Home Page Discussion Boards

Salon du Tapis d'Orient

The Salon du Tapis d'Orient is a moderated discussion group in the manner of the 19th century salon devoted to oriental rugs and textiles and all aspects of their appreciation. Please include your full name and e-mail address in your posting.

Kurdish Rugs and Related Weavings - an 8000 Year Weaving Tradition?

by Michael Wendorf


Kurdish rugs have long been misunderstood and/or ignored. This is exemplified by an old dealers joke that goes: "How do you know if a rug is Kurdish or Caucasian?" The answer? "Well, if you are selling it, the rug is Caucasian. If you are buying it, the rug is Kurdish." Though this is slowly changing, there is still little appreciation or understanding of traditional Kurdish rugs from historic Kurdistan - or what William Eagleton has called the Kurdish heartland.

Part of the problem has been political. With the possible exception of the brief Kurdish Republic of Mahabad (modern Sauj Bulaq) in 1946, it is centuries since Kurds have had either a Kurdish government or country. In modern Turkey use of the words Kurd, Kurdish and Kurdistan is illegal. Use of the Kurdish language and dialects have been discouraged.

Another significant problem has been the tendency to call Bijars, Sennahs and Kolyai Hamadans, each with their distinctive weave, a Kurdish rug. This tendency seems to have originated with Cecil Edwards in 1953 and remains in use by some today. While some or all of each of these rug types may have been woven by Kurds, they are each a response to different commercial influences and shed little or no light on any Kurdish weaving tradition. Moreover, none of these rugs are produced in the Kurdish heartland. Likewise, the weave structures are not what I would consider to be traditional Kurdish weaves. Rather, they are each place-specific structures arising in a specific commercial context.

A third problem is what Jim Klingner has described as the "Kurdish Synthesis," a tendency among Kurdish weavers to adopt and adapt influential patterns into their own style. In adapting such designs, more traditional rugs can become lost. In addition, the adopted and adapted patterns may become so popular and well known that they themselves begin to be what we visualize when we think of a Kurdish rug.


To understand traditional Kurdish rugs and related weavings, we need to understand something about Kurdistan and the Kurdish people who live there. Historic Kurdistan encompasses parts of western and northwestern Iran, eastern Anatolia, northern Iraq and the southern Caucasus as well as a bit of Syria. A convenient map of Kurdistan and the weaving heartland is found in Eagleton's Introduction to Kurdish Rugs at page 6. Elevations range from 3000 feet above sea level to peaks over 8000 feet.

Kurds inhabiting Kurdistan are first and foremost a people of the mountains.

These people are Iranian in origin and trace their roots as a people back at least as far as the Medes - who ruled parts of modern Iraq and Iran in the 9th - 6th centuries B.C. Eagleton concludes that the Kurds "...entered history as a mountain people occupying a part of northwest Iran from which they gradually moved south, north and west in Asia Minor and Iraq" (Introduction to Kurdish Rugs, p. 9). Eagleton does not state when this occurred. I tend to believe the Medes were but one group who mixed with the indigenous population of Kurdistan, a population that existed long before the Medes.

We may never know the precise history of the Kurdish people. However, I believe their weaving tradition supports the conclusion that Kurds and their ancestors have inhabited the mountains of Kurdistan since antiquity. I also believe that the tradition of goat and sheep herding in Kurdistan is consistent with this.


Kurdish weavers have been called "imaginative and prolific" by Murray Eiland. Other commentators have noted the Kurdish love of deeply saturated color and soft glossy wool. None of these descriptions really say anything about Kurdish rugs or a weaving tradition. To get a sense of the Kurdish weaving tradition that I have come to appreciate, one has to take a journey back in time.

It comes as little surprise to learn that many of the world's earliest surviving textiles (textiles dating to 6000 B.C.) have been found in sites within the so-called Fertile Crescent - including sites within historic Kurdistan which may have been home to the ancestors of Kurds or known to these ancestors since antiquity. This area, home of man's initial plant and animal domestication, has also played a significant role in the history of weaving. Kurdish weavings document, in my opinion, this history - a history that spans 8000 years in Anatolia. One example of this tradition is a group of utilitarian weavings which Marla Mallett calls "weftless soumak" and which John Wertime has called "simple soumak." Marla Mallett's term is more descriptive and is the term I use to describe the structure of these weavings.

As Marla Mallett has noted, in both structure and technique, weftless soumak is unique.

Why unique? Unlike other forms of soumak wrapping, no ground wefts are interlaced between rows of wrapping. As a result of this absence of ground wefts, weftless soumaks have slits much like slit tapestry. Patterns are typically created by wrapping small, distinct sections with yarns that are discontinuous.

In my mind, weftless soumak represents the earliest form of weft wrapping. Marla Mallett in her book, Woven Structures and John Wertime in his article "Origins of Pile Weaving" in Hali 100 have expressed the same opinion.

The existence of weftless soumak is significant to placing the Kurdish weaving tradition in its proper perspective. This significance is twofold. First, it seems to be a Kurdish technique. I am unaware of other weaving groups using it. Second, it can be linked to 8000 year old carbonized fragments. In 1961, carbonized burial fragments were excavated in Level VI of Catal Huyuk and published by Harold B. Burnham. These burial fragments, now in the Anatolian Archaelogy Museum (in Ankara), include fragments woven in weftless soumak. Although probably woven with bast fiber rather than wool, these fragments establish weftless soumak as a technique that predates the invention of loom shedding devices. As such, it is not dependent on the shed or the heddle, but is a simple technique related to basketry.

It cannot be said that the weavers who wove these 8000 year old carbonized burial fragments were Kurdish. We cannot even be sure what kind of loom was used, although it is possible that it was a warp weighted loom. What can be said is that Kurdish weavers seem to be the only weavers who wove in the 19th century using this ancient technique. It is possible that this is a coincidence. However, I think the better inference is that the weftless soumak pieces displayed represent the final product of a long and deeply ingrained tradition. This is a weaving tradition Kurdish weavers maintained long after other weavers had abandoned it, if they ever used it, and long after the need to weave in this technique had vanished. Why did Kurdish weavers continue to weave these pieces into the 19th century? Because that is how they had always been woven.

Though generally undocumented in rug literature, weftless soumaks, as represented by the examples shown here, come in a variety of bag and chuval formats. These weavings seems to have been woven throughout much of historic Kurdistan and for a variety of utilitarian purposes. These bags are almost without exception carefully crafted using excellent, glossy wool. Many of these weavings in chuval and sack format, as opposed to saddlebag faces, are only partially woven in weftless soumak. As seen above, the alternating blue and red stripes are woven in a weft faced plain weave. I do not know what significance, if any, this observation has.

The tradition of weftless soumak may have had significant influence on the design development of Kurdish bags generally. The restrictive nature of this structure with the wrapping of small sections with directional yarns and the absence of ground wefts resulted in diagonals and strong reciprocal elements. However, as Marla Mallett has pointed out to me, it is the reversal of wrapping threads that has influenced design development in Kurdish bags. That is the carrying back of each wrapping thread so that wrapping can always proceed in the same direction. One result is the tendency for pattern parts to be narrow. We see this tendency continued in a variety of knotted pile weavings where the structure, knotted pile, does not dictate such a result or tendency.

If we admit the possiblity that Kurdish weaving may represent an ancient tradition that is intertwined with the very history and development of weaving as we know it today, then we must also conclude that this tradition arose most likely out of simple flatwoven weavings and that these flatwoven weavings had a profound influence on all aspects of Kurdish weaving. Other typical flatweaves include reciprocal brocading,

overlay - underlay brocades and warp substitution weavings.

Kurdish weavers also weave a variety of so-called primitive rugs.

Here we see an east Anatolian example, possibly woven as a sleeping rug, using unspun wool, symmetrically knotted wool pile on a wool, weft-faced plain weave ground. The reverse shows some weft float brocading, typically 3/3.

Another distinctive weaving is the so-called Siirt rugs. Woven with a weft faced plain weave with goat hair wefts and cotton warps. This rug has no knotted pile. The effect of pile is created by teasing the face of the weaving. Examples such as this one suggest that this is a tradition of long standing not just because of the simplicity and primitive nature of the weaving, but also by the subtlety of it. Note that the weaver created a lattice pattern in the rug using only a teasel.

Next we find a Goyan or Hartushi rug with 5 1/2 vertically aligned concentric hooked diamonds. The weaver here used symmetric knots tied on four 2 ply tan wool warps with about 18 knots per square inch. Unlike the large sleeping rug with unspun yarns depicted above, the yarns comprising the knotted pile here are Z spun.

This Goyan or Hartushi rug seems to be imitative of a slit tapestry design underscoring the influence of flatweaves on Kurdish knotted pile weaving. But it also seems to link weavings like the sleeping rug into a larger continuum that may one day help us better understand how and where knotted pile developed.

Moreover, the rugs and related weavings depicted above are illustrative of a weaving tradition that seems to include techniques and structures that, when considered together, seem to suggest that Kurdish weavers and their ancestors were not only weaving in antiquity but were squarely involved with the innovations and developments that brought weaving forward into today. In this sense,the "Kurdish synthesis" that Klingner mentioned in the Chicago ACOR exhibition catalog needs to be reevaluated. Rather than being known simply as weavers who boldly adopted and adapted influential patterns into their own style, Kurdish weavers need to be recognized as having possibly contributed and most certainly preserved the most basic and traditional techniques we know. In short, the place of Kurdish women in the history of weavers needs a major reassessment.

Although I began collecting Kurdish rugs by collecting rugs with so-called Persianate designs, I believe this tradition emerges only out of the 16 - 17 th centuries. Although this is a significant span of time in the mindset of most rug collectors who think of a rug being really old if it could reach the 18th century, it is largely insignificant if considered in the context of an 8000 year weaving tradition. Some of these Persianate weavings have been labelled "Proto-Kurdish" after Levi's article in Hali 70. I believe that the label "Proto-Kurdish" used in connection with these Persianate weavings mischaracterizes Kurdish weaving and the tradition of Kurdish weaving.

The long rug depicted above appears at first glance to be what some might call Proto-Kurdish based on its use of a Persianate shrub design in a lattice. However, even here (or perhaps especially here) what speaks to me is a tradition much older than the Safavid Dynasty in Persia. The lattice itself outlines a series of stepped polygons arising out of a flatweave tradition. And that border, that I see as coming out of a warp substitution pattern. Both of these traditions were probably thousands of years old by the 19th century.

Proceed to Part 2

Discussion Boards