Posted by Richard Farber on 11-27-2002 12:33 AM:

Where did the motive go ?

Many thanks for the well researched and documented essay.

I understood that the floral motive which you for the sake of continuity retain the 'tuning fork' name appears in the carpets shown which were made before 1850. You see their 'root's in earlier textiles and carpets [although it would be nice to see some images of the motive in earlier pieces]. Perhaps you or the public have information as to what happened to this motive. What did it evolved into. Did it just disappear?

thank you

Richard Farber

Posted by Guido Imbimbo on 11-28-2002 12:39 PM:

Floral form

Dear Richard and all,

in the Salon we state that the "Tuning-Fork" is an elegant and effective rendition of a floral motive with calyx, stem and opened petals. These floral motifs are organized along vertical lines where also other simpler flowers appear as filling device along those lines.

Alberto Levi believes that the vertical lines are derived “from the shape of the watercourses” that would "have been extrapolated from the early garden iconography".

While I agree that the "Tuning-Fork" composition is rooted in the Persian garden iconography, I doubt that the vertical lines are directly connected with the watercourses that have been used in a specific and well known sub-group of "garden carpets" (see for example plate 91 in the Sphuler book on the Berlin Museum).

I believe that the vertical lines may represent just floral tendrils and actually along these tendrils we can find some knots of different size that again recall floral motifs.

Of course I am aware of the dangers and pitfalls of this kind of "represantational" analysis of carpets design. I also reckognize the limitation of exercises based on a simple "design" comparison that John Howe as correctly pointed out in another post.

I just wanted pointed out one typical feature of the early Persian floral iconography that is represented by "repetitive" floral schemes. These schemes are constituted by flowers, shrubs, etc. that are organized, with little variations, in a repetitive mechanism often using grids, compartments or lattice frameworks. The rendition of the "organizing" devices is explicitly floral and often recall often tendril forms. See for example plate 107 of the same Berlin Museum book.

We know that these repetitive floral structures had a vast influence in later carpets production in many centres that go from Persia to Caucasus. The latest rugs present an increasing geometricalization of the original scheme. In a previosus Salon I even argued with Wendorf that the apparent geometric forms of some Kurdish rugs are a particular interpretation of this original floral frameworks.

I believe that the "tuning-fork" structure belong firmly to the original Persian tradition. It is interesting to note that in the "tuning-fork" rugs the repetitive framework is organized on vertical lines. The "organizational" device would be represented by vertical tendrils.

Where did the "tuning-fork" go?
Well, first of all in the Salon we argued that it went to ... Group B carpets where the motif became more geometrical and rigid.

Secondly, we noted (see note 1 at the end of the Salon) that the "tuning-fork" device is very close to a more simplified "trefoil" motif, often organized in a lattice scheme, that is found in others Persian weavings.
An example of the "trefoil" scheme is published in plate 295 of the ORAC book but there are also other carpets in the Wendorf exhibition Kurdish carpets in Washington that is listed in our references at the end of the Salon.
Finally there is a pair of Kurdish bags that belong to my friend Dante that have been discussed on Turkotek a couple of years ago.

I know that this is only a partial answer but I hope that it helps to the discussion.



Posted by Richard Farber on 11-29-2002 12:54 AM:

Dear Guido Imbimbo,

thank you for the detailed reply.

I am trying to understand the phenomena that you have described -- out of a body of floral motives a new variant appears that is so different that researchers [or dealers] have given it a non floral name. The motive appears in a limited number of examples in a restricted geographic area and then within a very short period of time disappears with seeming to have left any [or very very little ] resonance in the continuing expression of Kurdish weavers let alone weavers of other groups.

I have been trying to compare this to what I know of the history of motives in music and have been thinking about closure motives. Yes I can think of some motives that have come along been used in a small area and disappeared - particularly in the middle ages. I have been thinking about why this is so.

I wonder if you would be prepared to say something on why the use of this motive did not expand in the Kurdish community let alone other proximate groups and why it so quickly died.

I also would greatly appreciate some images of the Persian motives that you think are those that were the precedents of the motive in question.

Thank you

Richard Farber

Posted by Guido Imbimbo on 11-30-2002 06:38 AM:

Kurdish Trefoil Carpets

Dear Richard,

here I have included some pictures of three Kurdish carpets that have been exhibited on October 23rd, 1999 at the Keshishian Gallery. The rugs present an overall deisgn in the field with repeated floral design that has been labelled "trefoil".

I believe the relation between the "tuning-fork" motif and the "trefoil" design is very strong. My guess is that the three "trefoil" carpets are somehow later than the rugs in the Group A discussed in the Salon. The difference in the degree of sophistication in the organization of the design and the execution of the details between the two groups of carpets make me guess that the "tuning-fork" rugs in Group A are older. Michael Wendorf may confirm or not this perception.

The evident relation between the "tuning-fork" and "trefoil" rugs confirms to me that the "tuning-fork" motif is definitely a floral one. I am convinced that the design is rooted in the early garden/shrubs persian iconography that is organized in repeated structure. In another thread John Howe have advanced the question if the "tuning-fork" motifs can be seen as species of "animal trees".

Best regards


Posted by M. Wendorf on 11-30-2002 10:11 PM:

Floral motif


I believe that my friend Guido must be correct when he writes that "tuning fork" motif is a floral one and that it is part of a Persian tradition (rather than, say, Kurdish). I tend to think of the vertical lines as stems rather than tendrils.

I also agree with Guido that Alberto Levi was not correct when he attempted to connect these stems or tendrils to from the shape of watercourses ... extrapolated from the early garden iconography. The Persian garden is a quite formal, even architectural. No doubt, watercourses played a central role in the design of these gardens. That said, I am not aware of any formal Persian garden plan involving the use of vertical watercourses that approximate anything even remotely resembling these vertical stems or tendrils. Moreover, there are well documented formal garden carpets that quite clearly show these formal Persian gardens in some considerable detail, none bearing any relationship to these stems or tendrils that I can readily discern.

By the same token, there is quite clearly a fairly large group of Kurdish carpets that have design motifs drawn from the formal garden carpets, many of which were woven in Kurdistan, and the formal Persian garden. Many of these rugs have fantastic color and wool. Whether this amounts extrapolation or whether the motifs themselves are iconographic is difficult to know for sure. However, I am quite certain that none of the rugs Mr. Levi illustrates are earlier than the 19th century and that they are not what one could fairly call Proto-Kurdish.

Guido is also careful to call the floral tuning-fork motif as being part of the Persian design tradition rather than more narrowly Kurdish. I think that this is also correct and a small but important point. We might even say "Pan-Persian." I also think the relationship to the "trefoil" motif is intriguing and part of the same Pan-Persian tradition. I have not handled the rug Levi illustrated in his article. But I have handled all the rugs Guido illustrates above and several "tuning fork" rugs. In addition, I have had several related bagfaces including one just like the pieces in the Dante collection. I cannot say that one group is older than the other. Among the trefoil long rugs, the rug on the far right in both illustrations seems very old - at least as old as any tuning fork piece I know of. The other two pieces both have good age as well. One thing I am quite confident of is that the borders on these long rugs is related to that found on many formal Kurdistan garden carpets. I call it rosette and shrub. See Spuhler plate 91 referenced by Guido as one example.

John Howe's idea about the animal tree is an intriguing one that I had not considered and that may be worth further exploration in connection with these motifs.

Thank you for the Salon, Michael Wendorf

Posted by Patrick Weiler on 12-01-2002 12:18 PM:

You can't see the Forks for the Trees


Is it possibly the branches of trees that Levi is referring to when he relates the "tuning fork" to garden carpets, not the water course itself?
Later versions of garden carpets seem to show tendrils entering the "plots" of the garden from the main water courses. Earlier versions show that these tendrils are actually trees with tendril-like branches.
A later version is shown here:
The tendrils enter the plots diagonally, then venture vertically, similar to the iconography on the familiar Tekke Engsi discussed by John Howe in an earlier salon. It is easy to suspect that, to Levi, these diagonal tendrils may have represented not trees, but water coming from the main channel to irrigate the plots of the garden.
This next link shows, at the very bottom of the page, an earlier version of a garden carpet, with some of the trees entering the plots in a diagonal orientation:
Most of the trees enter vertically or horizontally, but only the diagonal type survived in the later versions of garden carpets.

Patrick Weiler

Posted by M. Wendorf on 12-01-2002 07:16 PM:

garden carpets

Hi Patrick:

I do not know what Alberto Levi was referring to. He wrote the following: "Another of the many patterns that have been extrapolated from early garden carpet iconography is derived from the shape of the watercourses, which are indicated by a motif that resembles a tuning fork. This recurs on the three vertical axes that define the orientation of the composition of a proto-Kurdish rug from the Sauj Bulaq area." Hali 70, page 92.

The best that I can come up with is that he was responding to the watercourses in certain formal garden carpets in which the weavers attempted to weave a pattern into the watercourses that can be understood as the water moving through the watercourses. See plate 9, page 90 of his article. These vaguely resemble the shape of the tuning fork motif. But it is also possible that he was referring to branches, only he could answer this.

One problem that Mr. Levi did not address but that I think must be confronted when trying to link Kurdish carpets to garden carpets is to understand or decide what gardens are being depicted. An allegorical garden, a garden in Kashan, one in Isphahan, or one in Kirman? Another is who wove the formal Kurdistan garden carpets and why? Here Mr. Levi seems to suggest that ex-Safavid court weavers, possibly Kurds, went back to Kurdistan in the late 17th or early 18th century and commenced what became an ongoing local village production of rugs with garden and/or floral motifs in the Safavid tradition, but much simplified. If so, it seems similar dispersions also occurred among other peoples and areas - even among the Baluch.

These issues are emeplified by the links you provided. Quite a bit of variety within a basic structure. In your last link the Persian Garden Carpet is the well known Wagner Garden Carpet, Burrell Collection, The Glasgow Art Gallery and Museum. Note the four-fold garden, or four gardens in one, orientationwhich is the classical garden layout and something not seen in the tuning fork rugs. This is a Kirman piece dated to about 1750.

If anyone is interested in garden carpets you might consider Hali Vol. 5, No. 1.

Regards, michael wendorf

Posted by Guido Imbimbo on 12-02-2002 12:32 PM:

Searching for Tuning-forks origin

Dear All,

Richard Farber has asked at the beginning of this thread to see some pictures of early Persian rugs that show the garden iconography to which represent the root of the sub-group of Tuning-fork carpets.

I am convinced that this linkage actually is discernable, though it is not an obvious one and its is very difficult to identify. I just hope to give a small contribution in this direction.

As starting point, I prefer to focus to the early Persian “shrubs” rugs. This is a vast term that includes many examples that differ in type, quality, age and origins. Nevertheless, consent me to distinguish two groups of carpets.

1) Compartments Shrubs Rugs
In these rugs, the shrubs, usually of small dimension, are organized in compartment or lattice structure. Floral tendrils often represent these connecting structures. The repeated small shrubs/bushes in each compartment vary in some floral e color detail. Here I enclosed the typical example is the Berlin Garden carpet (plate 107 of the Sphuler book). This carpet takes origin by the Garden carpets mentioned by Michael Wendorf and illustrated in Hali vol. V, No. 1 and that recall the structure of the elegant gardens that were popular in the Persian courts. In the Berlin rug simplification the watercourses have been omitted. It also important to remember that there is also a group of early shrub rugs organized in the form of lattice, some of them attributed to India (see Eskenazi book, plate 37).

2) Free-Field Shrubs Rugs
The size of the shrubs represented in these carpets is much larger then the previous group. Sometimes the dimension of the floral device can reach the size of a tree. The details of the design can be very articulated. In the best pieces, it seems there is an effort not to repeat the same forms but instead to maintain a high degree of design differentiation. Note that in general in these schemes the design does not include tendrils or other floral device that connect each shrub. Here I enclosed a picture of a beautiful fragment (104x265cm) taken from Hali Issue 39, page 98. There are also a number of others significant examples that follow this design scheme, some of them attributed to the Kirman area (see 9th ICOC book, plate 55).

The organization of the floral devices in the "Tuning-forks" rugs seems taking origin to both design structures illustrated above. While the flower elements are not organized in compartment or lattice, it is possible to identify in the bleu field, vertical grids that resemble floral tendrils with small knots that looks flowers (see the direct scan of my fragment). These tendrils connect the different floral elements represented in the field of the rug.

At this point it may also useful to recall also another group of early Kurdish carpets, which age probably is not very distant from the Tuning-fork group. In the field of these carpets we find an all-over repeating shrub pattern. Here I enclosed a picture of a Kurdish carpet (168x274cm) taken from the John Thompson book (page 46) that is also commented in the Levi article (page 89). I believe there is an interesting connection between the pattern of the shrubs Rogers rug and the "Tuning-fork group" and between this and the repeated "Trefoil" rugs in the Wendorf exhibition.

Finally, I enclosed a picture from a beautiful Sauj-Bulag fragment published in the exhibition Sovereign Carpets in occasion of the IX edition of the ICOC in Milan. In the beautiful harshang rendition of this carpet it is possible to detect the presence of small bushes/shrubs that recall the "trefoil" and the "tuning-fork" devices.

Best regards


Posted by Bob Kent on 12-03-2002 09:38 AM:

'Proto-Kurd' = Ur- or Uber-Kurd?

Guys: Perhaps the 'Proto-Kurd' label was meant more as 'Uber-Kurd' (i.e., height of) versus the 'Ur-Kurd' (i.e., firstever) that it seems to imply? As used, 'ProtoKurd' subsumes some very different things. For example, the fine tuning-fork rugs might be put in there, and Sotheby's waxed ProtoKurdishly about another rug with the same major border and related shrub field as my cut up bijar (reconstruction image below, it's minor borders are maybe like the one in the open-field shrub rug above?). I assume the construction of these 'ProtoKurdish' things is completely different?? .... see you, B Kent

Posted by M. Wendorf on 12-03-2002 12:16 PM:

The Ur and the Ueber

Hi Bob:

The exact term used by Levi was "proto-Kurdish." I spent years trying to figure out exactly what he meant by this term. I even asked him a few times. Each time the answer was different and increasingly vague. Perhaps I used the wrong methodology?

In his article, see page 87, Levi wrote: "Among the several classical traditions of carpet design that appear to have migrated from central and southern Persia to Kurdistan and then become characteristic of certain classes of Kurdish rugs, the so-called vase carpet group is of particular significance (reference to illustration 1, the Gulbenkian vase carpet.). This extensive body of carpets, which encompasses a great variety of designs, exerted a major influence on the development of a family of relatively early Kurdish rugs, which I will call proto-Kurdish." Levi also references Martin's description of dislocation and unheavels in the era of Nader Shah. Page 85.

This leads me to conclude that Levi believes that Kurdish weaving arose out of this dislocation and upheavel, in effect arising in the 18th century. Moreover, that a variety of classical traditions characterize Kurdish weaving with the Vase technique pieces exerting a direct influence on the these so-called proto-Kurdish pieces. In this sense, it seems he is arguing both that this dislocation and dissemination, if it occurred, is the Ur and the pieces within this group with the most lustrous wool and most saturated color the Ueber.

I think it suffices to state that I do not share Mr. Levi's views of Kurdish weavings although there seems to be little doubt that there is a substantial group, or family to use Levi's terminology, of Kurdish carpets with very lustrous wool and deeply saturated colors that have designs that seem related to Safavid era carpets. I would note that none are structurally related to Vase carpets. I would also note that while of great beauty, they make up only a small fraction of Kurdish weaving output over the past 200 years.

Regards, Michael

Posted by Richard Farber on 12-04-2002 12:21 AM:

Dear Mr Imbimbo and Mr. Deschuyteneer,

there is a jarring dissonance between your well researched and presented essay and the continued use of the misnomer 'tuning fork'.

I was presented with a similar problem in finding a catagory name for the niche form textile that I began to describe in a previous salon. I did not call them 'prayer' embroideries because I honestly believed that that was not the best catagory name for the knowledge [or at least the speculations] available at that moment.

I believe that you should find another name for the group and put the tuning fork in brackets after the new name. Hopefully the new name will catch on and be some indication of the contents of the group.

with most sincere regards

Richard Farber

N. B. thank you for the images. They are a great help in understanding.

Posted by Patrick Weiler on 12-05-2002 05:56 PM:

Motive Located

Where did the motive go?

Check out this thread on Salon 35:

There are references to Afshar and Khamseh weavings sharing this design, which may provide a road map of where this design went after leaving the purported Kirman Vase rug region of southeast Iran. It went west, then turned north, following the mountains, only to be located in hidden Kurdistan by Guido and Daniel, intrepid explorers!

The tuning fork itself certainly seems to be the stem of a floral device.

Patrick Weiler

Posted by Alberto Levi on 12-08-2002 12:54 PM:

Neither Ur nor Uber

Dear All,
I have been informed about this Salon by a friend that visits this site quite frequently. I read the work by Guido Imbimbo and Daniel Deschuyteneer, which I found very interesting, and read all of the comments made subsequently. Since I have been quoted (and misquoted) a number of times, I would like to take this opportunity to clarify a few points.
First of all, the misquotes. Michael Wendorf's statement that he asked me a few times for an explanation of the term 'protokurdish' is false. The only time we discussed about this was at the recent ACOR at Indianapolis. There Mr. Wendorf gave a talk about Kurdish rugs, where the common thread was that 20th century Kurdish rugs are the heirs of a five thousand year weaving tradition. His criticism of the term 'protokurdish' was that it is a misleading term, since clearly Kurds were weaving rugs certainly before the beginning of the nineteenth century. My reply was that I never disputed this: as a matter of fact, I tried to trace some very ancient origins in some Kurdish typologies and addressed this both at a convention on Kurdish rugs held at the Textile Museum in Washington and on a paper published thereafter on 'Ghereh'. I can give further details about this to anyone interested. The term 'protokurdish' attempts at defining a cluster of weavings with common structural and chromatic features, which also share a common lineage in that they seem to re-interpret a number of patterns that are characteristic of certain Safavid typologies. Furthermore, these 'protokurdish' rugs seem to be immediate ancestors to many nineteenth and twentieth century Kurdish rugs from west and northwest Persia.
They represent a kind of bridge between the earlier court weaving tradition and the post-Safavid establishment of a village weaving tradition.
Which brings us to the subject of 'tuning forks', or the quotes. Indeed I see a strong resemblance between the 'tuning fork' motif as it appears on the Sauj-bulagh rugs illustrated in this work and what I call 'the shape of the watercourses' on Safavid Garden carpets (such as the Aberconway Garden carpet in the Kuwait National Museum, for example). The directional floral motif is most probably a result of an extreme stylization of the design of some earlier 'shrub' rugs (some of which are illustrated in Guido Imbimbo's post), and in that sense it is probably a hybrid design. Nevertheless, the directional shrub design is represented on a different Sauj-bulagh typology (the Rogers Thompson rug, illustrated both on my 'Hali' article and on Imbimbo's post), and therefore should be seen as independent from the 'tuning fork' type.
Common to all the rugs belonging to the 'tuning fork' sub-group is a repeating unit composed of a central seven-petaled flower flanked by a pair of four-petaled flowers in contrasting colour. From each of these four-petaled flowers sprout a three-petaled flower. This blossoming unit 'rests', if you will, on the cup of the 'tuning fork' motif. Could it be possible that the designer of this type wanted to synthetize both the representation of the flow of water on Garden carpets and that of the so-called 'islands' - from which blossom flowering trees (see for example the 'islands' on the Davis Garden carpet fragment in Metropolitan Museum of Art, N.Y., published in Dimand & Mailey, fig. 117, p. 85)?

Best Regards,
Alberto Levi

Posted by M. Wendorf on 12-09-2002 06:33 AM:


Greetings to all:

Welcome to Alberto Levi on this Board. As some of you may know, Mr. Levi has made numerous contributions to the rug world including his service as chair of the local committee for the ICOC conference last in Italy. In addition, Alberto has long been a promoter of Kurdish weavings and has documented several rare types of Kurdish weavings over the last years.

This Salon references Alberto's article in Hali 70, Renewal and Innovation - Iconographic Influences on Kurdish Carpet Design. In his article, Mr. Levi coined the term "proto-Kurdish" and attached it to four rugs illustrated (among numerous others) within the article. All four of these rugs were attributed to "SaujBulaugh". Two of these carpets were identified as being part of Mr. Levi's collection. For precision and consistency, I continue to use the original termand spelling of "proto-Kurdish." However, I think it is preferable in English to use Sauj Bulaq rather than SaujBulaugh and do so here. In fact, both spellings refer to a town that today is Mahabad. Sauj Bulaq was an important Kurdish town and Mahabad remains so today.

Mr. Levi refences some misquotes. No one here wants to misquote him or anyone else. In one instance, Alberto states that my statement that I asked him "a few times" for an explanation of the term proto-Kurdish is false. This is a strong word, so I will respond. I attended the Textile Museum conference devoted to Kurdish rugs referenced by Mr. Levi. In fact, several of what Alberto calls my 20th century rugs were used to illustrate the lectures and discussion. I recall a discussion of proto-Kurdish even then. I also recall exchanging emails with Alberto in 1998 or 1999 in which the subject was raised. In each case, I recall discussions about the use of the term proto-Kurdish. But perhaps I am mistaken. In any event, we can discuss it now.

Proto is derived from the Greek protos. Protos is generally understood by me to suggest the earliest or first in time, the first formed. What Bob Kent referred to as the Ur. In language and science it can have more specific meaning, but it is consistently used in the context of "first formed". In his post, Alberto writes that proto-Kurdish as used by him "attempts at defining a cluster of weavings with common structural and chromatic features, which also share a common lineage in that they seem to re-interpret a number of patterns that are characteristic of certain Safavid typologies." (By typology I understand Mr. Levi to mean some systematic classification such as the so-called vase structure in which all the members of the type or class share an assymmetric knot and a depressed back). "Futhermore, these proto-Kurdish rugs seem to be immediate ancestors to many 19th and 20th century Kurdish rugs from west and northwest Persia."

I see several issues arising from this explanation and the original article. First, if this cluster of weavings share common structural and chromatic features what are they? Mr. Levi provided no structural or chromatic analysis of any of the pieces he illustrated or discussed in Hali 70. So far as I recall, the only structural discussion was found on page 87 wherein we are told:

"Most of these early proto-Kurdish carpets are characterised by a lustrous, symmetrically knotted woollen pile. In many cases they have rust colored wefts."

Based on this conclusion, he refences Eagleton and concludes this group comes from Sauj Bulaq. Id. I do not believe words like "most" or "in many cases" and the scant structural information provided is suffient to talk about a typology much less a label as grand as proto-Kurdish.

The same is true for the common chromatic features. What exactly are they? Of the four pieces illustrated in Hali 70, two are woven on an ivory ground and two others on dark grounds either a dark corrosive brown or something else. All I recall being written on this issue is the observation that:

"The coloured yarns employed for the pile of these latter rugs is extremely saturated, and are an indication of the well-known competence of the Kurds in the art of dyeing." Page 87.

In my view, this hardly amounts to common chromatic features. Moreover, as Mr. Levi points out in his footnote, dyeing was largely carried out by Jews during the relevant period. So what exactly does the observation of saturated dyes have to do with the argument that these rugs, even if we assume they form a single group based on other common features have to do with them being Kurdish or proto-Kurdish?

And that, it seems to me, is the real issue. What makes these rugs protoanything? Alberto's observation is that the rugs he calls proto-Kurdish (and which I would simply call rugs with floral motifs interpreting Safavid patterns or designs) must be the ancestors to a larger group of carpets that also interpret Safavid era patterns or designs. This is apparently because he believes the four rugs he illustrates are older than most. Well, I think the most one can say is what Alberto adds in his post - "they form a kind of bridge between earlier court weaving and the post-Safavid establishment of a village weaving tradition." But even this (an Ur limited to "village weavings") is a can of worms. We cannot really say whether or not village weaving or Kurdish village weaving existed during, before or arose only after the end of the Safavid era. Mr. Levi's article offers no insight or new information on this question other than to surmise that weavers may have dispersed into Kurdistan after the fall of the Safavids.
In addition, the article and use of the term proto-Kurdish completely ignores a non-village or tribal weaving tradition among Kurdish groups that I understand Alberto himself believes may well have existed.

I have nothing to add to the discussion/use of tuning fork to describe the motif. I still find it to be a floral stem. But Alberto's view is clarified.

I am not familiar with Alberto's article in Ghereh that attempts to trace some very ancient origins of certain Kurdish weaving types and do not recall this specifically from the TM lecture either. I hope Alberto will expand on or at least summarize his ideas on this. Alberto refences a talk I gave at ACOR which was also the subject of a previous Salon - I believe it was #88. I will allow that to speak for itself except to point out that I refer to 6000 years of weaving history, not 5000.

Thanks in advance to Alberto and the prospect of more discussion and clarification of our points of view.

Regards, Michael

Posted by Alberto Levi on 12-09-2002 08:59 AM:


Dear All,

during the last few years the term 'proto-Kurdish' has been used in many instances, more often on auction catalogues. My original intention was that of isolating a structurally and chromatically consistent group of weavings, and placing it in a time line with respect to what existed before and especially to what developed later. From a structural point of view, all of the proto-Kurdish rugs of the Sauj-Bulagh type that I examined in person have an all-wool foundation,with natural wool warps and almost always two shoots of very fine rust colored wefts, a fine knotting with very little to almost no warp depression, and a soft pliable handle. I never recorded any precise data on these, but all of these features gave me the clear impression that we are dealing with a structurally consistent group. The palette of all of these rugs also seems to be very consistent, with very similar shades of yellow, red, light and dark blue and aubergine often on a dark brown background, more rarely on a yellow background and very rarely on an ivory background. In every case the colours are clear and vibrating, with a translucent quality that gives almost the impression of looking at a glass mosaic.
However this is limited to the Sauj-Bulagh group. Since publishing my article, many other types of proto-Kurdish rugs came to light, some with completely different structures and colours. But this could be the subject of a new Salon.
In any case, my 'proto-Kurdish theory' does not imply that there wasn't any village and tribal weaving carried out by the Kurdish people of west and northwest Persia prior to 1800. The problem is how to identify them. I have attempted at connecting the hooked lozenge design of the Jaf tribe, as well as the use of offset knotting, to some early Turkish rugs possibly from eastern Anatolia. This was part of my contribution at the Textile Museum Kurdish rug convention (Michael: was year was it?) and also part of an article on Kurdish rugs that I wrote for 'Ghereh'. I have since 'forgotten' about everything Kurdish, but perhaps this Salon will motivate me enough to start working again on this fascinating subject.

Best Regards
Alberto Levi

Posted by Steve Price on 12-09-2002 09:45 AM:

Hi Alberto,

If you would like to prepare a Salon essay on the proto-Kurdish designation, I will be glad to run it. As you may know, we open a new Salon on the 24th of every month, and barring anything unforeseen I can put this into the first slot after I receive it.

Would you let me know if you are willing to do this? If so, I will ask people to leave the subject of proto-Kurdish for the moment rather than embark on a more detailed discussion of it within the context of Guido and Daniel's Salon.

Thanks, and welcome to our forums.


Steve Price

Posted by M. Wendorf on 12-09-2002 02:07 PM:

Kaleidoscope of Kurdish rugs

Greetings All:

Alberto, the TM conference called Kaleidescope of Kurdish Rugs must have been '95?

Steve, I would prefer to leave the subject of proto-Kurdish rugs forever. More realistically, I think it would be of broader interest to ask Ghereh if Turkotek can use Alberto's article there as the basis for a Salon. We have previously had some discussions of the Jaf hooked lozenge and offset knotting; the connection with old Turkish rugs - are we talking rugs from Divirgi? - could be great.

Regarding isolation of rugs with common structural and chromatic features, I am all for it. However, isolating a group of Kurdish rugs on the basis of all wool foundation, natural wool warps, and two shoots of rust colored weft with a flat back and pliable handle includes many different kinds of Kurdish rugs in my experience. We have discussed here at length about what we might call a traditional Kurdish weave - two ply natural or ivory warps, flat back, wool or wool with animal hair for foundation, pliable or floppy handle with glossy wool, mulitiple wefts of various colors but often reds, corals, rusts and browns etc. - not much isolated from the general description supplied by Alberto. And I have seen rugs seemingly in the same group Alberto discusses with coral and red wefts.

Likewise isolation based on general palette consistencies seems dangerous. In my view, a more detailed dye analysis is necessary before we can even start to isolate a group within the production of these rugs with floral designs interpreting Safavid models.

Finally, the timeline that was Alberto's goal beyond isolation of a group is always speculative. Alberto thinks his rug is 1800 and mine 1920, I think the opposite. Who is right? Probably neither. This is not to say that someone like Alberto who has handled a lot of these rugs cannot get a feel for such a timeline, it is just that we can never really know because there are no anchors or base from to start with any real confidence. Great dyes continued to be made after 1800 and inferior ones prior to 1920.

Personally, I doubt there are any real close structural or chromatic consistencies sufficient to start narrowly isolating groups of these rugs and establish a typology. In any event, I think side and end finishes would have to be considered as well. In the end, it seems likely that several small shops were responsible for these rugs, however organised, and made them over some period of time, perhaps even 100 years or more. And we likewise need to consider that similar enterprises seem to have arisen in a variety of areas across Persia, not just Kurdish areas.

I for one do hope that Alberto will host a Salon.

Regards, michael

Posted by Steve Price on 12-10-2002 09:10 AM:

Hi All,

Alberto sent me a message saying that he will try to prepare a Salon on proto-Kurdish for our January 24 slot. This being the case, I ask that we suspend discussion of that topic for now.


Steve Price

Posted by M. Wendorf on 12-12-2002 03:34 PM:



I went and found my notes from the TM conference. Believe it or not, the dates was October 15-17, 1993. The topic of your lecture was "Classical Kurdish Rugs from Persia and Anatolia." My notes, such as they are, make no direct reference to the Jaf lozenge and the only reference to offset knotting is to the effect that if there are early Kurdish rugs, offset knotting would be one way to identify them. I believe you showed some slides of Anatolain rugs with offset knotting that might be Kurdish including one fragment with big hooked devices in the field. Perhaps this piece was your connector to the Jaf lozenge?

My notes also state that in 1993 you and Christina Bellini were working on a book on Kurdish rugs, what happened to that project?

Best, michael