The piece is a mafrash, a bag most likely used to hold personal belongings like a woman's jewelry. The face is 1'0"x2'1". The fringe is original and the goathair hanging ropes probably are, too.
The pile is of asymmetric knots open to the right, which indicates that it is Tekke, as do the brown wool outlining the narrow bands of pile and the slight depression of alternate warps. Although some pieces attributed to the S-group use asymmetric knots open to the right, S-group weaving is also characterized by outlining with brown wool dyed with a corrosive black and by deeply depressed alternate warps. The piece is very finely knotted by the standards of any Turkoman group: 12 knots per inch horizontally and 32 to 36 per inch vertically. The three white stripes are flatwoven cotton; the decoration on them is wool pile. The only Turkoman weavings with pile decoration on a flatwoven ground of which I was aware before seeing this bag are tentbands. In tentbands, the pile generally uses knots on three warps each; in this piece the knots in the pile decorating the flatweave are, like those in the other pile areas, of the usual asymmetric form around two warps.
The three broad red stripes are flatwoven wool. Each encloses three narrow red bands of pile that are essentially isolated, typical, Turkoman secondary borders. The ivory ground panel at the bottom is wool pile, but there are tiny purple silk triangles, less than a quarter inch on the side, scattered within the main motif.
The panel's main motif itself looks "tentbandish" to occidental eyes, although I've not seen any actual tentband with this design in it. Of course, just because a design reminds us foreigners of tentbands doesn't mean that it is a tentband design. Practically any Turkoman weaving with predominantly red and blue pile decoration on an ivory or white ground will look "tentbandish" to most of us.
The care put into the weaving of this piece is evident even in the pair of very narrow stripes in the topmost flatwoven band. Superficially, these appear to just be a few colored weft shoots in a band of white cotton wefting, and of no real importance. Each actually consists of alternating red and blue segments, achieved by using floating wefts in two colors rather than single weft shoots across the entire width of the bag.
The design itself is nearly unique. Interestingly, the only published analogy of which I am aware is more than just similar to it. The Amstey collection includes a pair of donkeybags with almost identical layout, motifs, materials and designs, a little more finely woven in the horizontal direction. The drawing of the motifs and the colors in the donkeybags differ somewhat from those in this mafrash, and they do not have pile decorated flatweave in their structure. I know of no other piece with a design resembling this one, and of none at all including this combination of weaving techniques.
The ak-juval is the most common type of Turkoman bag that is made with alternating stripes of pile and flatweave, although there are torbas and saddlebags with this arrangement (Amstey donkeybags, for example). To my knowledge, there are no other known mafrash using this format. Like the subject of this article, ak-juvals have a wide pile band at the bottom of the face, usually with an ivory ground. It is generally accepted that ak-juvals were stacked horizontally in the tent, with the bottom ends facing the tent interior. Thus, the wide pile bands at the bottoms of the faces would be the only part of the bag that could be seen except on the one atop the stack. This presumably explains the relatively sparse decoration everywhere else on the face.
Furthermore, it is often asserted, the alternating pile and flatweave bands helped to prevent the stacked bags from sliding off each other. How the stripes could reduce the tendency to slip is not immediately obvious from examining ak-juvals or photos of them. Pile occurs only on the face and never on the back, so the pile and plainweave on consecutively stacked bags wouldn't interlock unless the decorated surfaces faced each other, and even in that seemingly unlikely arrangement only two consecutive bags could be stabilized. In fact, Turkoman bags are woven as pairs along one set of warp cords, the pile faces flanking the plainweave on the loom. Most are then cut into two separate bags. In the case of ak-juvals, this is not done, so they resemble a pair of enormous saddlebags and are stacked with the pile faces on the outside of each pair.
There is no doubt that the subject mafrash was made to be hung. This is suggested by the goathair ropes having colored wool fringes at their ends and by the fringe at the bottom of the bag, which would make no sense at all in a bag that lies flat on its back. Nevertheless, the piece includes all the features -- alternating pile and flatweave stripes, a broad decorative band at the bottom of the face -- that are so conveniently rationalized for bags intended to be stacked. Even the touches of silk in the mafrash are restricted to the bottom panel. It is interesting that, if this piece were in the juval size and missing its fringe, nearly any expert would unhesitatingly aver that it was clearly intended for horizontal stacking.
What about age? There is hardly anything in attribution of Oriental weavings that is more likely to be in error than dating tribal pieces. Still, we persist in doing so, probably as much a reflection of hubris as of expertise. The uncertainty is even greater in pieces that are very unusual, and especially so when, like this one, it is nearly unique. (What would an older one look like? A younger one?) The bag has seen some service, as the inside and back are dirty and gray, although the face is bright and clean; apparently the surface of the face has been washed. All of the dyes appear to be natural, which suggests a 19th century rather than 20th century origin. The fineness of the weave and the attention to detail is consistent with this.
As a general rule, early Tekke work (prior to 1850) includes rather small amounts of silk when it is used at all; lavish use of silk is associated with later dates. The amount of silk in this mafrash could hardly be smaller without eliminating it altogether; there is a total of less than a half square inch of silk pile in the piece.
Unusual motifs can be taken to suggest either great age -- "dead" motifs that were already out of use when most existing pieces were woven -- or relatively recent production -- new motifs that have not yet disseminated widely. On the other hand, it is customary in Turkoman work that different kinds of items often use motifs that are specifically associated with the type. Tentbands, for example, use a vocabulary that seems to be unique to them. My estimate is that the mafrash under discussion was woven around the middle of the 19th century, but my confidence in this isn't very high. It would not astonish me to learn that someone has evidence either that it is early or late 19th century.
This bag includes a vocabulary of motifs that is seldom encountered, even the "secondary border" bands occur only when associated with a major border, except in ak-juvals. Its mixed structure of plain flatweave, floating weft flatweave, pile-decorated flatweave, and pile is so different from other Turkoman work that it is tempting to suggest that it is a representive of a previously undiscovered group. It certainly differs more from mainstream Tekke weaving than some Yomud subgroups do from most typical Yomud work or from each other.
It has some elements of being a sampler; only embroidery would be needed for it to be a virtual tour-de-force of Turkoman weaving techniques. It already includes goathair, cotton, silk and wool, the complete repertoire of fibers used by the Turkoman weaver. The cotton whites, the superficial resemblance to designs associated with tentbands, the white and ivory grounds, the very fine weave and the use of silk all suggest that this bag was associated with something special, perhaps a wedding. The loving attention to details of which nobody except the weaver would be aware -- the few lines of floating wefts and the touches of silk in areas so small that they are nearly invisible -- also suggest that it was created by the weaver for some personal ceremonial event and that it had some importance to her.
Does any reader have ideas to contribute to clarifying the origin or significance of pieces of this type? Turkoman bag, complete. Dimensions (face only), 1'0" x 2'1". The hanging ropes are goathair. White bands are flatwoven, with cotton wefts and wool pile decoration. The three red bands are flatwoven wool; the three narrow bands within each are wool pile. The ivory ground bottom panel is of wool pile with silk highlights. Asymmetric knots, 12 per inch horizontal and 32 to 36 per inch vertical (384 to 432 per square inch).
2. Plate 20, in, "Vanishing Jewels", Catalog of an Exhibition of the Collection of Frederica and Marvin Amstey. Rochester Museum and Science Center, Rochester, NY. 1990.
