<---previous article

Dividing the Chodor

by Kurt Munkacsi

Originally appeared in HALI 77 (Oct/Nov. 1994)


by Tom Cole

The weavings as well as textiiles of the Chodor are very different in many respects from that of their neighbors in Central Asia. The textiles represent a completely different design tradtiion, seemingly closer to ancient Seljuk architectural work rather than the design pool observed in other Turkmen texiles, ie. those of the Tekke or Yomud. Apparently in the past, the Afghan marketplace considered Chodor embroidery to be non Turkmen Central Asian material from Ugench, rather than Turkmen.

Since this article was written in 1994, other groups of Chodor weaving have emerged, the most notable one being a group with deeply depressed warps, exhibitng characteristics of many of the types discussed below. Provenance as well as period of weaving has much to do with all of this, and the real story of the Chodor has yet to be completely revealed as far as I know. What I do know is that some of their weavings are truly spectacular but in a different sense that I get from studying a great torba or chuval of one of the other Turkmen weaving groups, including the Saryk, Tekke, Yomud or Salor.


Plate 1. Chodor Group III (?) main carpet fragment, Khorezm region(?), ca. 1740(?). 0.76 x 0.46m (2'6" x 1'6"). This very old fragment is clearly related to a piece advertised by George Gilmore in The 1994 HALI Annual (p.118). The gül centres in particular are identical. I am not entirely certain that either of these pieces really belong to Group III. If they do, it is possible that they were woven in Khorezm prior to the Chodor being forced out by Nadir Shah. The very early form of the zig-zag lattice is of particular interest here, as it appears within the güls rather than outside them. Courtesy Peter Hoffmeister, Dörfles Esbach.

I have been a ‘hard-core’ collector of Turkoman rugs for about eight years. Ever since I acquired my first few pieces, the Chodor have held a special place in my heart. I don’t know what it is about them, but when one day I walked into a well-known Bay Area gallery and saw a Chodor chuval hanging on the wall, I just had to have it. Then another came up at auction and the same thing happened, and again, and before I realised it I had wound up with a large number of these pieces..

Why did I acquire so many? Maybe it’s because the Chodor were one of the original twenty-four Oghuz tribes as long ago as the 16th century. Maybe it’s because while still being part of the Turkoman design pool they also seem to stand apart. After looking carefully at this Chodor material for a few years and

trying to make some sense of the designs, I began to understand them, and definable groups started to coalesce. I noticed that the majority of weavings with the so-called Ertmen-gül design could be placed into one or another of these groups.

At this point I started asking myself the classic questions about Turkoman weaving: when and where? In each of the groups I could see the designs changing and evolving, the influence of other tribes (especially the Yomut) creeping in, as well as different palettes and structures and materials, particularly with regard to the amount of cotton in their construction. Was it possible to use these factors to date and place the Chodor?

17th Century (Map A)

• The Chodor are located in the northeast Caspian region on the Mangyshlak and Buzachi Peninsulas.
• The Yomut are located further south, with summer pastures in the Balkhan Mountains.

Early 18th Century
• The Kalmuks drive a group of Chodor into the northern Caucasus, where some still remain.


In 1990 an exhibition of the Amstey Collection was held in Rochester, New York. Its catalogue, Vanishing Jewels, contains what is probably one of the most important historical articles written in English about the Turko-man, ‘Turk-men Ethnohistory’ by William Wood.1 This covers the movements and interactions

of each of the main tribes from the mid 18th century to the present. Mr Wood’s sources and references are excellent, especially with regard to Russian material, which is usually inaccessible in the West. I therefore decided to see whether the groups of weavings I had already tentatively identified might possibly match his account of their history and movements, as summarised below and depicted in the series of maps (A-H):

Circa 1720 (Map B)

• The Kalmuks force the majority of Chodor from the Caspian to the Khorezm oasis region on the lower Amu Darya River.

• The Yomut form permanent settlements along the rivers and canals of the Amu Darya Delta.

• Possible first Chodor interaction with the Yomut.

Circa 1740 (Map C)

• Nadir Shah drives a group of Chodor from Khorezm to Charjui on the middle Amu Darya, where they remain to the present day.

• The Yomut briefly seize control of Khiva from the Uzbeks.

Circa 1760
• The Chodor join the Yomut in revolt against the Uzbek Mohammed Amin Inaq and seize control of Khiva.
• Second Chodor interaction with the Yomut.


Before going into details of the individual groups of Chodor Ertmen-gül weavings, I would like to make some general observations. The first concerns structural characteristics. Differences among the chuvals and torbas presented here are given in a detailed table (see Appendix). However, all the Chodor pieces analysed share a number of characteristics. Warps tend to be of undyed Z2S wool (goat hair, sheep’s wool or camel hair?), in a variety of natural colours and mixtures. There are two weft shoots. The pile is wool, unless otherwise mentioned, and the knot is invariably asymmetric, open on the right.

Moving to design, one element in particular unifies the Ertmen design and separates it from the rest of the Turkoman repertoire: the zig-zag lattice that surrounds the güls and fills the spaces between them. This lattice is not present in any other classic Turkoman gül designs and seems to have undergone several stages of development.

I believe the oldest form of lattice to be the simplest, a zig-zag following the outer edge of the gül, rather like a lightning bolt, which imparts a sense of movement to the güls. This may be seen in the form of the lattices on chuvals (4) and (12), which I consider to be very old pieces, with the latter probably the older of the two.

1768 (Map D)
• The Chodor, defeated by the Uzbeks, return to Mangyshlak.
• The Yomut, driven from Khorezm, return to nomadic life; some in the Caspian region, others west of Ilyali.

1779 (Map E)
• The Yomut return to Khiva to help suppress Uzbek rebels.

• The Chodor return to the Khorezm area.
• The Yomut return to the lower Gazavat and the area north and west of Kunya Urgench.
• Third Chodor interaction with the Yomut.

1811 (Map F)
• Part of the Chodor return to Mangyshlak.

1835 (Map G)
• The Aday Kazakhs drive the Chodor from Mangyshlak, for the last time, forcing a return to Khorezm.
• Fourth Chodor interaction with the Yomut.

1860 (Map H)
• The Chodor concentrate around Porsu, where they remain.

• Russian conquest of Khiva.
• Subjugated Chodor remain in the area.
• The Yomut are slaughtered and driven out of the city.

The next stage of lattice development can be seen on chuvals (2) and (3), where ‘box’ devices are starting to develop. The lattice on (3) is a particularly interesting variation, though am not sure what the weaver was trying to achieve. In the last stage of lattice development (8), the box devices have become ‘box flowers’, while the lattice itself no longer has an ‘electric’ feel and has almost become a trellis supporting the flowers.

What might be a precursor to the lattice appears on an early main carpet fragment (1) published by Peter Hoffmeister.2 Here the incipient lattice consists of a line of small diamonds and squares just within the gül, following its basic outline. It is as if the lattice is just starting to emerge from the gül.

A related chuval belonging to George Gilmore appeared in The 1994 HALI Annual.3 In this, the lattice, now outside the gül, retains the same basic outline as the güls in the Hoffmeister fragment, even though the shape of the güls themselves is different.

Another item of interest concerns the so-called ‘minor’ güls. In all descriptions of Chodor design that I have read, the blue güls are invariably described as ‘minor’ and the red and white ones as ‘major’. I think this assumption should be rever-sed. If anything, it seems that the blue güls, being more informative and specific, define the identity of the weavings. In the divisions I have established, the blue güls are the main identifiers.

Plate 2. Chodor Group I chuval, Amu Darya Delta (?), ca. 1780 (?). 1.12 x 0.91m (3' 8" x 3' 0"). Ground colour and con-struction link this ‘blue güls with bars’ chuval with (3), (4) and (5). It has an early design without Yomut influence, and an old form of the lattice. All illustrations Nancy Jeffries & Kurt Munkacsi Collection, New York, unless otherwise credited.

Below - Detail of Plate 2

Pieces started to fit very nicely within the tentative groups. I was discovering. Colours, relative amounts of cotton in the foundation and design influences all seemed to make sense, but I was very surprised when I attempted to fit pieces and groups within the historical chronology established by William Wood. The usual time scale for these weavings was expanded by his account by 75 to a hundred years, and pieces that I would normally have dated to the first quarter of the 19th century seemed to fall into the first half of the 18th century. For the sake of argument I am going to use these earlier dates for the pieces in this article. We will see if they hold up as time goes on.


I have been able to fit the Chodor Ertmen-gül rugs and bags into four main groups and two subgroups. Further divisions

will doubtless come to light. One thing is certain with tribal weavings – just when you think you have it all figured out, something happens to surprise you.

My groups and subgroups are as follows:

I.             Blue Güls With Bars – type piece (2),                                                 ‘pinwheel’ subgroup – type piece (10);

II.            Blue Güls With Stars – type piece (11),                                                 subgroup – type piece (C);5

III.            Missing Blue Güls – type piece (18);

IV.            Tall Blue Güls – type piece (22).6

Plate 3. Chodor Group I chuval, Amu Darya Delta (?), ca. 1780 (?). 1.40 x 0.79m (4' 7" x 2' 7"). Close to (2) in palette, design and structure, this chuval has an old form of the zig-zag lattice and great scale of drawing. It may even date from before the Chodor joined the Yomut uprising against the Uzbeks in the 1760s.

Below - Detail of Plate 3


This is the largest of the groups and shows the highest incidence of design and structural variation (see Appendix). These pieces must have been made by the main body of the Chodor tribe. Several different palettes are used, with varying amounts of cotton in the foundation (from all-wool to all-cotton wefts). Some have elems and borders with Yomut designs.

I believe the amount of cotton used in a piece can indicate its place of origin. Khorezm in the lower Amu Darya (between Khiva and the Aral Sea) is an oasis region and to this day cotton growing is very important there. It may therefore be argued that

the less cotton we find, the further away we are from Khorezm. I think these varying features can be used to locate these pieces in place and time.

There are two chuvals that I believe are from the oldest stratum of this division, (2) and (3). Their lattices are of the older form and their designs do not show any Yomut influence. They are similar in colours and structure, though while (3) has only cotton warps, (2) has a mixture of cotton and wool. Because of the large amount of cotton, brownish-purple ground colour and pure Chodor design, I think these pieces may date back to the first half of the 18th century, when the tribe first arrived in Khorezm from the Mangyshlak Peninsula (Map B).

Plate 4. Chodor Group I chuval, Amu Darya Delta (?), ca. 1800 (?). 1.14 x 0.76m (3' 9" x 2' 6"). The form of the lattice suggests that this chuval, with its Yomut-type elem, is somewhat older than (5).

Two other pieces that seem to belong in this group are (4) and (5). In palette and structure they are virtually identical to the previous two bags, but we also begin to see Yomut influences in the design, especially in the elems. While (4) appears to be the older of the two, based on its lattice, the elem of (5) has an old Yomut design. These two bags may therefore be dated after about 1760-1770, as might another piece in my

collection (28). This is the period when the Chodor joined the Yomut in alliance against the Uzbek Muhammad Amin Inaq. By 1768 they had been defeated and forced from Khorezm back to Mangyshlak (Map H).

There is also a subgroup associated with Group I, the ‘pinwheel’ group, which includes (10) as well as a chuval belonging to Charles Grant Ellis exhibited in Washington DC in 1980,7 and a torba in the Straka Collection.8 The shape of the blue gül indicates these pieces are related to each other. The palette, use of cotton and Yomut-type elem might link the Ellis piece with those mentioned above. The Straka piece is perhaps later. While (10) is very similar to the Ellis piece in design, it has a much more purple ground. Because of the palette and absence of cotton from the foundation, I think it was made after the second Chodor return from Mangyshlak, in about 1835, probably in the Porsu area (Map G). Chuval (8), while not in this subgroup, is probably contemporaneous with (10), as its palette is similar and it too lacks cotton in its foundation.

Plate 5. Chodor Group I chuval, Amu Darya Delta (?), ca. 1810 (?). 1.17 x 0.84 (3' 10" x 2' 9"). Both palette and structure link this slightly later bag, which combines Yomut influence in its elem and main border with an early lattice form, to (2), (3) and (4).

Below - Detail Plate 5

A chuval with a similar, if slightly darker, palette (6) also lacks cotton and is much more finely knotted. This might be

an example of the tribe’s second Mangyshlak period, from 1811-1835 (Map F). Both (7) and (25) might be examples from the later Porsu period, 1860 to the present (Map H). These have fully developed lattices, use cotton in the found-ation, have a deep purple ground colour and aflat yellow absent from older pieces.

Group I attributes include:

• The blue güls contain a vertical pole in the centre with kochaks at both top and bottom and an ashik-type device through the centre.

• There is either a red or white gül in the centre of the field.

• The red and white güls contain vertical poles with opposing hooked devices top and bottom, and are trisected by three diamonds, dividing the pole into four equal segments.

• The red and white güls never contain ‘animal’ or ‘bird’ heads.

• Four complete blue güls surround the central red or white gül.

• The blue güls usually have two shades of blue or blue-green, diagonally opposed.

• The gül shapes are serrated and tend to be elongated, the blue güls being more elongated than the red or white ones.

• Rounded ‘S’ forms are often used in the minor borders.

• This group seems to have the most Yomut-influenced design elements.

• This group has the greatest use of cotton in the wefts.

plate 6

Plate 6. (Chodor Group I chuval, Mangyshlak Peninsula (?), ca. 1830 (?). 1.09 x 0.79m (3' 7" x 2' 7"). A Yomut elem design and the absence of cotton in the foundation indicate this chuval was probably made after the Chodor and Yomut joined forces against the Uzbeks, and were driven from Khorezm back to Mangyshlak. It shows the final stage of lattice development.

Below - Detail of Plate 6


This is a very small group in which I am able to place only five chuvals – (11), (12), (14), (29) and one belonging to Dr Erich Menzel9 – as well as several torbas, including (26), and a few main carpets. They are all so distinctive that I am sure they form a separate group.

One of these chuvals (12) is very old. The lattice is of the old form. Even by traditional standards it would probably be dated to thefirst quarter of the 19th century. By the expanded scale I am using it probably originates from when the Chodor first arrived in the Khorezm region on the lower Amu Darya, 1720-1760 (Map B).

The next two pieces, (11) and (29), are very similar to each other and must be from the same place and period, although they are later than (12). They have cotton in the wefts, indicating that they are probably still from Khorezm.

The torba (26) may be contemporaneous. All have an early form of lattice. I believe these three pieces probably date from around 1770, just before the first return to Mangyshlak (Map D). The last two chuvals in this group are later. Chuval (14) has a later stage of lattice and a typical Yomut elem design, but still uses cotton in the foundation. The Menzel chuval lacks any lattice at all, suggesting to me a later date. These latter two pieces could be from as late as 1880.

There are several main carpets in this group. One in particular, exhibited in Hamburg in June 1993,10 has coloured wefts similar to chuvals (11) and (29). I should also mention a torba, shown in the same exhibition, with coloured wefts but lacking cotton in the foundation,11 indicating it was made away from Khorezm, perhaps during the second return to Mangyshlak.

Plate 7. Chodor Group I chuval, Porsu region (?),ca. 1860 (?). 1.09 x 0.89m (3' 7" x 2' 11"). I think of this and (25) as coming from the same area, at the start of the tribe’s decline. This chuval nevertheless has a beautifully drawn version of a classic Chodor design.

Below - Detail of Plate 7

Two additional torbas, (13) and (15), are either part of,or related to, this group, but seem to be later than the pieces described above. They lack the lattice and (15) has dyrnak güls in place of blue güls. A piece illustrated by Werner Loges may also be a very late example of this type.12

There also appears to be a Group II subgroup, in which the stars at the centre of the blue güls have diamond devices above and below (also found in the blue güls of Group IV pieces). These weavings also tend to use a green for the ground of some of the ‘blue’ güls. I know of four chuvals – (17), one published by Raymond Benardout,13 one advertised by Jay Jones,14 and one in a private collection – as well as one torba fragment (16).

Group II attributes include:

• Blue güls with eight-pointed star centres, the stars containing a rectangle with a diamond centre.

• The main field design on the chuvals always has three complete blue güls, with the central blue gül surrounded by red and white güls.

• Along with the standard Chodor natural wool and white cotton wefts, many pieces also have some coloured wefts.

• The palette is consistent, especially the brown-purple ground colour.

• The red and white güls always contain well articulated ‘bird’ or ‘animal’ heads with what appears to be a ‘tail’.

• When used, the ‘S’ minor border is of the squared ‘S’ type.

• The main chuval border is typically (four out of five) of the ‘X’ and diamond type.

• The blue güls are more elongated horizontally than those in Group I.

• The warps are on one level, without depression.

Plate 8 Chodor Group I chuval, Mangyshlak Peninsula (?), ca. 1830-1840 (?). 0.91 x 0.71m (3' 0" x 2' 4"). Cotton is absent from the foundation of this piece, which also shows the last stage of lattice development, perhaps indicating that it was woven when the Chodor were in Mangyshlak rather than in Khorezm.

Plate 9 Chodor Group I torba, Khiva region (?), ca. 1835 (?). 1.17 x 0.43m (3' 10" x 1' 5"). Chuvals of this group greatly outnumber torbas. Perhaps made during the second return to Khorezm, this torba has an inter-mediate lattice and uses a lovely mid blue colour.

The ground colour is much browner than that found among the rest of the Chodor. Most of what we call ‘rose’ palette pieces belong to this group. The design appears to be very stable over time, indicating a long undisturbed presence in one location. There is hardly any Yomut influence – only two of the published main carpets show any Yomut design features.15 The shape of the güls on these carpets is not consistent with the rest of this group as they lack the ‘stepped’ outline.

I have already mentioned two very old pieces in this group in relation to lattice development (1). They could even be ‘precursors’, perhaps woven in Khorezm before the Chodor were forced out in 1740. Next comes a cluster of three main carpets and one small rug, which link the ‘precursors’ with the rest of the group. Their main feature is the interior design of the güls, which have the ‘animal’ headed box devices of later pieces but lack the vertically stacked ashiks. I believe the ‘animal’ heads in these güls derive from a different iconographic source, and are not a degenerate version of those seen in Groups II and IV.

Plate 10. Chodor Group I chuval, ‘pinwheel’ subgroup, Khorezm (?), ca. 1835 (?). 1.19 x 0.76m (3' 11" x 2' 6"). This chuval may have been made when the Chodor returned from Man-gy-shlak for the last time. The main border shows obvious Yomut influence.

Below - Detail of Plate 10

This intermediate cluster comprises a small rug in the Straka Collection,16 which does not have any cotton and with güls arranged vertically rather than diagonally; a main carpet offered by Lefevre & Partners in London in 1976, which has all-cotton wefts and a few knots of silk in the pile;17 a main carpet in The Textile Museum, Washington DC (exhibited in 1980), with the güls arranged vertically and part-cotton wefts;18 and a main carpet with a very odd arrangement of güls (21), exhibited by Eberhart Herrmann in 1980.19

This cluster is not as old as the two pieces discussed earlier. I would date them to about 1800 – even though they have cotton in the foundation, I think they were woven in Charjui. They all

have the same fairly old version of the lattice and a typical Chodor use of an ashik-based design in the elems. As far as one can judge from published photographs, they also all have the ‘rose’ type palette. One has a few knots of silk in the pile, which also turns up in a chuval from this group (19).

This brings us to the remainder of Group III. The oldest chuval from this section, once with Simon Crosby,20 lacks cotton in its foundation, has the typical güls and palette of this group, no Yomut design influence and an old form of lattice. Chuval (18) is particularly interesting, for while güls, elem and structure are typical, its palette is closer to traditional Chodor colours and the main border shows Yomut influence, perhaps the consequence of intermarriage.

Plate 11. Chodor Group II chuval, lower Amu Darya region (?), ca. 1770 (?). 1.17 x 0.71m (3' 10" x 2' 4"). Though later than (12) this chuval still uses an early form of the lattice. It has a typical Chodor ashik elem.

Detail of Plate 11

Another chuval that merits discussion is (19). It differs in that while the güls are arranged by colour in diagonal rows, the colours themselves are inconsistent with the rest of the group. The elem too has an unusual kejebe type design. While this design appears on many torbas attributed to the Chodor, I have seen it in the elems of only two chuvals, this and (27), which is so late that it almost does not count. Chuval (19) also has an intense ‘rose’- type palette. Initially, because of its condition,

I thought it was very old, but now I am not so sure. The elem design is related to that of (27). It has one further unusual feature

that sets it apart from almost all other Chodors I have seen – a very small amount of red silk in the pile which appears to have been unravelled from a piece of insect-dyed silk cloth.

I have been able to find only two torbas that are obviously part of this group, (20) and one illustrated in the Macmillan Atlas of Rugs and Carpets.21 Another torba in my collection could be placed in this group from the point of view of palette and structure, but it has a kejebe field design.

Plate 12. Chodor Group II chuval, lower Amu Darya region (?), ca. 1760 (?). 1.02 x 0.84 (3' 4" x 2' 9"). The oldest of the ‘blue güls with stars’ group, with an early zig-zag lattice and an old Yomut main border design.

Below - Detail of Plate 12

Group III attributes include:

• No blue identifier güls.

• The güls have interiors of four box devices with ‘animal’ heads, two either side of three vertically stacked ashiks.

• Güls are arranged in diagonal rows of alternating colour.

• Typically the diagonal rows are alternately red and white.

• There appears to be little or no cotton in the foundation.

• The palette of much of this group is the ‘rose’ type.

• Most of the elems have an ashik-based design.

• The güls have a ‘stepped outline’ more like typical Turkoman chuval güls.

• Light blue or blue-green is used, especially in the elems

of main carpets, and to outline güls.

• Hardly any Yomut influence is detectable.

Bogolyubov illustrates a typical main carpet of this group.22 He states that it was purchased in Khiva, but in my opinion it was made in Charjui. As I have noted above in respect of tribal weaving, one never knows when pieces will appear that contradict one’s theories. One such, a main carpet offered at auction in Germany in 1991,23 has Group II blue güls mixed with typical Group III güls. To confuse things further they are arranged vertically and the lattice form is unique in my experience.

Plate 13. Chodor Group II torba, Khorezm (?), ca. 1850 (?). 0.99 x 0.41m (3' 3" x 1' 4"). In this example, formerly in the Dr Jon Thompson Collection, the lattice has disappeared and the interior of the blue gül is changing. This probably represents the first step of this group being absorbed by the Chubbash. The next stage of design development can be seen in (15).


This is the smallest of the four main groups and the most enigmatic. It is represented by one chuval in the Toni Woger Col-lection at the Völkerkunde Museum, Munich (22) and five torbas. These include the oft-published piece in the John Gilbert Collection;24 one formerly in the Jon Thompson Collection, sold in 1993 at Sotheby’s, New York (23),25 one in the Wher Collection, Switz-erland;26 one published by Werner Loges,27 and one in the Van Aardenne Collection at the Rijks-museum, Amsterdam. I do not know of any main carpets, ensis or other items that obviously belong to this group. Perhaps some of the main carpets with tauk-nuska güls or ensis with other designs could fit, but determining this is a task for a future study.

When I first looked at this group I thought that it must surely be the oldest of all. The designs were so well executed, the lustrous colours so beautiful and the knotting so very fine, that these

apparently ‘prototypical’ weavings must have come from the earliest times when the Chodor were still on the Mangy-shlak and Buzachi Peninsulas.

But now I am not so certain, mainly because the condition of all the pieces in this group appears too good for them to be truly ancient.

Judging from published illustrations, the palette is very similar in all the pieces, except the Thompson torba, which has a browner field and lacks the beautiful clear light blue typical of the group. The iconography is virtually the same in all the pieces. Of the three technical analyses published, two torbas (Gilbert and Thompson) lack cotton in the foundation, while one (Loges) has both cotton and wool wefts. The diamond ‘eye’ device in the blue güls is also used in the Group II subgroup.

Plate 14. Chodor Group II chuval, lower Amu Darya region (?), ca. 1840 (?). 1.35 x 0.81m (4' 5" x 2' 8"). Linked by palette to the rest of this group, this fairly late chuval with cotton in its foundation shows the final stage of zig-zag lattice development in combination with an obvious Yomut-style elem.

Below - Detail of Plate 14

I would guess that this group was woven in the span of one or two generations by a wealthy family. This speculation is based on their superb dyes, the quality of wool, excellent workmanship (workwomanship?) and wonderful detailed drawing. The lack of surviving main carpets or other weavings suggests that few were actually woven.

I think the Wher piece is the oldest, followed by the Munich and Gilbert pieces (which must be contemporaneous), then the Loges piece, with the youngest being the Thompson one. I think they probably originate from the time when the Chodor briefly shared control of Khiva with the Yomut in the latter part of the 18th century.

Plate 15. Chodor Group II torba, northwest Afghanistan (?), ca. 1870 (?). 1.19 x 0.36m (3' 11" x 1' 2") Perhaps a Chubbash weaving – the lattice is lost, the blue güls have been replaced by dyrnak güls and the palette has changed. The main border is similar to (13).

Group IV attributes include:

• The blue güls are tall and elongated.

• The interior design of the blue güls has horizontal poles, as opposed to the vertically oriented ones in Group I, with kochaks at the ends. At the centre of the poles is an ashik with an arrow and ‘bird’ head device in the interior.

• The red and white güls have extremely well articulated ‘animals’ in their interiors.

• All have high knot counts and use very fine materials.

• The design is consistent, with very little variation.

• The palette includes a very beautiful clear light blue.

• A box and kochak device is used in the main border, similar to one found in Salor and early Tekke work.

Plate 16. Chodor Group II subgroup torba, Khorezm (?), ca. 1830 (?). 1.47 x 0.48m (4' 10" x 1' 7"). A good example showing an early form of lattice and a bottle-green shade typical of this very small subgroup.

Plate 17. Chodor Group II subgroup chuval, Khorezm (?), ca. 1850 (?). 1.27 x 0.76m (4' 2" x 2' 6"). The red and white güls are related to those of Group III, but the star in the centre of the blue güls places this chuval in the Group II subgroup. This dating is based on the form of the lattice. Courtesy Eberhart Herrmann, Munich.

Below - Detail of Plate 17


There is a group of Yomut torbas that seem to be ‘copies’ of Group IV pieces (24). They are without doubt Yomut, based

on their symmetrically knotted structure and their palette .28 Most of them appear to be very old, perhaps from around 1760 when the Chodor and Yomut seized control of Khiva.

I should re-emphasise that, with one notable exception,29 I believe the Chodor used rectangular trappings instead of five-

sided asmalyks, and that I have used the term torba interchangeably with wedding trapping. I also wish to note that while I have concentrated on Chodor weavings with the Ertmen-gül in this study, and have omitted all other designs attributed to the tribe, most notably main carpets with tauk nuska güls, torbas with the kejebe design and the group of chuvals that use chuval güls, I nevertheless believe that the methodology and groupings I have proposed could easily be extended to encompass all Chodor weaving.

Plate18. Chodor Group III chuval, Charjui region (?), ca. 1850 (?). 1.19 x 0.69m (3' 11" x 2' 3"). Absence of cotton and the ‘missing blue güls’ design suggest that this mid-period piece is the work of the Chodor whom Nadir Shah forced to migrate to Charjui around 1740.

Plate 19. Chodor Group III chuval, Charjui region (?), ca. 1860-1870 (?). 1.12 x 0.76m (3' 8" x 2' 6"). The intense ‘rose’ palette and the use of a light blue that fades to grey indicate to me that this is a mid-to-late period Charjui piece. The presence of a few silk knots in the pile is very untypical of Chodor work.

Above - Plate 20. Chodor Group III torba, Charjui region (?), ca. 1860 (?). 1.27 x 0.43m (4' 2" x 1' 5") One of only two torbas published from the ‘missing blue güls’ group. The form of the lattice indicates it is from the later period in Charjui on the middle Amu Darya.

Below - Plate 21. Chodor Group III main carpet, Charjui region (?), ca. 1800. 1.80m x 2.77m (5'1 0" x 9' 1"). The güls are typical of a second period middle Amu Darya ‘missing blue güls’ piece, although their arrangement is unusual. Courtesy Eberhart Herrmann, Munich.

Plate 22. Chodor Group IV chuval, Khiva region (?), ca. 1780 (?). 1.21 x 0.85m (4' 0" x 2' 9"). One of the most beautiful of all Chodor Ertmen-gül pieces, this is the only published chuval from the ‘tall blue güls’ group. It appears to be closely related to the famous large torba or wedding trapping published by Mackie & Thompson and others. Woger Collection, Staatliche Museum für Völkerkunde, Munich, inv.no.86-308029.

Below - Detail of Plate 22

Plate 23. Chodor Group IV torba, Khiva region (?), ca. 1860. 1.68 x 0.66m (5' 6" x 2' 2"). In most respects this is the least and latest member of a distinguished design group, lacking the intense colour and lustrous wool of other examples. Courtesy Sotheby’s, New York.

Plate 24.. Yomut ‘tall blue güls’ design torba, Khiva region (?), ca. 1760 (?). 1.07 x 0.46m (3' 6" x 1' 6"). This symmetric knotted torba is one of a number of Yomut versions of a Chodor Group IV torba, perhaps woven in the period when the Yomut, in alliance with the Chodor, seized control of Khiva from the Uzbeks.

Plate 25. Chodor Group I chuval, Porsu region (?) ca. 1870 (?) (3' 11" x 2' 2"), The chuval, which is slightly later than (7) and with a similar palette, was probably made after the Choodor returned to the Khva regino from Mangyshlak for the last time.

Plate 26. Chodor Group II torba, Amu Darya Delta (?), ca. 1830 (?), 1.07 x 0.38m (3' 6" x 1' 3". A piecee with an early form of the lattice and the brownish palette which I associate with teh lower Amu Darya region.

Plate 27. Chodor(?) Group III chuval, Charjui or nothern Afghanistan(?), ca. 1900 or later (?), 1.07 x 0.76m (3' 6" x 2' 6"). Thsi very late example of the "missing blue gols" group includes a synthetic red dye. It has a Yomut minor border, and could even be of Yomut origin.

Plate 28. Chodor Group I chuval, Kunya Urgentch region,(?), circa 1875 (?), 1.09 x 0.71m (3' 7" x 2' 4"). This is something of a problem piece as the palette is not easy to place, perhaps Kunya Urgentch, rather than Porsu. The design of the lattice and the fairly compressed gul probably indicate a fairly late date, perhaps after 1860. It has a typical Yomut elem design.

Plate 29. Chodor Group II chuval, lower Amu Darya region (?), ca 1820(?), 1.14 x 0.71m (3' 9" x 2' 4"). Later than (12) but contemporaneous with (11), this chuval with its typical Chodor ashik elem retains an early form of lattice.

Below - Detail of Plate 29