نواحی و مناطق قالی بافی

In Search of the Turkmen Carpet

This article tries to illustrate the richness of firsthand descriptions of Turkmen weaving and the nature of the historic sources which contain them.


One group of documents consists of indigenous biographies, histories, and travelogues. There is, for example, the history of Mir Abdoul Kerim Boukhary I and what it has to say about the location and condition of Turkmen tribes early in the 19th century. An account of a diplomatic mission by Riza Quoly Khan is similarly useful for a picture of the fractious and independent Turkmen of the 1830s.2 This literature, however, tends to be descriptively weak concerning ordinary artifacts.
  Olufsen offers a photograph (1) of such a rug, now in a Danish museum. What he means by "mosque-pishtak" is explicit: "...there is opposite to the entrance porch a porchlike structure with a large pointed niche, called pishtak... This pishtak dominates the whole complex of buildings. It looks toward Mecca, and here the ecclesiastic who presides...takes his seat...."5. The gate is accordingly established as an image for the prayer rug in Central Asia at this time.

Plate 1. Olufsen offers a photograph of such a rug, now in a Danish museum. What he means by "mosque-pishtak" is explicit: "...there is opposite to the entrance porch a porchlike structure with a large pointed niche, called pishtak... This pishtak dominates the whole complex of buildings. It looks toward Mecca, and here the ecclesiastic who presides...takes his seat....". The gate is accordingly established as an image for the prayer rug in Central Asia at this time.. -RW
Another rich vein is accounts by outsiders, especially those of Russian veteran Central Asian hands. An excellent general account of the Turkmens was written by Kuropatkin in 1879. He names, locates, and counts the principal groupings (306,220 tents altogether); he describes tribal conflict and identifies trade patterns. He accurately sums the political situation: "The authority of Bokhara, Khiva, and Persia on certain Turkmen tribes, such as, for example, the Ersari, the Tchodors, the Iomuds, the Goklans, so scarcely affirmed, began little by little to strengthen and to turn from nominal (power) to actual (power)." He quotes a Turkmen saying: "No Persian is able to enter Turkmen territory other than at the end of a rope,"3 that is, as a slave.   Kuropatkin, as well as a few other Europeans, understood the sociology of the Turkmens. Each tribe, village, and family contained sedentary (chomur) and pastoral (charva) elements, reflecting a caste distinction which turned on herd ownership. Status, however, was not immutable. To maintain flocks was to have means; to till the soil was to be poor. The poor and the well-off were mingled.

A further important document stream consists of Russian government reports concerning the kustar (home) industries. These reviews cover all production from 1883 on 4 in Transcaspia district, and thus portray Turkmen weaving except for that of the Bokhara Khanate, an autonomous area outside the normal Imperial administrative structure.

Plate 2. Inclusion of Turkmen carpets in this bazaar description is typical. Observation after observation, photograph on photograph indicates that the Turkmen rug was a staple furnishing in Central Asia. The Bokhara lady at leisure in the 1890s illustrates the point. - RW
A final resource is the photograph. With the introduction of rolled film circa 1890, both the Central Asian traveler and resident functionary had equipment for recording the appearance of textiles, among them many Turkmen rugs. The camera is powerful in that it renders a rug specific in appearance, time, and place. These pictures - and they are abundant - are an important aspect of the written record. But, it is crucial to get timeand place straight. It is also necessary to remember that at this time photographs were posed, not candid.

These various sources add up to a rug history. They permit intercomparison and are useful tests for the principal Russian rug dissertations (Felkersam, 1914; Dudin, 1928).

  The paragraphs which follow present details on both Turkmen weaving and the nature of the source materials.

There is, naturally, a geography involved. A simple, but not wrong, view would categorize Turkmen tribal groups as Western (Chodor, Yomud, Goklen), Eastern (Tekke, Salor, Saryq), and Northern (Ersari). There were, of course, others. Roughly, the Western and Eastern Turkmens occupied the then Transcaspian district, now the Turkmen S.S.R., and the Ersaris were located to the north along the southeastern rim of what was the Bokhara Khanate and is now the Uzbek S.S.R.

Plate 3. The photograph of the doughty Annette Meakin, taken in 1896 or 1898, shows her with recognizable textiles, one of which she and others regularly refer to as "Pende." - RW
In the Bazaars

Marketplaces are frequently described. Sometimes the relation of contents becomes fairly detailed. During his 1898-99 trip, O. Olufsen toured Bokhara bazaars:

"Close to the harness bazaar is that of the saddle-bags (kurdjum), wallets of woven carpeting laid across the saddle... These kurdjums, often beautifully woven in geometrical patterns in yellow, red, brown, blue, and black contain provisions for the rider for several days.

Far out in the northern part of the town the carpet dealers have their bazaar, Tim-i-Gilam, the principal part of which is made up


of an open square place encircled by open shops with stores of thousands and thousands of carpets. Here we find the carpets with nap (gilam) both from Bokhara and the neighbouring countries, Persia, Turkestan, Afghanistan, Belutchistan and from the Kirghiz in Turkestan, carpets without nap (pallas) generally with a white ground and decorated with various coloured stripes or geometrical figures and splendid red, blue, black, or white felt carpets (kigis), sometimes with patterns in geometrical designs, all the latter made by the Kirghiz tribes.

Plate 4. Illustrated is a formal portrait of the pious, young khan of Charjui, taken there sometime between 1880 and 1882. The photo shows a popular pattern of likely Tekke origin involving a "turret" main gul drawn in a height-width ratio of 1:1:1. - RW
The carpets made in Bokhara and, as the others, on primitive looms in the open air, are long and narrow and in quality rank between the gilam of the Kirghiz and the Youmuts; they can neither as to solidity nor beauty be compared with those of the Turkomans or Persians but they surpass considerably the Kashgar, Afghan, Belutshee and common Kirghiz carpets. The colours of the Bokhara gilam are mostly brown or red with black, yellow, green or white geometrical patterns. A special sort, used as prayer-carpets in the mosques are, as a rule, red with a mosque-pishtak in white."   Olufsen offers a photograph (1) of such a rug, now in a Danish museum. What he means by "mosque-pishtak" is explicit: "...there is opposite to the entrance porch a porchlike structure with a large pointed niche, called pishtak... This pishtak dominates the whole complex of buildings. It looks toward Mecca, and here the ecclesiastic who presides...takes his seat...."5. The gate is accordingly established as an image for the prayer rug in Central Asia at this time.


Plate 5. "The Turcoman carpets, too, were very much in demand, and sold readily, in spite of the high prices demanded for them ..." The bazaar in Merv, circa 1918 with a stack of Tekke Turkmen main carpets being offered for sale.
Inclusion of Turkmen carpets in this bazaar description is typical. Observation after observation, photograph on photograph indicates that the Turkmen rug was a staple furnishing in Central Asia. The Bokhara lady at leisure (2) in the 1890s illustrates the point. Numerous writers use the phrase "Turkmen and Bokhara rugs"; these were the two recognized genres. The presence of Turkmen, as distinct from Bokhara, rugs can be noted in urban places throughout western (Russian) Central Asia. The rugs were, in one scholar's phrase, woven by Turkmens to suit an Uzbek taste.

Russian military activity gave rise to a process which might ironically be termed a special kind of bazaar. Here is an account of Samuel Gourovitch concerning the assault on the Akhal Tekke at Geok Tepe in 1882:

  "The day after the storm of Geok Tepe he (Gourovitch) assisted in the pillage of the Tekke fortress, and secured a large number of valuable carpets, which, however, were taken from him by the military authorities... He had a tolerably good specimen of a small Merv carpet, for which he asked 15 roubles, and another from Geok Tepe, for which he wanted six... (and he had) several carpets, some pierced with bullets and bought with other loot from the soldiers at Geok Tepe...."6

A similar situation arose after the defeat of Khiva in 1872 and involved collection of an indemnity which the Russians levied on the Khiva Yomuds, who had offered the only neaningful military resistance.


Plate 6. A Tekke Turkmen family in their yurt, with utensils for spinning wool. Note the pile of rugs behind them, suggesting they were engaged in making rugs for commerce. - TC
"The Turcoman carpets, too, were very much in demand, and sold readily, in spite of the high prices demanded for them and of the fact that hundreds had been "looted" in the campaign against the Yomuds. A carpet, four yards long by two wide, brought 4 to 5 (pounds). A curious feature of the sale was, that although the Turcomans must have been hard pressed for money to pay the indemnity, they could not be induced to lower their prices a single kopek. They simply named their price, and you might take the article or leave it, as you pleased. The carpets are made by the women, and will compare favourably with the best carpets made anywhere. Each family has a different pattern, which is handed down from generation to generation as an heirloom, without undergoing the slightest change. The colours are principally red and white, interspersed with small patches of green and brown, and are really very pretty, as well as durable. 7

What is pointed up by these stories of looting and booty is that Turkmen carpets were an accepted item of value.

  In Situ

A difficulty with bazaar observations is the possibility of mistaken attribution. Such is not a problem for on site accounts. Two rather interesting observations involve the sedentary Goklen who occupied a small area within Persian jurisdiction. Here Fraser in 1825 remarked on the weaving of both felts and carpets.8

Yate, however, in 1894 described a different situation:

"The interior of their kibitkas was even dirty too, and they had none of the cleanliness and fine carpets and wall-bags of the Tekkes and Sariks... The Goklans did not appear to me to be such an industrious race as their brethren the Tekkes or the Sariks. They made no carpets, and only a few coarse rugs. Felts apparently were their only manufacture....9"

Plate 7. A Yomud woman at work, weaving within the confines of the yurt in which she lives. Note the flat woven chuvals suspended by ropes to the interior lattice of the yurt, situated just behind her. - TC
Two observations do not a conclusion make; it is the case, however, that attribution of rugs to the Goklen must deal with early and late characterizations which are quite different.

An on-site report of 1884 originates with Lessar, a Central Asian veteran, later to become Russian political agent in Bokhara. He toured Transcaspia's eastern extremity to the Pende oasis and the Persian border town of Seraks. Among other things he describes the weaving activity of the Saryq Turkmens:

"There are also hand-made goods which the Saryks export.

1) First place among them is taken by rugs; their pattern is slightly different from those of Merv, and the quality is worse than those of Merv, due to the admixture of cotton and absence of an admixture of silk. In the oasis, due to the lack of mulberry trees,

  silkworm breeding is not practiced. The price of the rugs is nearly the same as those from Merv.

2) Felts are produced in significant amounts in Pende; a piece 5 arshins (11 feet) in length and approximately 3 (7 feet) in width costs 10 krans.

3) From the hair of young camels (one to two years old) a beautiful material for robes is made. One woman can make one piece of the material about 9 arshins (21 feet) long and fourteen to fifteen vershoks (1 foot, 3 inches) wide. This material is very highly valued in Persia and Herat; one piece costs 200-300 krans. Among the Saryks themselves there are not people wealthy enough to wear these robes... Other goods manufactured by the Saryks' artisans serve only to satisfy local needs."10

Plate 8. Members of the ruling family in Khiva, shown sitting on a Chodor carpet displaying the classic 'ertmen gol' pattern, taken no later than 1896. - TC


These are solid particulars about a woven product, fixed in time and place. The photograph (3) of the doughty Annette Meakin, taken in 1902, shows her with recognizable textiles, one of which she and others regularly refer to as "Pende."

Weaving in Ersari territory is glowingly described by Logofet who was there shortly after the turn of the century: "Kerki and all of the left bank of the Amu-Dariya to the kishlak Bassaga has been from earliest times a center of rug production in the Bukhara domain. Passed from generation to generation, this art in the above-mentioned region, was brought to a high level of market.

  perfection; thereby Kerki rugs, or as they are often called, kizil-ayak, from the name of the village which is 40 versts from Kerki, justifiably, after Tekkes, occupy second place in the rug commodityLately the quality of their production has deteriorated and the main reason will be found in the fact that the producers of the rugs could not afford to dye their wool with expensive vegetable dyes, in view of the fact that the market price of the rugs is at a comparatively low level, and the expenditure of time and effort in the production of each square arshine of rug is tremendous. So counterfeiting touched the rug industry as well.11"

Plate 9. A photograph by S. Prokudin-Gorskii showing two Tekke Turkmen posing for the camera in front of their yurt, with a main carpet on the ground in front of them. Photo taken circa 1918 - TC
Quite a bit has been written about this particular weaving locale. Felkersam, author of the standard work on Central Asian weaving, reports an earlier description (1902, by Laurenti) of "Kerki and Bashiri" weaving.12 Of considerable interest in this regard is Komarov's census of 32 Kerki district and 26 Charjui district hamlets in the Amu Daria left bank territory for the year 1886. While he does give particulars about a hamlet, Kizil-Ayak (300 family units), it is his general description of the area which is significant. He identifies the four Ersari tribal wings -- one of which is the Kizil-Ayak -- and their locations in the Kerki and   Charjui bekdoms. He leaves no doubt that the entire area was irrigated, and hence had sedentary, not nomadic, population.13

In the early 20th century Bokhara Khanate commercial weaving activity may have been extensive. Wm. Eleroy Curtis visited Bokhara City in 1910 and asserted that every house had a loom. While he certainly did not limit home looms to rug weaving, his description of locally produced rugs is specific and concrete. He noted that they had 10 or 12 designs, but only four or five were commonly used.14 Who the weavers were he does not say, but the rug he discusses is the Bokhara type.

Plate 10. Turkmen weaving in a photo taken by Dudin, circa 1902 - TC
In Commerce

The historic sources document an established, longstanding Turkmen rug trade vis-a-vis Central Asia and northeastern Persia. Fraser had extensive experience along the Russo-Persian border. Writing after his trip of 1822-23 he stated: "The manufactures of the Toorkomans consist chiefly in carpets, which they weave of very beautiful fabrik, and which are highly prized, fetching very huge prices. They are chiefly of the twilled sort, but they also make them of a fabrick resembling the best Turkey carpets, and of very brilliant patterns. They seldom exceed in size an oblong of from twelve to sixteen by eight or ten feet, and for the most part are greatly smaller. They also weave cloth of camels' hair, and coarse woollens, chiefly for their own use; as well as numuds (felts) of an inferior quality."15 At this time the term manufactures has a clear commercial connotation, underscored by Fraser's later reference to products made for home use.

Here is a general comment from the early 1880s on the home industry of Tekke Turkmens, by one who unwillingly lived with them for several months:

"When a Turcoman is blessed with a large number of daughters, he contrives to realize a considerable sum per annum by the felt and other carpets which they make. In this case an ev (felt tent) is set apart as a workshop, and three or four girls are usually occupied upon each carpet, sometimes for a couple of months.


Each girl generally manufactures two extra fine carpets, to form part of her dowry when she marries. When this has been done, she devotes herself to producing goods for the markets at Meshed and Bokhara, where the Turcoman carpets fetch a much higher price than those manufactured in Khorassan or beyond the Oxus. Sometimes these carpets are made partly of silk, bought from Bokhara. They are generally twice the size of the ordinary ones, which are made from sheep's wool and camel hair mingled with a little cotton, and are almost entirely of silk. They fetch enormous prices."16

Illustrated (4) is a formal portrait of the pious, young khan of Charjui, taken there sometime between 1880 and 1882. The photo shows a popular pattern of likely Tekke origin involving a "turret" main gul drawn in a height-width ratio of 1:1:1. The photo demonstrates that this gul form was extant and in fashion at this time. A well known photograph of a Tekke weaving shed showing work underway appears in Felkersam (1914), but in fact the picture was taken in 1902. Depicted is a flattened form of the same gul with a quite different height-width ratio. The two photographs may be indicators of the time this motif changed shape.

Plate 11.
A Tekke Turkmen woman working at a loom. Photo taken by Dudin, circa 1902 - TC
Inside the Bureaucracy

At the same time Curtis was writing about Bokhara City, the Statistical Committee, Transcaspia Oblast, was fully describing the kustar industry of the Turkmens. Both Merv and Ashkabad had warehouses for furnishing the weavers with vegetable dyes and high grade wools. A government effort to improve the quality of the product in this way had begun around 1907. All of the woven products are reported on and include rugs, kilims, saddle-bags, prayer rugs, juvals, torbas, and tent bands. In 1911 Merv produced over 1,000 rugs and more than 2,000 juvals, torbas, and tent bands (combined). Ashkabad wove but 200 rugs but

nearly 4,000 tent bands.17 There was a large overseas export; in the year 1913 the product value of overseas shipments was 1,244,000 roubles, 62% of an annual production estimated at 40,000 pieces.18

Furthermore, an established weaving industry existed in the heart of Turkmen territory in the early 1880s. While the data are sketchier, a Transcaspian commerce in woven products is described and the same product line revealed.19 There are several particulars:

Plate 12. A Tekke family posing for a photograph, circa 1918. Photo by S. Prokudin-Gorskii - TC
One, primarily a curiosity, is the fact that during the years 1884 and 1885 the Merv Tekkes were making Persian design rugs. There is, of course, a principle here: design shifts can easily occur in commercial weaving. The serious question, not illuminated by this episode, is whether weaving for the trade differed in any important way from weaving for home use. The record, so far at any rate, suggests they were the same.

Another item of interest is the statement that other than for one local plant used to produce a yellow dye, dyes were imported from Khiva and Bokhara. This practice means that dyes for Turkmen rugs, particularly Yomud, have to be seen in this perspective.

  A matter of deeper significance is revealed in these reports: they state it was the poor who wove. How matters might have worked is hinted at by the nature of Turkmen society. Since chomur and charva lived and worked together, labor supply was co-located with raw material. Indeed, while the reports note hamlets specializing in the home industry of silvermaking, they make no mention of weaving hamlets.

Kustar activity also extends into the Soviet period. Circa 1930, the product categories were: Tekke, Pendi, Yomud, Kerki, Ensi, Juval, and Torba, and various small products including door surrounds and tent bands. In other words, the traditional product mix continued. There is, in short, a considerable continuity to commercial Turkmen weaving, with no a priori distinctions between the pre-revolutionary and Soviet periods.

Plate 13. Photographed no later than 1896, the retinue of the Khan of Khiva is shown here, a Yomud group 'kepse' gol rug at their feet. - TC
In Other Regards

While the home industry included piled and flatwoven products, the dominant product was felts, a thing to be expected in such a society. All was not monochromatic tent frame and floor covering. Here is an observation made by Orsolle among the Yomuds in the 1880s: "...these Turcomen are occupied peacefully in the sale of thick felt rugs, decorated with red and green arabesques...."20. Moser at about this time was further east, with the Akhal Tekke, and commented: "The woman who marries brings as dowry a certain number of felts which she has made in her spare time, among which must be included a very fine covering for the horse of her spouse."21

The sources make it clear that felts could be a quality decorated object and were an important element of both the weaving economy and of Turkmen society.

The many photographs of and statements about the dwelling door curtain indicate overwhelmingly that they, too, were made of felt. There does not seem to be a photograph of a Turkmen pile door covering; one etching is apparently all there is by way of visual evidence. The kustar documents do not help, for the term ensi is not used; ensi is a Soviet period label. Rather, these reports speak of "prayer carpets" for both Eastern and Western areas. Whether what was being reported was misnamed ensi or proper prayer carpets is not clear.


There is also information from the past concerning dyes. The list of Schuyler, one of the standard encyclopedic sources on Central Asia, is typical. He identifies the local sources for yellow and black and mentions both local madder and imported indigo. Concerning red he observes: "Cochineal is frequently used for dyeing silk red. It is chiefly brought from Bukhara, although the insect is found in abundance in the spring in Tashkent and the neighborhood, on the young leaves of the ash, mulberry and other trees. Since the introduction of fuchsine from Russia, the use of cochineal and of other native dyes have fallen off. For that reason in Khokand the Khan prohibited (in 1876) the importation of fuchsine, as being an inferior dyestuff."22

Schuyler's treatment can be taken as typical; an equivalent source of the same period (1880s) would be Lansdell.23. Later (1902) Annette Meakin identified dyestuffs in Bokhara as indigo from India, cochineal from Russia, native madder, and sophora japonica for yellow.24

Chemical analysis of dyes is an important research tool. But the historic sources reveal the messiness of reality: an apparent widespread use of imported dyes by the Turkmens, a concentrated effort to import Bokhara natural dyes into Turkmen territory circa 1910, and the apparent use in Bokhara of "Russian," i.e. imported, insect red as late as 1902. The trek from dye analysis to date is slippery, and any attempt to use synthetic red as an absolute dating method would be simplistic..

Plate 14. A Tekke Turkmen rug vendor in the Merv bazaar, circa 1918- TC
In Perspective

Unquestionably many as yet unknown sources of information about Turkmen weaving await discovery. Since the material is there and wants only the digging, there is no great hurry to reach conclusions about the Turkmen oeuvre.

But two or three notions insist on emerging now. One is the unremarkable fact that the 20th century is proximate to the l9th. Given the existence of an active Turkmen weaving for regional and subsequently international commerce during the late l9th/early 20th centuries, there should be some good reason (or at least some reasons) to look elsewhere for the origins of the many Turkmen rugs surviving today.


Another is that Turkmen weaving for the Central Asian market may extend well back in time. The record consistently notes the distinction between weaving for commerce and weaving for home use, with the former present throughout the l9th century. Central Asia has been Uzbek dominated since 1600 and Turkic for much longer. One possibility is a longstanding Turkmen weaving role in this culture.

What matters most, however, is the esthetics of Turkmen weaving. Here the roots may be deep. Color combinations and patterns (be they girik knot) or tamgha (clan brand) or something else) may well have developed long ago. The terms to be encountered in such a quest -- Sogdian, Oghuz, Sassanian, Kushan, etc. -- will differ radically from those of latter days. This task lies ahead.


1. Mir Abdoul Kerim Boukhary, Histoire de il'Asie Centrale, ed. and trans. Ch Schefer, Paris, 1876.

2. Riza Quoly Khan, Relation de l'Ambassade au Kharezm, ed. and trans., Chas. Schefer, Paris, 1 876, p. 58.

3. Kuropatkin, Lt. Col. A., Voienny Sbornik, Sept.-Oct., 879, p. 322.

4. Obzor Zacaspii'ski Oblast; vol. I, 1883-90, annually hereafter until 1916.

5. Olufsen, O. , The Emir of Bokhara and his Country, London, 1911, pp.329, 332, 546.

6. Marvin, Charles, The Region of Eternal Fire, London, 1888, p. 71.

7. MacGahan, J.A., Campaigning on the Oxus..., New York, 874, p. 412.

8. Fraser, James B., Narrative of a Journey..., London, 1825, p. 602.

9. Yate, Lt. Col. C.E., Khurasan and Sistan, London, 1900, p. 233-4.

10. Lessar, P.M., Yugo-zapadnaya Turkmeniya, 1885, p. 53.

11. Logofet, D. N., Na Granitsakh Srednei Azii, St. Petersburg, 1909, Vol. I, p. 124-5.

12. Felkersam, Baron A., Starye Gody, April-May, 1915, p. 5.

13. Komarov, Capt. G. Sh., Sbornik geograficheskikh...po Azii, Vol. 25, p. 278 ff.

14. Curtis, Wm. Eleroy, Turkestan, The Heart of Central Asia, New York, 1911, p. 167.

15. Fraser, James B., Into Khorassan, Appendix B, Part II, p. 41.

16. O'Donovan, Edmund, The Merv Oasis, New York, 1883, Vol. II, p. 252.

17. Obzor zakaspiiskoi oblasti za 1911 g., Ashkabad, 1890, p. 104 ff.

18. Kustarnaya I remeslennaya promyshlennost' natsionalnikh respublik, Moscow, 1926, p. 7.
19.Obsor zakaspiiskoi oblasti, Ashkabad, 1890, p. 104 ff.

20. Orsolle, Le Caucase et la Perse, Paris, 1885, p. 330.

21. Moser, Henri, A Travers l'Asie Central, Paris, 1885, p. 330.

22. Schuyler, Eugene, Turkestan, New York, 1876, Vol. I, p. 182.

23. Lansdell, Henry, Russian Central Asia, Boston, 1885, Vol. I, p. 482.

24. Meakin, Annette M.B., In Russian Turkestan, London, 1903, p. 38

This is an adaptation of part of a paper given by Richard E. Wright at the Interatrional Conference on Central Asian Carpets in Leningrad in 1988.
Copyright 1989 by Richard Wright. Original text and photographs appeared in Oriental Rug Review, June/July, 1989
With much thanks, this article was reproduced with permission by Ron O'Callaghan and Richard Wright
No parts of this text nor the photos may be reproduced in any form without permission from Ron O'Callaghan, Richard Wright or myself

by Richard Wright

Richard Wright's continuing research appears at -

Originally appeared in Oriental Rug Review, Vol. 9, #5
Original captions with corrections provided by Richard Wright, noted as (- RW)
Addtiional captions provided by Tom Cole where noted as ( - TC)

منبع :