| Yomud kepse gul torba, l9th century. Overall dimensions, 2'10" x 1'6". Woven from the bottom. Symmetric knots, 11 to 12 per inch vertical, 8 per inch horizontal; approximately 90 per square inch. |
The quickest way to tell a workshop rug, made with a cartoon, from a freehand drawn tribal or village rug is the absolute regularity of the design in the workshop product. Even though it may appear at first glance to be regular in layout, with motifs that seem almost uniform, the village or tribal weaver's product is full of variations and evidences of spontaneity that are important aesthetic components. These features give the rug certain personal characteristics and a charm that just doesn't exist in workshop products.
There's an interesting form of variation in some tribal and village weavings. It is the change in motif type or dimensions or in layout that occasionally occurs partway through the weaving. It is as though the weaver either changed her mind about what she intended to create or realized that she misjudged the space needed for some part of the design, and corrected that error. It is rather uncommon to see examples of this, probably for two reasons. First, pieces with major errors almost surely make up a relatively small fraction of the total that were woven. Second, it seems likely that early collectors tended to bypass pieces with obvious errors, so such weavings were not even preserved in proportion to their production. Unlike stamp collectors, rug collectors don't place a premium on pieces with mistakes in them. Perhaps there is no good reason why they should.
Although not really commonplace, there are enough examples so that errors are not extreme rarities either. I've seen a number of Caucasian and Persian village and tribal weavings in which the weaver abruptly altered the width of the field, creating discontinuity in the line of the borders, in order to accommodate a field design of a size that was evidently underestimated1. Clearly, the continuity of the border was of less importance to the weaver than the intended design of the field. Turkoman pieces with this feature are very uncommon, if they occur at all.
On the other hand, there is a kind of error that is occasionally encountered on Turkoman pieces. It is a change in the dimensions of the guls in order to allow what must be a predetermined number of them fit into a field size that would otherwise be too small to accommodate them.
Figure 1 is a 19th century Yomud kepse gul torba2. Although not very finely woven for a small Yomud piece (around 90 knots per square inch), it has very good colors and excellent, glossy wool. It is obvious that the weaver realized that the scale of the guls was too big when the field was a little more than half completed. Trying to reason out how this might have happened and how she dealt with the problem tells us some things about the aesthetics of 19th century Yomud weavers.
How did she happen to make the guls too big? Two possibilities come to mind. One is that she misjudged how much space she would need for them and couldn't violate the prescribed dimensions of the torba to correct things. This seems very unlikely to be the explanation. The torba is 18 inches deep. Had she completed it by continuing to use the same scale, it would have been only two inches deeper. This is well within the range of depths for Yomud torbas, which can exceed 24 inches.
Furthermore, this piece was woven from the bottom, so it was the second of its pair3. With its mate already finished, she must have known how big it would be. I suspect that she began the project by putting warp threads that were a few inches shorter than what she needed onto the loom, and made the scale change when she discovered her error. Alternatively, she may have made the long flatwoven section that becomes the backs of both members of the pair two inches too short, requiring that the second bagface be shortened so its back would fit it properly. It would be interesting to have the second torba of the pair for comparison. I suspect that it is two inches deeper than this one, with the kepse guls made to the scale with which this one was begun.
Now, realizing that she could not complete the piece according to her original intention, she had some options. I believe that her choices from among them tell us something about her aesthetics. Let's briefly consider the elements of design of nomadic Turkoman bags. They consist of a rectangular field surrounded by one or more borders. The field design is typically a repeat of a gul or of alternating major and minor guls, with partial guls ending at the edge of the field. The borders are surrounded by pile that is usually undecorated, although the bottom -- the elem -- is sometimes decorated with a repeating motif that is most often a stylized or realistic plant.
One way the weaver might have dealt with the error in the piece shown in Figure 1 would have been to just continue weaving the guls to the same scale. This would have allowed her to include the two complete guls, but she could not have finished the field with the partial guls that "disappear beneath the border." We must conclude that it was more important to have those partial guls than to keep the scale consistent. A similar situation occurs in an 18th century Arabachi juval4 in which the weaver evidently realized that she would not have sufficient space in which to complete the design, and reduced the sizes of the major guls, but not the minor ones.
She could have eliminated the space above the upper border, reduced the thickness of the upper border, or done both, and created more space for the field design by doing so. The fact that she didn't do any of these things shows that preserving the proportions of the border to the field and the space above the border were more important than maintaining a consistent morphology of the guls. It is often asserted that the field on a Turkoman piece is conceived as a "window" on an infinite pattern, and that this is an important component of the Turkoman aesthetic. That concept is very clearly demonstrated here, although with an unexpected twist. We tend to think of the window as showing a segment of an infinitely repeating design, but in this piece the design does not really repeat because of the change in the morphology of the guls that resulted from the mid-gul change in scale. The weaver chose to preserve the window illusion -- that is, there are still partial guls disappearing into the borders on all four sides -- even though it necessitated a rather conspicuous change in scale that altered the proportions of guls that were more than half completed. It would appear that the concept of the border enclosing a window on a much larger field was even more important than maintaining the morphology of the gul itself. Many of us would have guessed that the morphology of the guls on a Turkoman piece would be almost inviolate, although it is well known that guls of the same general type will vary somewhat from one weaving to another.
When she opted to change the scale, she made another interesting decision. She could have kept the colors the same but switched from a medium blue to a very dark blue at the point of the scale change, as though to emphasize it. What are we to make of this? Clearly, an error was not something to be concealed. Perhaps the color change, making the error more obvious than it would have otherwise been, intentionally calls the viewer's attention to it. Making the scale adjustment to have the design fall perfectly within the remaining dimensions does, after all, reflect a particular kind of skill. This skill may have been a source of pride. On the other hand, it may have been a requirement of the culture that she acknowledge (confess?) the error by making it conspicuous.
One of the basic paradigms of an experimental approach in the natural sciences is to perturb a system and see how it responds to that perturbation. From that we learn something about the rules governing how it works. An error in weaving is, in a real sense, a perturbation in the normal course of its production, and we should be able to learn something from it about the rules governing how the weaver works. It doesn't necessarily follow, of course, that I have correctly interpreted the outcome of this experiment. Perhaps some readers will have other ideas about the significance of this weaver's response to her error.
NOTES
1. An example of this in a 20th century Caucasian soumak khorjin can be seen in Oriental Rug Review, vol. 13, No. 1, p. 31, 1992.
2. I am grateful to Saul Barodofsky (Sun Bow Trading Company, Charlottesville, Virginia) for allowing me to examine and photograph this piece.
3. Turkoman bags are woven in pairs. The weaver begins with a short length of flatweave, then the first pile face (starting with its top), then a flatweave somewhat more than twice the length of the face, then the second face (starting with its bottom), and finally another short flatweave. The whole thing is cut across the long central flatweave, which becomes the back for each bag. The short flatwoven pieces with which the weaving begins and ends become the turned-under lips atop each bag.
4. Illustrated in an advertisement in Hali, No. 71, p. 23, 1993.
