Thirty Rug Masterpieces from
the Collection of S. M. Dudin, Part I

by Elena Tsareva

From Oriental Rug Review, Vol. 11/1

Editor's Foreword

When we were asked if we would be interested in publishing Elena Tsareva's manuscript on the Dudin Collection, we responded positively for several reasons. First, Ms. Tsareva is a world renowned scholar of Turkoman carpets, and it is an honor to publish what is in effect the full text of her address to the VI ICOC. We also saw this as an opportunity to publish a full catalog of the Dudin rugs which will be exhibited at the Conference, all of which could not be included in the Pacific Collections catalog.

The challenge was whether time, international communications, and enough effort could be organized to print all the pieces in color. Thanks almost entirely to the untiring efforts of Ms. Tsareva, that has been possible. By publishing this catalog as a part of Oriental Rug Review's I.C.O.C. issue, this Conference, its activities, exhibitions, rug fair, and other activities will remain a permanent part of every Oriental rug enthusiasts' library. We are proud to have had the opportunity to make this contribution.

George W. O'Bannon, Editor

ISBN 0-9628054-0-8
Copyright 1990 Oriental Rug Auction Review, Inc.

Epilogue

The S. M. Dudin Collection of Central Asian Tribal Rugs was to make its U. S. debut at the 6th International Conference on Oriental Carpets at San Francisco, CA. It was to hang November 18 through November 25th.

As those dates approached there were furrowed brows and increasing hand wringing among Conference organizers. The collection was lost in transit. As the Conference was concluding word came that the collection had been located, but it was on its way back to the Soviet Union. So, its' only appearance at the VI ICOC and in the U. S. was in the pages of Oriental Rug Review and, now, here.

About the Author, Elena Tsereva

Elena G. Tsareva is the Curator of Carpets, Felts and Weapons at the State Museum of Ethnography (SME) Peoples of the U.S.S.R. She is in charge of one of the major Oriental carpet collections in the Soviet Union, including the famous Turkoman collections of Bogolubov and Dudin. A graduate of the Oriental Department of Leningrad University, she is the author of Rugs and Carpets from Central Asia, The Russian Collections and numerous other articles in journals and books on Oriental art. She was the curator for the special exhibition of Turkoman carpets at the International Conference on Central Asian Carpets in Leningrad in 1988.

The most remarkable of all known Central Asian rug collections is the pioneer group of textiles collected by Samuil Martynovich Dudin. The majority of the pieces are in the State Museum of Ethnography of the People of the U. S. S. R., Leningrad, and some are in the State Hermitage, Leningrad. The history of how the collection came into existence is also interesting. At the end of the l9th century, the Russian Emperor's family decided to enlarge the Russian Museum and to organize a new department of ethnography.1 In order to form a Central Asian collection, they invited S.M. Dudin (1863-1926),2 already a well known amateur specialist of Oriental art, to travel to Central Asia to compile ethnographic material and create a photographic collection. Later Dudin wrote in his report:

"In the winter of 1900 Academician V.V. Radlov suggested to me that I prepare a short program for a journey to Central Asia, aimed at compiling an ethnographic collection on the Sarts of Russian Turkestan... I prepared a program then submitted it for consideration to the Grand Prince Georgiy Mikhailovich, and it was accepted by him...."3

Making a Rug. Note the manner in which the loom is supported. Also the pile area
in front of the weaver on the right is not even with the rest of the weaving. S.M. Dudin, 1902

From his two main trips - 1901 and 1902 - Dudin returned with an enormous number of objects (2,526), unsurpassed in their richness and variety. Part of them were rugs and carpets. Most of the rugs - 350 pieces - were turned over to the Russian Museum by Dudin, and in 1937 the widow of the artist donated 23 more pieces to the State Hermitage. The difficult conditions of life and travel in Central Asia on the eve of the century, together with the enormous objective of the undertaking, required such remarkable skills and efforts to make the expedition successful that it became an heroic deed. One becomes acutely aware of this when reading Dudin's report of the expedition.

Most of Dudin's rug collection are Turkoman pieces, as he considered the Turkomans to be the best carpet weavers in the world; still Uzbek, Kirghiz, Baluch, and Afghan rugs of Central Asia are well represented. Dudin himself wrote in his report:

"Among the objects of the material culture my interest was aimed mostly at rugs, embroidery and jewelry. Besides the common and artistic interest which they demonstrate, they are the only items through which you can study Turkoman art. I considered it necessary to collect as many of them as possible as these pieces have become rare among the local population. Many years before they were sold to the market following increasing interest in them from the various merchants who sold these objects. The largest number of pieces were obtained naturally enough from the shops and bazaars. Visiting yurts, I aimed at another very important purpose: I wanted to know about the characteristics and ornamentation of the rugs, to confirm all of the terms which the merchants and local ‘experts' use."4

Koshi Village. Obviously a very settled and agricultural environment. Similar scenes
may still be seen today in northern Afghanistan. S.M. Dudin, 1901-1902

So the special value of Dudin's attributions for the pieces in the collection is that he often received information either from the owners or from the merchants who knew the local products well.

Besides the material objects, Dudin made a great photo collection (over 1,800 items) and obtained important theoretical information. Based on his knowledge of the material, Dudin wrote some articles, the most important of which for those studying carpet weaving is "Rugs of Central Asia," published unfortunately only after his death.5

The methodology used by Dudin in studying Central Asian rugs sounds surprisingly up-to-date even today. His ideas and methods were used by those who followed him as well. Admiring the artistic feeling of this great collector and the variety of his interests, one must not go to extremes and canonize his every word. Many of his attributions now look questionable, not unlike the attributions of his contemporaries. First, Dudin got some incorrect ideas from local merchants; second, in Dudin's time no one paid much attention to the technical characteristics associated with the attributions.6 Dudin himself based his attributions on the artistic impression of the rugs and this inevitably led to mistakes. From this point of view Dudin's work can be studied as an historical source. Taking the opinions of Dudin and his contemporaries as a starting point and comparing them with modern knowledge, it is easy to see progress in the field. This principle was taken as a starting point for this introduction, which is why there will be many citations from Dudin's works and long discussions on some subjects.

Spinner, S.M. Dudin

The organizing ethnographic principle chosen for the catalog is the same as that used by Dudin in his article: Turkoman, Uzbek, Kirghiz, and Baluch. Such an approach may seem old-fashioned but as it is on the Dudin Collection, using it as a starting point for showing the difference between today's and yesterday's knowledge of the topic, I considered it logical to follow the principle chosen by the great collector of 90 years ago.

Twiner, S.M. Dudin

SALORS

"The best in all respects of Turkoman pile articles indisputably must be those named Salor or (according to the place of their production in the past) Pendeh. They are also the oldest ones, for until now only among them could one find an example 200 years old and more. Their high textile qualities, amazingly beautiful coloring, and the beauty of ornamentation are highly appreciated not only by Europeans but by the Turkomans of other tribes as well. Undoubtedly this was the reason that even now a great number of their articles are preserved, as they were the very pieces which were in every way possible taken care of, not used in everyday life, and exposed to light only on especially important and exceptional occasions (arrival of guests, festivals, etc.). Passed from generation to generation they lasted in such condition until our times not only preserving all their merits, but bringing nearer to us the fascination of time's patina.

"The group of Pendeh pile items includes properly speaking the works of two tribes - Salors and Saryqs (according to A. Semyonov). But according to my information, gathered in 1901 during my journey to the country on the instructions of the Russian Museum, the Saryqs (at that time) were not engaged in carpet weaving; regarding the Salors, the rugs produced by them at the present time have departed both in quality and in ornamentation from the ‘Salor' carpets in the truest sense, that is from the antique Salor production. They approach the Akhal Tekke rugs, forming an intermediate group between one and another. These later pile pieces really have mixed characteristics and peculiarities of both groups, to an extent where it is often difficult to decide how to attribute them. It must be noted here that with very old Pendeh rugs, kaps7 and so on, there is never any doubt as to their attribution to one and the same group; but the newer the pieces are, the sooner you find doubtful items which you don't know whether to call Salor or Akhal. But if you take into consideration that the basic mass of Salors, being pressed by Saryqs, nearly 80 years ago moved from their ancestral territory across the Persian border and because of the requisitions committed by new authorities, they stopped making carpets. At the same time, the small group which remained among the Saryqs and Akhal Tekkes naturally had to submit slightly to their influence but at the same time exerted some influence upon them. Thus, both phenomena will become understandable. That is: among old pieces we find Salor products made before they abandoned carpet weaving and left the boundaries of their ancient settlement, while in the new products those of mixed type, made some time later - on one hand by Salors who remained on Russian territory and on the other by the representatives of the Akhal and Saryq tribes which lived in the vicinity and adopted some of the Salor decor....

"Study of their ornamentation presents a special interest as we are entitled and desire to find the most refined features of the Turkoman style, especially as the Salors are the most ancient Turkoman tribe from which all the other tribes descended...."

This long citation was necessary to demonstrate in a concentrated way both Dudin's presentation of Salor carpet weaving which remained unchanged during the last 90 years and the ones which differed. So, Dudin's affirmation of the character and importance of early Salor carpet weaving has become common and it was even singled out as an S-Group in recent works. At the same time, knowledge of the characteristic features of Salor carpet weaving is greatly enriched and now there is no difficulty recognizing Salor pieces from Tekke ones (I mean Akhal Tekke) or Saryqs, although Dudin and some of his contemporaries did not single out the latter into a separate group. Today Salor carpet weaving is one of the best known8 and is recognized, concurring with Dudin's opinion, as the classical version of Turkoman carpet production. As Dudin noted, "the starting point" of Turkoman carpet weaving traces from Salor weaving, though he primarily considered the artistic features of the rugs.

Turkoman horseman. Note the embroidered felt saddle blanket and felt bedroll, S.M. Dudin, 1901

Dudin seems to understand Turkoman carpet weaving at its heart. He again and again underlines the major difference between the Central Asian and Caucasian, nomadic and settled carpet weaving. These main differences are: the immobility of design, the setting of color and dynamics of pattern against the ground in Caucasian carpets, and the sublimation of pattern colors with the ground color of the central field in Turkoman rugs.

Going back to Dudin's description of the Salor group of rugs, it is worth noting that Dudin emphasized the great rarity of Salor main carpets:

"Salor carpets in the exact meaning of the word, especially of large dimensions, are a great rarity. At least I, myself, never met them and am given to doubt they were ever produced in any significant number."

This citation makes it possible to conclude that, though Dudin often attributed some fine pieces as old Salor, they could have been Saryqs or Tekkes. He never employed in his attributions the external signs, for example the names of patterns. So he never thought many carpets with the so-called "Salor gul" were produced by this tribe, while other connoisseurs do so practically automatically. In reality the majority of these rugs were produced by Merv Tekkes.9

Dudin's rug collection is very rich in Salor pile pieces in comparison with the number of Salor pieces which have survived to the present: seven items in the State Museum of Ethnography and three in the State Hermitage. Dudin wrote that "... Salor pieces are met with rather often and are not rare...," but it is difficult to agree with him. It seems that he formed this opinion because of his incorrect attributions, identifying as Salor most Saryq weavings, some Tekke and even some Yomud pieces. It doesn't matter now what Dudin thought in his time, as we now know that practically all antique Salor items were treasures. This is both because of their rarity and because of the beauty of the rugs, combined with the highest technical achievement.

1. SALOR Turkoman, JUVAL
4'3"x2'I0" (130x85 cm)
18th-early 19th century, SME No. 26-32

Design: Juval gul,
Warp: Wool, ivory; mixed with gray, Z2S, rather thick, depressed
Weft: Wool, natural light brown and dyed dark brown, Z2S, loose, two shoots, one light, one dark
Knot: Wool, Z2S and Z3S; asymmetric, open right; count: 12 horizontal(46), 17 vertical (68), 104 per sq. inch (3, 128 dm2); pile looks down
Colors: Dark red, crimson red, terracotta red, brown, yellow, dark blue, blue, dark green, ivory; natural dyes
Sides: Three-four warps overcast with red wool
Ends: Upper: Red and ivory flatweave folded over to the front and covered by a sewn on dark blue braid and red tape; Lower: Missing
Purchased: 1901, Merv, "kap Tekke, torba pile woolen"
Published: Tsareva, Rugs; pl. 8
One of the most characteristic features of Salor carpet weaving is its stability, which can be demonstrated in observing any representative object. If we take juval No. 1, the analysis will be: mixed ivory and gray wool warps, Z2S; two shoots of weft, wool, Z2S; weft shoots alternate one light brown and one dyed red; and so on, up to the final, upper end. The top kilim is turned over to the face and covered with a plaited, sewn-on band, this being very characteristic for Salor upper end finishes. The only difference from the majority of other Salor pieces is the absence of silk in the pile. In most Salor items, main carpets among them, silk is used at least in small quantities. Sometimes the area of silk pile occupies up to one-fifth of the ornamented surface.

The stability of Salor carpet weaving is evidenced not only in technique but in ornamentation as well. Hence this juval with 4x4 juval guls, secondary diamond gul pattern, kotshak pattern for the main border, and minor border with chakmak pattern is very characteristic of the type. Slightly less usual is a patternless elem, as most Salor juvals have the kelle pattern in the elem.

2. SALOR Turkoman, TORBA
3'I0"xl'4" (116x41 cm)
18th-early 19th century. SME No. 26-90

Design: Shemle gul
Warp: Wool, ivory, Z2S, depressed
Weft: Wool, brown, mixed with ivory, some red in elem, Z2S, rather fine, two shoots
Knot: Wool and silk, Z2S and Z3S (silk and orange-red); asymmetric open left; count: 14 horizontal (53), 18 vertical (74), 252 per sq. inch (3,930 dm2); pile looks down
Colors: Dark brick red, bright red, pink (silk), crimson (silk), crimson (wool), yellow, violet-brown, dark brown, dark blue, blue, ivory; natural dyes
Sides: Three warps overcast with dark blue wool, red and green plait sewn on the edge which goes into a tape at the upper end (missing) and finishes with a tassel at the lower end
Ends: Upper: Red and ivory flatweave folded over to the face and sewn down, over the edge sewn on red and brown tape with silk embroidery; Lower: Ivory flatweave folded over to the back and sewn down, late multicolored fringe sewn on
Purchased: 1900, Samarkand, "mafrach woolen, pile, of extremely fine design, old Salor"
Published: Tsareva, Rugs, pl. 13; Felkersam, plate, page 13
One of the finest examples of Salor weaving is the shemie torba No. 2. Dudin wrote about such items:

"...among the well made rugs you can more often find hundred year old examples than among the more recent ordinary pieces of the household. Regarding the close relationship of artistic and stylistic aspects to technical quality, there is no need to prove that the finer the weaving, the sooner you can expect a weaver to more precisely depict the artistic taste and style of the tribe; every rug, being either individual or collective work, in fulfillment represents more often the reflection of tribal than of individual taste and style."

Very often these "well made" items differ slightly in technique from the majority of the tribe's production. They have unusual knots, high density of weaving, richer coloring and elaboration of ornaments; their designs are of the most ancient and difficult in fulfillment.

The shemle pattern of torba No. 2, in my opinion, is one of the most antique Turkoman patterns and represents a highly stylized image of the Tree of Life motif. Similar adaptations of this pattern can be seen on contemporary weavings, with the kejebe pattern, while their common ancestor, may be seen in the central field of some Seljuk rugs.10

Close in feeling and related in coloring is the mafrash with the aksu pattern, No. 3. As with most Salor pieces, it is harmonious in appearance and refined in composition. The number of colors used, 11 in the torba and 12 in the mafrash, seems to be the maximum for Salor production.

3. SALOR Turkoman, TORBA
3'4"x1'6" (102x45 cm)
Late 18th or early 19th century, SME No. 26-23

Design: Aksu
Warp: Wool, ivory, Z2S, tightly twisted,depressed
Weft: Wool dark brown 7Z2S, loosely twisted,two shoots
Knot: wool and silk, Z2S, Z3S (silk), Z4S; asymmetric open left; count: 10-12 horizontal (4-47), 17-20 vertical (65-79), 170-240 per square inch (2,700-3,700 dm2)
Colors: Red, crimson red, crimson (silk), pink (silk), terracotta red, violet-brown, dark brown, dark blue, blue, green-blue, ivory, orange (in two places); natural dyes
Sides: Missing
Ends: Upper: Ivory and red flatweave folded over to the back and sewn down, late finish; Lower: Ivory flatweave, missing, remains of dark blue fringe on 8 warps
Purchased: 1901, Merv, "mafrach Salor, pile, woolen, with silk, old"
Published: Tsareva, Rugs, pl. 1, Felkersam plate page 12
Mafrashes, being the items which left the yurt more often than other pile pieces, accompanying their masters during visits to relatives and neighbors, were usually decorated with unusual patterns. The meaning of most of them is hopelessly lost. As to the aksu design, it attracted the attention of rug scholars long ago. At first, it was believed to belong to the Chinese sphere of influence; it was then believed to be purely Central Asian in origin, "having a Chinese shape by chance." Today, I think that the first researchers were correct and that the pattern originated from the town of Ak Su in Chinese Turkestan, which had ancient trade contacts with Central Asia and was famous for its rugs, the aksu composition being rather popular. It can be assumed that the motif came to Central Asia rather early and was surely "homeless" (by contrast with guls, for example), as it was used by all Central Asian carpet weaving peoples, Turkomans, Uzbeks, Kirghiz. It was called "aksu" everywhere. These facts point indirectly to one and the same source, apparently Khotan rugs.

4. SALOR Turkoman, ASMALYK
4'10"x2' (148x62 cm)
Late 18th early 19th century, SME No. 26-19

Design: Kejebe with Single Medallion
Warp: Wool, ivory, Z2S,.depressed
Weft: Wool, brown, Z,2S, in the middle one of sometimes two red wool, two shoots
Knot: Wool and silk, Z2S, Z3S (crimson red and orange-red in elem, ivory and silk); asymmetric open right; count: 12 horizontal (48), 17 vertical (68) 210 per square inch (3,264 dm2)
Colors: Orange-red, crimson red, crimson (silk), yellow (small details), orange, terracotta , red, dark blue, ivory; natural dyes
Sides: Two warps overcast with red wool; red and blue plait sewn to the edge, going into band at upper corner, at the lower end into a tassel
Ends: Upper: Dark blue and ivory flatweave folded over to the back and sewn down; Lower: Ivory flatweave folded over to the back and sewn down; dark blue fringe on 4 warps near the pile, 16" long (41 cm)
Purchased: 1901, Samarkand, "mafrach Salor, pile, woolen, old"
Published: Tsareva, Hali 6/2 1984 p. 131; Felkersam, plate page 12

One of the best examples of Salor production in Dudin's collection is the asmalyk No. 4, notable with a rather rare composition, used also on some camel trappings of the Ersari group. The coloring differs in tints from the previous items and in some ways is close to the previously mentioned Ersaris (Beshir group). As differences in coloring are one of the few current bases for attributions within groups of rugs, it is possible to suggest that the asmalyk was produced by Amu Darya Salors, not by Serakhs Salors, as were the other two bags.

Dudin himself often referred to asmalyks as mafrashes, which undoubtedly points to his bazaar information, merchants having no idea of the use of these items. At the beginning of the century there was a great confusion with terminology; now the situation is standardized,11 and it doesn't matter if some modern names differ from those used by the local population (mostly as regional variants).

5. SALOR Turkoman, KAPUNUK
4'3"x4'3" (128x128 cm)
18th century, SME No. 26-94

Design: Curled Leaf
Warp: Wool, ivory, mixed with brown,Z2S,depressed
Weft: Wool, ivory, mixed with brown, Z2S, very fine, two shoots
Knot: Wool, Z2S, Z3S (white in some places); asymmetric open left; count: 11 horizontal (45); 22 vertical (88), 242 per square inch (3,960 dm2)
Colors: Terracotta red (two tints), orange, yellow, dark brown, dark blue, blue-green, ivory; natural dyes
Sides: Terracotta red wool overcast over 3 warps
Ends: Upper: Red and ivory flatweave folded over to the back and sewn down 1.8 cm wide tassel of olive brown color, with embroidery in pink silk and with orange overwrapping sewn on over the edge; Lower: Similar to upper end in the center, the lower ends of vertical stripes; warp ends knotted in fringe, late silk fringe is sewn on over the original
Purchased: 1900, Samarkand, "kapunnuk, pile, woolen, door surround... Yomuds"
Published: Tsareva, Rugs, pl. 4; Felkersam, plate page 26, detail

6. SARYQ Turkoman, TORBA
3'8"x1'1" (110x33 cm.)
Late 17th/early 18th century, SME No. 26-27


 Design: Diamond gul
Warp: Wool, ivory, Z2S, slightly spun, slightly depressed
Weft: Wool, brown, Z2S, two shoots
Knot: Wool Z2S; symmetric, count: 10 horizontal (41), 26 vertical (103), 260 per square inch (4,223 dm2); uneven knotting; natural dyes
Colors: Cherry red, orange, olive brown, dark blue, blue, green, ivory
Sides: Red-orange wool plait on 3 warps
Ends: Upper: Red and ivory flatweave folded over to back and sewn down; Lower: Ivory flatweave cut down; remains of multicolored fringe on 4 warps, dark blue fringe sewn on later
Purchased: 1901, Merv, "mafrach Tekke, pile, woolen, old"
Published: Tsareva, Rugs, pl. 2l; Tsareva Hali 1/3 1978, p. 10 (top)

7. SARYQ Turkoman, JUVAL
4'4"x2"10" (132x88 cm.)
18th century, SME No. 26-75

Design: Salor gul
Warp: Wool, ivory, Z2S, depressed (?)
Weft: Wool, brown and gray, Z2S, loosely twisted, two shoots
Knot: Wool, silk and cotton Z2S, (crimson red); wool and silk loosely twisted; symmetric; count 13 horizontal (50), 18 vertical (70), 234 per square inch (3,500 dm2)
Colors: Bright cherry red, crimson red, pink (wool) pink (silk), salmon, dark yellow, olive brown; dark blue, dark blue-green, white (cotton); natural dyes
Sides: Missing
Ends: Upper. Ivory flatweave folded over to back and sewn down; Lower: Missing
Purchased: 1901, Samarkand, "Kap, piled, woolen, with silk, Tekke, old"
Published: Tsareva, Rugs, p. 24; Tsareva, Hali 1/3, 1978, p. 279, Hali, 7/3, 1985, p. 19

8. SARYQ Turkoman, MAFRASH
2'3"x1'1" (69x32 cm)
Late 18th-early 19th century. SME No. 26-22

Design: White Panel
Warp: Wool, ivory, some reddish, Z2S, slightly depressed
Weft: Wool, gray-brown, in places brown mixed with ivory, in the center dyed red, Z2S, loosely spun, loosely twisted; two shoots
Knot: Wool and silk, Z2S; symmetric, multitude of offset knots, some sharing knots (two knots on three warps); count: 12 horizontal (47), 16 vertical (64), 192 per square inch (3,008 dm2); pile looks down
Colors: Deep red, orange-red, orange, pink (silk), violet-brown, brown, bright dark blue, bright blue-green, blue, ivory; natura1 dyes
Sides: Dark blue wool plait on 2 pairs of warps; remains of late tassel in the right corner
Ends: Upper: Red and ivory flatweave, folded over to the back and sewn down; Lower: Dark blue flatweave cut down, remains of dark blue fringe on 8 warps close to the pile
Purchased: Merv, "mafrach Salor, pile, woolen, with silk, fringe cut down
Published: First publication

9. (Catalog Cover): SARYQ Turkoman, ENSl
6'4"x4'9" (192x144 cm)
19th century, SME No. 26-17.
Photo by D. Anderson

Design: Ensi Nagysh
Warp: Wool, ivory, Z2S, rather tight, slightly depressed
Weft: Wool, brown, Z2S, fine, two shoots
Knot: Wool and cotton, Z2S; symmetric; count: l0 horizontal (41), 15 vertical (59), 150 per square inch (2,419 dm2)
Colors: Cherry red, orange, olive brown, dark blue, dark green, white (cotton); natural dyes
Sides: Two warps overcase with brown wool
Ends: Upper: Brown flatweave folded over to the back and sewn down, one additional brown weft over two warps on the front; late brown and white loop in the right corner; Lower: 20 cm brown and ivory flatweave with two multicolored plaits
Purchased: 1901, Merv, "ensi, door hanging, Salor pile, woolen, with cotton"
Published: First publication

10. SARYQ Turkoman, TORBA
 4'2"x1'4" (126x40 cm)
First half 19th century, SME No. 37-13

Design: Kejebe
Warp: Wool, ivory mixed with brown, Z2S, Slightly depressed
Weft: Wool, brown, Z2S, loose, two shoots
Knot: Wool, silk and cotton loose, Z2S, silk Z2S and Z4S; symmetric; count: 12- 13 horizontal (48-50), 2.5 vertical (98), 325 per square inch (4, 802 dm2)
Colors: Cherry red, pink (silk), bright red, brown, dark blue, dark blue, green, white (cotton); natural dyes.
Sides: Some warps wrapped over with cherry red wool
Ends: Upper: Dark blue and ivory flatweave folded over to the back and sewn down; Lower: Dark blue and ivory flatweave
Purchased: 1902, "mafrach, Pendeh, woolen, with silk pile"
Published: Tsareva, Rugs, pl. 15; Tsareva Hali 1/3 1978, p. 279 (bottom)

Another even rarer piece in Dudin's collection is a kapunuk, No. 5, which Dudin himself attributed to the Yomuds (similar to many other more or less unusual items such as five-sided asmalyks, khalyks, and so on). Observe the shape of this door surround. It is absolutely square in perimeter and is larger in dimension than Tekke or Saryq ones. The shape demonstrates the highest level of a weaver's skill, while the size must follow the larger dimensions of Salor yurt door openings. It is even possible that Salors used kapunuks to decorate permanent dwellings which Salors constructed according to historical sources as early as their presence in Mangyshlak, e.g. the l7th century and earlier.12

The difference between old Salor and Saryq on the one hand, Salor and Tekke production on the other becomes especially evident when comparing similar form and design items made by different tribes. If you put together the two kapunuks -- No. 5 Salor and No. 14 Tekke -- you will see that although identical in shape and pattern, they present the essence of stylistic differences of both tribes. For the Salors it is an austere style, refined though somewhat darkly colored, striking in depth and harmony of tints -- a vigorous, laconic, and complete composition, possible only thanks to the highest technical achievement. With the Tekkes it is the finest knotting, love for a lighter coloring and minute and bright patterns. I think the best way for the beginner to understand the main peculiarities and distinctions of tribal carpet weaving is through comparison of such pairs as this.13

SARYQS

Dudin combined Saryq and Salor weavings into one group calling them "Pendeh," while we know that at the time of his travel to Central Asia the Pendeh Oasis was inhabited by Saryqs and the Salors lived mostly in Serakhs. Dudin's affirmation that Saryqs did not manufacture carpets even in the 19th century is very strange, and I wonder where he got this incorrect information. Even setting aside the technical features which were not taken into consideration by the scholars at the turn of the 20th century, it is still surprising how Dudin, a person of the highest artistic feeling, did not distinguish Saryq items but attributed them to different tribes, mostly to Salors but also to Tekkes and even to Yomuds.

Possibly the great differences in the age and look of the Saryq pieces bought by Dudin caused this error. The early ones, woven in the 18th-early 19th century, differ greatly from those dating after 1830, the period during which the Saryqs adopted many features of old Salor carpet weaving and thus adopted many of their images. We single out two groups of Salor carpet weaving: before 1830 and after. The same can be said about Saryq and even Tekke pile production, though the reasons will be completely opposite. The year 1832 brought death and destruction to the Salors, while the other two tribes became enriched; thus, it became something of a watershed for Turkoman carpet weaving, a date no less important than 1880 (Battle of Gok Tepe), which inaugurated the common production and merchandising of this craft in Central Asia.

In all there are 16 Saryq pile items in the Dudin Collection at the State Museum of Ethnography plus a yolami in the State Hermltage.14 All of them are true stars of the collectIon and among these, I believe, are the oldest and most beautiful of all Turkoman tent bags, splendid in coloring and design. In describing the Saryq portion of the collection, I arranged the material, in so far as possible; according to the age of the pieces, beginning with the oldest presented in the exhibition. Torba 6 was a very complicated piece to attribute. Dudin himself attributed it to the Tekkes. One can be misled by some peculiarities of the main border design, kojanak, the secondary field motif, chemche, and the coloring of the original fringe (not dark blue as in the 19th century torbas but mixed). But the deep cherry red of the ground and technique - symmetric knot, type of warp and weft - led us to think tnat it is of Saryq production and belongs to the early Amu Darya period of this tribe, e.g. late 17th-early 18th century. Undoubtedly, the guls of the central field are very archaic and ceased being used in a rather early period.

The second Saryq piece - surely of the Amu Darya period - is juval No.7, my favorite in the group. It shows a very popular composition with the so-called Salor gul and an unusual variant of intermediate juval guls. The noble shape of the naldag border and guls and their well spaced distribution on the surface of the bag, gives an impression of them floating in the deep cherry red ground color, glittering with pink silken centers of the guls. All this gives an unforgettable impression and makes this juval one of the most beautiful pieces not only among Saryq weavings but all Turkoman pile items. The prominent artistic values of the juval corresponds with the high craftsmanship: special softness of fabric, shiny silver patina of the pile, and fine correlation of threads organizing the structure of the textile.

Even among the extremely rich Saryq part of Dudin's collection, a unique example is mafrash No.8. At first glance it would be taken for a Tekke Avlia15 but a second glance reassures that it is not. There are four panels instead of the two or three on Tekke pieces, the minor elements of the main motif shows it to be a variation from the Tekke style, and the brighter but, at the same time, softer coloring set it apart. Other points to be mentioned are large quantities of silk in the pile and, of course, the technique: symmetric knot, combined with a multitude of offset and some shared knots, which are extremely rare in Turkoman weaving. Taking into consideration that Avlia, as with other Owlads, were a religious group, not tribal, it is possible to suppose that this mafrash is Saryq-Avlia production. Additional indirect features such as one pile color - a reddish tint with brown - has worn differently and the fact that in 1901 Dudin called the piece "antique" makes it possible to date it to the 18th century, which is again the Amu Darya period.

Ensi No.9 looks later and probably dates to the Merv-Pendeh period. Saryq ensis are the most complicated in composition of these items, representing the Turkomans' cosmic ideas: seven heavens at the top, the earth (zamin) at the bottom, and images of the surrounding world in between. As to the dating, attention should be paid to the extreme unevenness of the pile which Dudin considered to be a special trick, but it is more likely that it is the result of different corrosion rates of the pile over time. From personal experience of those studying carpet weaving, it is known that black dyed or bleached white wool is usually destroyed quicker than that dyed in red, while indigo (dark blue color) conserves wool. Based on these observations, it is usually accepted that pieces with definite differences of dark blue and red pile levels can be dated as 150 years old or older.

The last Saryq weaving presented in the exhibition is torba No. l0 with the kejebe design. The variation of this ancient pattern is one of the most elaborate and leaves no doubt as to the item depicted: it is neither "lamp" nor "sacrificial altar" but the Tree of Life in an extremely expressive shape. This torba is attributed to the classical Pendeh period - middle 19th century. The quality. and mastery of silk and cotton usage makes it a unique item demonstrating the highest level of Saryq weaving.

Dudin paid great attention to the extreme beauty of small pile Turkoman pieces and described this phenomenon in his visual, artistic style:

"Of special beauty and fine manufacture are kaps, mafrashes and asmalyks. In them, mostly in Salors and also in Tekkes, white and pink wool is often replaced by silk of the same colors and also in Salor pieces white wool by cotton which introduces into the pile surface another characteristic which increases their beauty even more. This quality of Turkoman rugs, in addition to other reasons, can be explained by the fact that all items are used, besides their natural purpose, as decoration for yurts and, when put on camels during migrations and wedding ceremonies, served as publicity for the family, testifying silently to the skill of the weavers - brides and wives. It was the competition of the inhabitants of various yurts that created the superb examples of carpet craftsmanship which one admires in his travel on the Turkoman steppes and in local carpet shops."

Notes

1 The State Museum of Ethnography of the Peoples of the U.S.S.R. separated from the Russian Museum and got its modern name in 1934.

2 About S. M. Dudin, see my article "The Dudin Collection," Hali, Vol. 7, No. 3 1985, pp. 14-23. Twenty-nine pieces from the Dudin Collection are illustrated in this article, 19 in color. Only five of the pieces are included in this exhibition.

3 Dudin's Report is preserved in the archives of the SME, f. 1, op.2, d.235 and d.277. 4 In quoting Dudin I have tried to preserve his style which I consider to be very expressive.

5 S. M.. Dudin, "Pile Rug Articles of Central Asia," presented by Academician E. F. Karski at the meeting of the Department of Historical Sciences and Phylogeny on April 21, 1926. Collection Studies of the Museum of Anthropology and Ethnography, v.VII, 1928, Leningrad, S.71-155. Also in German as "Teppiche Mittelasiens," Turkomenenforschung Band 5 , Reinhold Schletzer Verlag, Hamburg, 1984.

6 Dudin's article was an attempt to compliment and correct in some details the works of his contemporaries: A.A. Bogolubov, Carpets of Central Asia , St. Petersburg, 1908 (in English,Carpets of Central Asia , Crosby Press, Fishguard, 1973); A.A. Felkerzam, Old Carpets of Central Asia, Old Years October-December, 1914-June, 1915 (in German Alte Teppische Mittelasiens, Reinhold Schletzer Verlag, Hamburg, 1979); S.S. Semyonov, "Carpets of Russian Turkestan in connection with the edition of Carpets of Central Asia from Bogolubov's Collection, St. Petersburg, 1907-08", Ethnnographical Review, Nos. 1-2, 1911, pp. 137-179.

7 Kap is a term used by most of the early Russian collectors. They used it to refer to the various sizes of bags made by Central Asian weavers, including juvals, mafrashes, torbas, and khorjins.

8 The most important works on Salor carpet weaving are: Azadi, S., Turkoman Carpets and the Ethnographic Significance of their Ornaments , Crosby Press, Fishguard, 1975; Moshkova, V.G., Carpets and Rugs of the Peoples of Central Asia in the Late 19th and Early 20th Centuries, Tashkent, 1970 (in English, Oriental Rug Review Vols. III and IV); and Mackie, L. and Thompson, J., Turkmen, The Textile Museum, Washington, D.C., 1980.

For a bibliography of the Salors and their history, see in E. G. Tsareva, "Salor Carpets," Hali, Vol. 6, No. 2, 1984. A Salor kejebe asmalyk collected by Dudin and not included in the exhibition is illustrated in this article.

9 I wish to add some comments about the term "Salor gul," which many scholars considered for many years to be the tribal Salor gul and hopelessly tried to find it on Salor main carpets. Though it is now known what the real Salor gul looks like, there still exists an undying idea that this unique, complex, and beautiful element is a real Salor gul (in Russian its name is "Salor Rose"). The Salors themselves named it sagdag (that is Sogdian), also "Mary" or sometimes maida gul . There is a fine mess with rug terms thanks to merchants, who gave their own "trade" names to rugs and their ornaments, "Salor gul" being one of them, as it was often used in the products of this tribe. The same happened, for example, with the term "Bukhara carpets," which merchants used to designate half of all Turkoman rugs.

10 Serare Yetkin, Historical Turkish Carpets, Turkiye is Bankasi Cultural Publication, Istanbul, 1981, pl. 2.

11 The most thorough description of the types of Turkoman woven items are described and illustrated in the aforementioned works of S. Azadi and V. G. Moshkova.

12 A. Dizikiev, "On the History of the Distribution of the Salor Turkomans from the l6th - beginning of the 20th c.", Studies in Turkoman Ethnography, edited by S.G. Agajanov and A. Orazov, Ashkhabad, 1965 (in Russian).

l3 For further analogues, see Saryq kapunuk No. 87-9 in Tsareva, E.G., Rugs and Carpets from Central Asia, Aurora Art Publishers, Leningrad, 1984, pl. 19; and a similarly shaped and patterned item from the Ashkabad Collection, No. K-165, Dzumaniyazova, M., (sic Shakhberdyeva M.), "The Ashkhabad Turkomans," Hali, Vol. 10, No. 1, 1988, p. 40.

l4 see literature listed in Notes for Tsareva, E.G., "Saryk Tent Bags in the State Museum of Ethnography of the Peoples of the U.S.S.R.," Hali, Vol. 1, No. 3, 1978, p. 280. This article illustrates 13 Saryq pieces collected by Dudin in black and white, 10 of which are not included in this exhibition.

15 Avlias and Owlads are religious families descended from the Prophet Mohammed. About Owlads, see Demidov, S. Turkoman Owlads (in Russian), Ashkhabad, 1976; Atayev, K., "Some Data on the Ethnography of Turkoman Shikhs' in Trudy Instuta istorii, Archeologii i Ethnografiya, ANTSSR, v. 7, Ashkhabad, 1963 (in Russian).

Go to Part II

From Oriental Rug Review, Vol. 11/1

Go to Oriental Rug Review Home Page