Illustration 1: Yomud khorjin, late 19th century. Overall dimensions, 3'3" x 2'0". Wool warps, weft and pile; braided goat hair closure. Symmetric knots, 10 horizontal by 16 vertical per inch (160 per inch2). Rewoven area approximately 3" x 3" at side of bridge, otherwise original.

On the other hand, there are small numbers of Yomud, Ersari, Beshir, Arabachi, Kizyl Ayak, and Chodor prayer rugs in books and in exhibition and auction catalogs. Those of each tribal group are characterized by different designs and vocabularies of motifs, which suggests that the prayer rug designs do not predate the division of the major Turkoman tribes. In fact, except for the Beshir pieces, which are clearly commercial products woven by sedentary Turkomans, specimens meeting the usual criteria for attribution to a time much before the last quarter of the 19th century are extremely rare.

Prayer rugs that are actually used for Moslem devotions are treated very gently. They are touched with clean hands, they are never walked upon with shoes, and they are not normally exposed to the elements. Therefore, we cannot suppose that there were many older examples but that they didn't survive because of their disintegration from hard use. Since Turkish, Persian, and Caucasian prayer rugs were greatly prized by early collectors, it is unlikely that early Turkoman examples were simply discarded from lack of interest. The conclusion that they weren't being woven in significant numbers before the second half of the 19th century seems unavoidable.

Turkoman khorjin that could reasonably be attributed to the first half of the 19th century are also very rare, perhaps nonexistent. Indeed, there are very few specimens with palettes that are free of obviously synthetic dyes. For this reason, it is generally believed that, like prayer rugs, khorjins were late additions to the Turkoman repertoire of weavings. But unlike prayer rugs, khorjin usually are subjected to hard use and exposure to the elements. Thus, even if there were saddlebags among the earliest Turkoman weavings, older examples would be expected to be rare. Most would simply have disintegrated.

There are extant khorjins made by every Turkoman tribe except the Salor (at least, I am unaware of any Salor or S-group saddlebags). Since Salor weaving essentially ceased around the middle of the 19th century, the absence of Salor specimens could be due to whatever accounts for the paucity of early examples from other tribes. The conventional wisdom is that the explanation is simple: there never were any. That hypothesis will be challenged here.

It is customary to consider types of weavings or features that are common to most or all of the Turkoman tribes to be archaic. Some obvious examples of this are the basic Turkoman bag formats (juvals, torbas, and so on), and the juval gul. Most of us believe that the Turkomans wove juvals, used the juval gul, and made bags and rugs with a layout of columns and rows of repeating guls within a bordered field even before they separated into the major tribal groups. How else are we to account for the ubiquitous occurrence of these features among the products of the various tribes? The mere fact that every Turkoman tribe (with the possible exception of the Salor) seems to have made khorjin suggests that saddlebags were among the early Turkoman products. The khorjin made by the Turkoman tribes resemble each other in a number of characteristics and differ from those made by their neighbors. This, too, suggests that weaving saddlebags is actually a very old custom for the Turkomans.

Illustration 2: Illustration 2: Detail showing the largest area of synthetic dye on khorjin Illustration 1 in The relatively bright red dye is probably Ponceau 2R.

Illustration 1 shows a pair of Yomud khorjin that will serve to illustrate some points. Most of the colors come from natural dyes, although there are a few small areas of wool dyed with a synthetic red on one face (an area that includes Ponceau 2R, the "hot" red, is shown in detail in Illustration 2). The presence of the synthetic dye suggests that the piece was not woven before about 1880; it could not have been woven before 1878. The fact that the amount of synthetically dyed wool in it is so small (less than a total of one square inch of the pile surface) probably means that it was woven before the dye became readily available and inexpensive. On this basis, I estimate that it dates to between 1880 and 1890. Although there is a rewoven area in the bridge, the overall condition is quite good. The bags are stretched somewhat toward their bottoms, as might be expected if they had been used as containers. The pile and the backs show some wear near the lower corners, which also suggests that they had seen utilitarian service. Since most of the popular books on Turkoman weaving do not illustrate saddlebags, references to color photographs of a number of Turkoman khorjin published during the past ten years are provided in Table 1.

Most Turkoman khorjin, regardless of specific tribal origin, use a closure system unlike those found on saddlebags made by their neighbors. Indeed, such a closure system is practically unique to Turkoman khorjin. It is also found on a few Baluch saddlebags, perhaps having been adapted by the Baluch from Turkoman models. Turkoman influences on Baluch weavings are well known. The fact that this closure system occurs on saddlebags (and only on saddlebags) from every major Turkoman group suggests that it is not a recent invention.

Some characteristics of the bridge are common to khorjin of all the major Turkoman tribes. One is its short length. Another is the fact that it is decorated, usually with pile although some have flatwoven brocading. These features, like the closure system, would almost suffice to identify a pair of khorjin as Turkoman even in the absence of any other information about them₁.

Illustration 3: Detail of brocading on reverse side, near lower edge of khorjin in Illustration 1.

Saddlebags from all the Turkoman tribes typically have pile or flatwoven embellishment along the bottoms of the backs of the bags. Illustration 3 shows brocaded decoration on the back of the piece used for illustration. While the Turkomans are not the only weavers who decorate the backs of khorjin, most Turkish, Persian, and Caucasian weaving groups leave them as simple or striped balanced plain weaves, or decorate the entire back rather than just a strip at the bottom. South Persian saddlebags often have a decorated strip of pile that continues from the bottom of the face onto the back.

Thus, there are a number of characteristics of Turkoman saddlebags that suggest a common (therefore, early) origin. The very fact that every Turkoman tribe seems to have made them (the Salor being a possible exception, although the lack of Salor examples would be expected in any case), in itself, suggests that they were not a late addition to the repertoire. A closure system that is unique to Turkoman work, the size of the bridge, and decorating the bridge and the lower part of the backs of the bags are all features that every Turkoman tribe's saddlebags share. It is customary to consider Turkoman designs and characteristics to be archaic if they cross tribal lines. For this reason, I believe that it is likely that khorjin were woven by the Turkomans even during the early period, and that the reason older pieces aren't found is that they disintegrated from being subjected to outdoor use regularly.

Therefore, I suggest that there are different reasons for the rarity of early Turkoman prayer rugs and khorjin. Early prayer rugs probably never existed; early khorjin have probably disintegrated from use and exposure to the elements.

1. Helmut Reinisch, Saddle Bags, Verlag fur Sammler, Graz, 1985. 2. George O'Bannon, Vanishing Jewels, Rochester Museum and science Center, Rochester, NY, 1990. 3. Jon Thompson, Sotheby's, NY catalog for sale on December 16, 1993