Home Page Discussion

Salon du Tapis d'Orient

The Salon du Tapis d'Orient is a moderated discussion group in the manner of the 19th century salon devoted to oriental rugs and textiles and all aspects of their appreciation. Please include your full name and e-mail address in your posting.

Turkmen Tree-of-Life Paneled Mafrash

by Steve Price

Turkmen mafrash (or kap) include a small group of Turkmen bags about which little is written, although there has been considerable collector interest in them for some time. This is the group with a layout consisting of ivory panels, each containing what is generally interpreted to be a tree of life motif. Published opinion is that this layout and design originated among the Tekke, and diffused to other Turkmen groups, although most examples are either Tekke or Yomud. Here is a Tekke specimen (from Pinner, The Rickmers Collection, p. 53). It was collected between 1902 and1906, and is attributed to the first half of the 19th century

Those of all tribal groups show remarkable consistency in design and layout, even with regard to some details that we might think of as trivial. Let me begin by focusing on some of the features that this one shows.

First, the obvious. The most prominent feature is the ivory ground panels, each containing what I suppose is a stylized tree of life. This much of the description is common to every specimen (I recognize that this statement is tautological - we would not include any specimen in the group unless it fit that description). Let's push on. Notice the outer border. The ivory ground color and bird or flower motif shown here is also common to every specimen of which I'm aware. Next, look at the borders surrounding each panel, decorated with stylized flowers. Again, every specimen I've encountered has such borders, although in most of them the orientation of the motif is reversed relative to this one. That is, the devices usually look more obviously like flowers on a stem, with leaves on the stem and at the bottom. The only variation is in the number of pairs of leaves below the flower head (one or two pairs, with rare exceptions). Now take a look at the ashik - the large horizontal element within the field. There is a device in its center, always a rosette in Tekke pieces, but varying in the work of other tribes.

I will return to these matters. But first, have a look at a Yomud specimen.

Notice the similarity to the Tekke example. In formal terms - that is, the description that might appear in a major auction catalog - they are identical except for the number of panels. But other differences are obvious. The Tekke is much "busier". It has more fine detail, more fillers. The Yomud is much less crowded, much simpler. This is a consistent difference between Tekke and Yomud examples. In fact, it is a fairly consistent difference between most formally similar textiles of those two tribes. The rosette at the "hub" of the Tekke occurs in every Tekke example I've seen. Other devices are common at this location in Yomud pieces. In this one, it's a stylized bird or flower, similar to the motif that appears in the main border.

This is probably a good time to introduce a little database of the characteristics of Turkmen "tree of life paneled mafrash" that I assembled. The method was simple. I scoured every published source that I could get my hands on conveniently, looking for photos of these textiles. The ones I found, plus one that belongs to me (the Yomud piece above), comprise the database. The total number of specimens in it is 36. For each one, I recorded the tribal attribution given to it in the source (see note 1), plus the following characteristics: number of panels, number of pairs of leaves at the base of the flowers in the vertical borders, number of pairs of leaves at the base of the flowers in the horizontal borders, device at the hub of the ashik. I will be happy to provide the references to the sources to anyone who would like them. For most readers, I assume that this is not interesting information.

The table below contains the numerical results. I will briefly summarize them here. First, there are consistent differences between Tekke and Yomud pieces in each characteristic. Most Tekke have 3 panels, 2 pairs of leaves on the plants in the horizontal borders, and always have a rosette at the hub of the ashik. The Yomud usually have 2 panels, a single pair of leaves on the plants in the horizontal borders, and an approximately random distribution of hubs of the ashik motif, the hub types consisting of three different motifs or no motif at all. Pieces from the other three tribal groups (Saryk, Ersari and Chodor) did not show any obvious pattern of being more similar to Tekke or to Yomud, although it must be emphasized that there was only one specimen from each of the three.

What do we make of this? First, there are obvious, consistent differences between the Yomud and Tekke pieces. That is, despite the clear similarity between them, certain characteristics predominate in one or the other. The migration of the design was clearly accompanied or followed by tribe-specific modifications. Even the most prominent feature, the number of panels, differs.

Second, my initial hope that it would be possible to use my data to trace the migration of the design from one tribe to another met with disappointment. I had hoped that examples from one of the others would consistently show Tekke or Yomud characteristics, from which we might conclude that its ancestor was the one it resembled. Surprisingly, the sole Saryk and Ersari examples had layouts of four panels, a feature occurring in only about 10% of the Tekke and none of the Yomud specimens in my database. I say "surprisingly" because we might expect the number of panels to be so important a feature that it would not change when the design diffuses.

Now I ask you to consider the following as some of the discussion points:

1. What, if anything, can we learn about design migration from this fairly well defined design's variations among the Turkmen tribal groups?

2. Are there other categories of Turkmen (or, for that matter, non-Turkmen) weavings whose study might reasonably be expected to shed light on mechanisms and dynamics of migration of design?

3. What sense can we make of the fact that certain characteristics are invariant and others are highly variable when a category of textile migrates from one tribe to another? Do the invariant ones have any special significance?

4. Is there a historical explanation for the fact that the design is almost restricted to Tekke and Yomud weavings? It isn't surprising that there are no Salor and very few Saryk and Chodor examples in view of the fairly small numbers of pieces made by those tribes that are extant. But there are probably more Ersari textiles around than weavings from any other Turkmen group. Why so few paneled tree of life mafrash?

Here are the numerical results. There are 36 specimens: 24 Tekke, 9 Yomud, 1 Saryk, 1 Ersari, 1 Chodor.

1. Number of panels:

a. Tekke: Of 24 pieces, there were 6 (25%) with two panels each, 15 (62%) with three panels, 2 with four panels and 1 with six panels.

b. Yomud: Of 9 pieces, 6 (67%) had two panels each, 3 (33%) had three panels.

c. Others: The Saryk piece had four panels, as did the Ersari; the Chodor had two panels.

2. Pairs of leaves on the plants in the vertical borders:

a. Tekke: Of 24 pieces, 5 (21%) had a single pair of leaves, 19 (79%) had two pairs of leaves.

b. Yomud: Of 9 pieces, 7 (78%) had a single pair of leaves, 2 (22%) had two pairs of leaves.

c. Others: The Saryk and Ersari had single pairs of leaves; the Chodor had a two pairs.

3. Pairs of leaves on the plants in the horizontal borders:

a. Tekke: Of 24 pieces, 1 (4%) had no leaves, 22 (92%) had a single pair of leaves, 1 (4%) had two pairs of leaves.

b. Yomud: Of 9 pieces, 8 (89%) had a single pair of leaves, 1 (11%) had two pairs of leaves.

c. Others: The Saryk had 1 pair of leaves, the Ersari had none, the Chodor had 2.

4. Device at hub of ashik:

a. Tekke: All 24 pieces (100%) had a rosette at the hub of the ashik.

b. Yomud: Of 9 pieces, 2 had a rosette, 3 had a chemche device, 1 had a stylized device that might be read as a bird or as a flower; 3 had no device in that position at all.

c. Others: The Saryk and Ersari each had a rosette the ashik hub; the Chodor had the "bird or flower" device.


1. There is one exception. What is clearly a Yomud specimen in Amos Bateman Thacher's book is attributed by Thacher to the Tekke. I list it among the Yomud.