Afghan Carpets

By: Brian I. Spooner

Originally Published in 1973

View PDF

This article is being published to coincide with an exhibition of the carpets of Afghanistan which is to be held in the University Musuem from March 7th to April 15th 1973. The purpose of the exhibit is to display a range of carpets which have been selected explicitly for the purpose of demonstrating differences in materials, workmanship and overall quality.
This article is being published to coincide with an exhibition of the carpets of Afghanistan which is to be held in the University Musuem from March 7th to April 15th 1973. The purpose of the exhibit is to display a range of carpets which have been selected explicitly for the purpose of demonstrating differences in materials, workmanship and overall quality.

Weavers and Dealers

In the history of international trade. Oriental carpets are something of an anomaly. Although other exotic crafts have found a market in the West, no other has been so successful for so long, and marked by such lack of communication between producer and consumer. In the most extreme situation a rug is woven by women in a nomadic encampment in central Asia. It is meant for domestic consumption as a primary item of furniture. However, when times are hard, as in the recent famine, the nomad takes it to a local bazaar center where it is sold for cash. Through resale it finds its way to an emporium in a major city. Along with many others it is then exported by an Oriental merchant to a dealer in the West. The carpet may then be subjected to certain treat­ments which would affect its colors or sheen, and so enhance its appeal to the Western consumer. The cultural criteria according to which the carpet was woven were entirely independent of those that influence its ultimate purchaser.

In less extreme cases—and for a large pro­portion of the best carpets—the weaving is financed by a merchant in the outskirts of a city, or in outlying villages. Here the work is done by men as well as women and children. Today such a merchant might have fairly direct dealings with major Western importers and dealers, with the result that the Western dealer can exert some influence on the colors, designs and sizes of the Oriental production.

However, the dealer is not the consumer. He is a businessman. He purchases. and insofar as he is able directs the production of the types of rug that he estimates will best enable him to maximize his sales. Furthermore, the dealer does not have the field to himself. There is another role that has to be considered in this anomalous system of communication: the role of the art historian.

Karakul sheep awaiting sale in a market center. Karakul lambskins are well known in the fur trade. The adult animals are one of the main wool-producing species. The length and strength of the wool fibres differ from one species of sheep to another, and some wools dye better than others.
Karakul sheep awaiting sale in a market center. Karakul lambskins are well known in the fur trade. The adult animals are one of the main wool-producing species. The length and strength of the wool fibres differ from one species of sheep to another, and some wools dye better than others.
Black tents. This is a camp of Pushtun (Afghan) pastoral nomads in north-western Afghanistan. Pushtuns do not traditionally weave pile carpets, but they produce much of the wool used in carpet weaving by the Turkmen, who are not nomadic within Afghanistan.
Black tents. This is a camp of Pushtun (Afghan) pastoral nomads in north-western Afghanistan. Pushtuns do not traditionally weave pile carpets, but they produce much of the wool used in carpet weaving by the Turkmen, who are not nomadic within Afghanistan.

The art historian complicates the economics of the system by introducing the alien and sup­posedly abstract and absolute criteria of academe. If we may for the moment ignore recent developments in the field and characterize him briefly, the art historian is primarily interested in tracing the history of art forms, in this case of symbols and designs, and secondarily of the material culture and technologies associated with them. In the case of movable art forms. such as carpets, the dealer tends through his business dealings to be in closer contact with the place of origin of the objects than the art historian who pursues the academic study of them—though the dealer is still at one remove. It is thus in the nature of things that there should be a close association between art historian and dealer, as there once was between the archaeologist and the antique trade. The affluent public that patron­izes the dealer, also patronizes the academic. since it is he who raises the value of the objects beyond the purely economic plane, The role of the art historian therefore has tended to reinforce the lack of communication between the weaver and the consumer.

A wool dyeing shop. Few carpet weavers dye their own wool. Shops such as this all use imported aniline dyes in vats in the shop and is being hung out to dry.
A wool dyeing shop. Few carpet weavers dye their own wool. Shops such as this all use imported aniline dyes in vats in the shop and is being hung out to dry.

The Western consumer buys Oriental carpets as floor covering and as investments. In making his choice generally he has little to go on besides personal taste and what the dealer tells him, modified by some degree of familiarity with the literature which is created by both art historians and dealers. This literature suffers generally from excessive definition and overclassification, which is bewildering to the layman, and with few excep­tions contributes little to any practical under­standing of the craft. Although there are signifi­cant differences in design. color, knot and quality that are related to place of production, these are by no means all so closely related to the identity of the weavers that they may be used as reliable criteria of provenance. Some centers of produc­tion were more influenced by foreign demands than others and adapted their production accordingly.

Most libraries contain a shelf of books which, besides offering such a classification of Oriental carpets, cursorily introduce the reader to the history of the craft and the trade. It is interesting to note that in one such book published in Phil­adelphia in 1913 (The Practical Book of Oriental Rugs by Dr. G. Griffin Lewis, pp. 19-20) it was stated that although the trade in Oriental rugs was at that time well advanced in Europe “twenty-five years ago few American homes possessed even one. Since then a marked change in public taste has taken place. All classes have become inter­ested and. according to their resources, have purchased them in a manner characteristic of the American people, so that some of the choicest gems in existence have found a home in the United States…. This importation will continue on even a larger scale until the Orient is robbed of all its fabrics and the Persian rug will have become a thing of the past.” The last part of the prophecy has not of course come true. Production has been increased to meet the demand. How­ever, Lewis continued: “Already the western demand has been so great that the dyes, materials and quality of workmanship have greatly deteri­orated.” Sixty years later the enormously in­creased demand can be seen to have had signifi­cant effects on the product. Different types and qualities fit into the market at different levels from the rare antique and top quality modern piece for the collector and the investor, down to a price range to suit the pocket of the ordinary American family.

The Consumer

Dyeing with traditional vegetable materials. This takes more trouble, and seems to be practiced now only by very few high quality weavers.
Dyeing with traditional vegetable materials. This takes more trouble, and seems to be practiced now only by very few high quality weavers.
Girl spinning. The secret of the most expensive carpets lies in the processes of carding and spinning: the fibres are first sorted and then spun into wool threads of different strengths and thickness, according to whether they are to be used as warp weft of pile.
Girl spinning. The secret of the most expensive carpets lies in the processes of carding and spinning: the fibres are first sorted and then spun into wool threads of different strengths and thickness, according to whether they are to be used as warp weft of pile.
Turkmenwomen weaving on a home loom. They sit on what they have woven and move forward as they progress. After making a line of knots and inserting the weft threads, they are beating back the weave with comb beaters as the final stage in acheiving the desired density of knots. When the weave is beaten sufficiently tight, they will make another line of knots and insert more weft threads, and so on.
Turkmenwomen weaving on a home loom. They sit on what they have woven and move forward as they progress. After making a line of knots and inserting the weft threads, they are beating back the weave with comb beaters as the final stage in acheiving the desired density of knots. When the weave is beaten sufficiently tight, they will make another line of knots and insert more weft threads, and so on.

Most current purchasers of Oriental rugs do not expect to get an antique or a collector’s item. They are attracted by an exotic floor covering that promises, if not to appreciate, at least not to lose value as quickly as the comparable domestic product. They have little knowledge of the market or how prices are determined. A recent report on the foreign market for Afghanistan’s carpet industry prepared by the International Trade Centre in Geneva states that a typical carpet retails in Germany for three times its purchase price in Teheran. Similar statistics are not avail­able for the United States. These statistics were compiled from information supplied by reputable dealers. Lewis (op. cit.) discusses dealers who were not reputable, and it should not surprise the reader that such do still exist, and thrive.

The best known Oriental carpets are Persian, and in fact many of the best carpets were and are still made in what may be called the Iranian cultural area.

Although there is sufficient similarity in the techniques used throughout the Iranian cultural area and beyond to justify using the broad term “Oriental carpet” for the total production, there are also significant differences. It is the knowledge of these differences in technique and their implications, rather than places of origin by which the carpets tend to be known, that is most valu­able to the consumer. In the remainder of this article I briefly describe some of these differences with special attention to those which are features of the best Afghan carpets. I then conclude by indicating some of the changes which have taken place in the industry in Afghanistan during the last decade.

The Technology

The exquisite large carpets which began to be known in Europe in the seventeenth century were commissioned by the Safavid court of Persia, and appear to have had predecessors similarly commissioned by earlier dynasties. However. carpet weaving was never a court monopoly, and may well have originated in the nomad’s tent in Central Asia. Rugs which are now sold under the category “tribal”—which are by no means all woven by tribal people, let alone by nomads—are generally smaller and are produced on horizontal looms laid out on the ground. Most of the better carpets that are woven within the present borders of Persia and find their way into Western markets are woven on vertical looms. The vertical loom has the advantage that more and larger carpets may thus be woven within a smaller space. Therefore, although any type of carpet of any quality can be woven on either a horizontal or a vertical loom, traditionally carpet industries which were financed in cities (mostly within Persia) and staffed by employees, rather than produced by individual domestic operators, were woven on vertical looms, whereas most operators who wove for their own purposes used horizontal looms. All Afghan carpets formerly were, and the overwhelming majority still are, woven on horizontal looms. They are woven mainly by people of tribal origin. Turkmen and Baluch, most of whom until recently lived in tents or small villages close to the northern and western borders of the country. Recently some have moved into the outskirts of towns and cities, and members of other ethnic groups have also taken up carpet weaving.

Map-Teheran-Bukhara-KabulSuch small-scale operators used the mate­rials they had at hand. They used wool for the warp, the weft and the pile of the carpet. The raw wool from the sheep was washed, carded, graded and spun into different thicknesses and strengths according to whether it was to be used for warp, weft or pile. It was then dyed with vegetable dyes produced locally. Until very recently all Afghan carpets were made entirely of wool, and most still are. This requires a higher degree of skill than the use of cotton which is prevalent in carpets made on vertical looms (and also most of those made on horizontal looms outside Afghanistan), because, however strongly spun, a wool thread is not so strong as cotton and a wool warp thread is liable to snap at any time as it undergoes tension in the weaving process. When a warp thread does snap, it must be mended in place by spinning the snapped ends back together. Although the use of cotton thread for the warp obviates this problem, it changes the texture of the fabric without signifi­cantly increasing its durability. Since most of the warp is covered by the pile it is to a large extent a matter of taste whether cotton or wool warps should be preferred, but an appreciation of the craftsmanship involved and the feel of an all wool product can be persuasive. The same case cannot so easily be made against the use of a silk warp, but silk is more expensive and less readily available.

Warp threads are vertical; weft threads; horizontal. Left, shows the Ghoirdes knot, Right, the Sehna knot.
Warp threads are vertical; weft threads; horizontal. Left, shows the Ghoirdes knot, Right, the Sehna knot.

Across the warp threads, one, two, or three weft threads are woven between each row of the knots, which form the pile. The greater the number of weft threads, the stronger and more even will be the finished carpet, but this will obviously also make it more difficult to achieve the highest number of knots per square inch. There are two different knots in use (see figure). These may be made with single or double strands, and they may be made around single warp threads (as in the figure) or around double threads. Obviously if they are made around double threads the pile will be less dense and the carpet will be finished more quickly, and this is a notorious way of cheating. The difference between knotting with single or double threads, and between the two different knots, is not so simple. Generally, the finest, thinest and most pliable carpets are made with the Sehna knot (Persian: yak pith) knotted with a single strand (yak tar). Ghiordes knots (do pith) made with single strands (yak tar) are unusual. Both knots may be found made with double strands (do tar). In this case once again the Sehna knot produces a somewhat more pliable fabric, but the Ghiordes may be considered more durable in that when a carpet is damaged the Ghiordes will hold—since it is a real knot—whereas the Sehna will unravel. In a fine carpet the two may seem impossible to differentiate. However, if the fabric is bent back double, the Sehna will reveal every other warp thread uncovered, whereas with the Ghiordes every warp thread is covered with the knot.

boys weaving from a pattern. At the far side of the loom a man sets the heddle rod for them. After making each knot they cut off the excess length with a knife. This is the same "factory" situation seen in the preceding picture.
boys weaving from a pattern. At the far side of the loom a man sets the heddle rod for them. After making each knot they cut off the excess length with a knife. This is the same “factory” situation seen in the preceding picture.
Boy pulling back shed stick in order to shift the position of the even warp threads in relation to the odd threads snd fscilitate the insertion of the second weft thread on another "Afghan gold."
Boy pulling back shed stick in order to shift the position of the even warp threads in relation to the odd threads snd fscilitate the insertion of the second weft thread on another “Afghan gold.”

In the opinion of most fanciers, color com­position is a classic feature of Oriental carpets, but there is in fact great variation both in the colors used and the way they are combined. The most significant variation is between colors pro­duced by chemical dyes and those made from local traditional materials. Most carpets that reach the Western market today have been chemically washed and treated in such a way that their colors, for better or for worse, have been changed from their original quality. And some such treat­ment has been the practice for many decades. A less expensive and older treatment is to expose them to wear and sunlight artificially by leaving them out in the streets to be walked and driven over. This is still a common sight in cities such as Teheran and Kabul and has the effect of softening the colors. It is commonly believed that the colors of a good Oriental carpet are prepared from vegetable dyes. However, the process of vegetable dyeing requires much time and trouble, and ever since chemical dyes began to become available from the West in the middle decades of the last century they have quickly spread throughout the Oriental industry. For a long time now even the more isolated producers in small villages and nomad encampments have been using chemical dyes, and today there are very few carpets indeed which are made from wool dyed with natural dyes. Most if not all of these come from Afghanistan.

Since both natural and chemical dyes may be completely fast and both may produce ex­cellent colors, there may be no foolproof method of determining which has been used. Only a long familiarity with the potential of the different dye­ing materials is a reliable guide. And after all if the colors are good. it is not important which were used. However, in carpets which are known to be naturally dyed the colors tend to be noticeably softer. Further. it seems likely in many cases that natural dyes are kinder to the wool in the long term. In many older carpets which have received a fair amount of wear the different colors can be seen to have worn differentially, leaving a relief effect due to the different qualities of the dyes.

Finally, different designs in tribal carpets are traditionally the work of different tribal groups, and among the Turkmen the design is related to the identity of the tribal groups. like the Scottish tartan. There is an overall recognizable identity in the design of Turkmen carpets which is generally referred to as Bokhara—the central Asian city where many of the best examples of Turkmen carpets were once produced. With a little familiarity the varieties of this general design, particularly those produced by the Saruq Salor, Tekke, Yamut, and Ersari tribal groups, can be easily recognized. However, since the movements of these tribal groups and the change s in traditional Turkmen society occasioned by the inclusion of most of central Asia in Russia in the middle of the last century. these distinctions have become blurred. In Afghanistan today, among the Turkmen who still keep to traditional designs, a Tekke will weave a Yamut or a Saruq design if he considers that the economic prospect is better. Generally all these varieties are now referred to as Mauri—a designation derived from Marv, the Turkmen center now in the Soviet Union from which most of the best weavers in Afghanistan have migrated.

Most of the Turkmen who traditionally lived in the territory south of the present border between Afghanistan and the Soviet Union were Ersari. The Ersari have always woven rather thicker and less expensive carpets than the other groups. Most of the finest Afghan carpets that are being produced today come from Turkmen settlements in Barmazid. Shakh, Maurichaq and the city of Herat. Most of the Turkmen in these places came into Afghanistan from Soviet Turk menistan in the thirties and carved out small niches for themselves in areas not previously it habited by Turkmen. In the process of adapting to their new environment and their new neighbors they lost some of the distinguishing features of their cultural identity. In dress, male Turkmen now tend to be indistinguishable from their Uzbek or Pushtun neighbors. Turkmen women are seduced.

After every three to four inches of progress the pile is cut down to the desired thickness with shears.
After every three to four inches of progress the pile is cut down to the desired thickness with shears.

According to some of them, for some ten to fifteen years after they arrived in Afghanistan they did not make carpets, and when they were sufficiently well settled in to turn again to carpet making, they had to overcome a break in the continuity of the tradition. They still had in their possession many fine old pieces, which they could use as patterns, and gradually the quality of their product improved and approached tradi­tional levels. The secret of the finest carpets lies in the carding and spinning processes as well as in the weaving. It is among these groups (which are of Tekke origin) that a certain amount of vegetable dyeing is still carried out. In some cases their neighbors from other ethnic groups are learning from them by serving as apprentices. Not being Turkmen they are less tied to the traditional designs. and are experimenting, using for instance conventional Turkmen borders with center pieces derived from Persian designs.

Modern Change

Not all the changes now taking place in the industry are at such high levels of quality. To cope with the enormous increase in demand, especially in lower price ranges, cheaper carpets are now being produced in large quantities. Traditionally in Afghanistan, when town-based merchants financed carpet-production, they did it by giving out wool to the camps and villages where the women would card, spin, dye and weave and return the finished product to the merchant. In this way the merchant promoted production and ensured the size of his own stocks, but had little control over what was produced. Apart from the fact that the quality might suffer somewhat because the women were not weaving for themselves, the carpets that were produced in this manner were of the traditional types that the women had learnt as girls from their own mothers.

High quality carpet in process on a home loom.
High quality carpet in process on a home loom.

During the last decade a few more enterprising merchants, mostly Turkmen, but in direct contact with the emporia in the capital—a situation which is easier now with the great improvement in communications—have set up their own “factories.” In some cases fifty or more looms have been installed in one large building and men are for the first time in this setting employed to do the weaving. This has allowed a break with tradition. To begin with, on their own admission, the quality was not satisfactory but it has already improved. The great advantage—from the merchant’s point of view—is that he can control production. He can decide what size, color, design and (within limits) quality will be produced, watch the process, and enter into more definite contracts with exporters. In some cases these merchants travel abroad themselves and visit their foreign markets in Europe.

Taking the home-made rug to market.
Taking the home-made rug to market.

Most of the new designs and colors that have appeared on the market recently have come from this type of operation. They are directed at a new type of consumer, and for the first time producers, exporters and dealers can be said to be communicating directly and cooperating in the creation of new markets. The introduction of new “Afghan golds” during the last five years is an example of this collaboration.

Besides the Turkmen (otherwise known as Bokhara or Mauri) the other major category of carpets produced in Afghanistan is Baluch. These are mostly made in villages close to the western border of the country. They are generally smaller, more diverse, less densely woven, and therefore cheaper, than the Turkmen carpets. There has been little expansion in this sector, except that individual Baluch now weave more for local markets and less for their own consumption. As with the Turkmen carpets there has been a notice able change in design and color particularly in the cheaper pieces, which is due largely to the desire to adapt to a market, and the consequent mixture of native and alien criteria. Most of the new designs and colors that have appeared on the market recently have come from this type of operation. They are directed at a new type of consumer, and for the first time producers, exporters and dealers can be said to be communicating directly and cooperating in the creation of new markets. The introduction of new “Afghan golds” during the last five years is an example of this collaboration.

The production of Afghan carpets is also being expanded in other ways. Inmates of many of the prisons are employed by the government in carpet weaving. This source, as might be expected, produces items of mixed but seldom very high quality. A further source with perhaps greater potential is in the establishment of “Women’s Institutes” in the provinces. These allow some measure of “liberation” and public life to women who have been traditionally restricted to a domestic sphere of operation.

However, it is difficult to see how any expan­sion can take place in the production of the really highest quality carpets. And it is only these which will in the long term maintain the value and reputation of the Oriental carpet in the West. Furthermore, as the standard of living in Afghan­istan rises, as must steadily happen, the present level of production at this quality is likely to wane-The last part of Lewis’ prophecy may therefore come true after all. As he continues to say: “It therefore behooves us to cherish the Oriental rugs now in our possession.” Not only that. We should also seek to understand better the traditional technologies of their production while they are still there to be studied.

Cite This Article

Spooner, Brian I.. "Afghan Carpets." Expedition Magazine 15, no. 2 (January, 1973): -. Accessed April 18, 2024. https://www.penn.museum/sites/expedition/afghan-carpets/


This digitized article is presented here as a historical reference and may not reflect the current views of the Penn Museum.

Report problems and issues to digitalmedia@pennmuseum.org.