parts. We find ourselves searching to see if any two are alike. Where the medallion colors are the same, the interior motifs differ. “Cohesion” exists only in the lack of cohesion. With its brash use of bold primary and secondary colors, the effect is of course far removed from the serenity of several other examples on the thread. If one enjoys sensory overload, it’s here indeed. One strange feature: considering the character of the field, it’s difficult to understand the weaver’s reason for including the outside light blue border, a detail that is out of character with everything else in the rug.

Finally, Joel’s rug in Post 77 displays an approach to color organization that also isolates individual motifs, though the effect is radically different from the previous two examples. Variously colored outlines emphasize the isolation of each motif. Though there are interesting contrasts of dark motifs on light medallions and then the reverse--light on dark--the warm color palette assures a sense of unity.

None of these aesthetic approaches is inherently superior. They just display different attitudes toward constructing a design with the same basic serrated medallions. Attitudes toward color distribution play a major organizing role and the simple motif itself is almost irrelevant. Instead, the overall impact is important. With these rugs, if we consider the design concepts—matters of style--along with technical details, we might indeed begin to see some legitimate “groupings.”


Joel Greifinger
September 5th, 2014 06:24 PM

Seeing is...
I'm going to swerve from Marla's extremely interesting reflections on aesthetic issues back to the questions raised at the beginning of this lengthy thread. I began there by trying to get a handle (so to speak) on the shifting historical and contemporary referent for 'Sauj Bulagh' applied to certain Kurdish piled weavings.

Assuming that we've come to the end of the discussion for now, I figure it's best to cap things with some additional 'eye candy'. This rug seems to fit all of the criteria that have been posited for the Sauj Bulagh designation. In addition, it fulfills Justice Potter Stewart's justly famous epistemological guide (in his case, for recognizing obscenity) "I know it when I see it."

In terms of antecedents, the bold presentation of the field devices is reminiscent of this rug (shown earlier in the thread) that was published in Oriental Rugs from Atlantic Collections (1996) as "Shahsavan Rug, northwestern Iran".

I think these days it would get called Sauj Bulagh because, well...I know it when I see it.