Click on image to view in larger size
Click here to see image in extra large size.
This is a curious rug. The tree design looks borrowed from Baluch tree designs (see this one, for example) while the heavily depressed structure puts it in the Bijar area, ruling out a Quchan Kurd /Kordi attribution. The colourful palette with its clear yellow rose and medium green is also quite different from the Baluch origins of the design. Along the stem of the tree, quartered gul-like shapes appear which are reminiscent of Turkmen rugs, even though the internal detail of the guls has been lost (compare a simplified version of borrowed Turkmen guls with less loss of detail in this Baluch type rug attributed by Craycraft to the Jamshidi tribe). The quartered design of the 'guls' echoes that of the quartered serrated leaves adorning the branches. Interestingly, the horizontally extending leaves of the typical quatrefoil design look atrophied; in fact, those facing the stem are ignored while those facing outwards seem to have changed into an asymmetrical split leaf design, with a small blocky bud in the centre, if I read this correctly.
The tree scheme itself is used in a cavalier fashion; beyond the point where the horizontal branches end, another branch stud appears out of nowhere, with another pair of leaves attached to it. The figure below comparing this rug with a similar, probably older rug, seems to explain this strange feature. At three levels, two half medallions in a shape similar to the central 'guls' protrude inwards symmetrically from below the border, indicating an endless repeat that is alien to the self-contained tree structure. The internal treatment of these shapes is different, bringing it closer to the 'samovar' shape seen in many Persian main borders, a shape which is derived from a large palmette blossom in profile. Which raises the question whether the similarly shaped 'guls' (also a term for flower, incidentally) should also be interpreted as palmettes, not as derivatives of the Turkmen tribal heraldic emblems seen on rugs from the Tekke, Salor, Ersari, etc.
Most attractive perhaps are the small human and animal shapes that give the rug its tribal note. Especially nice the jolly scene of a human with outstreched arms touching a double headed animal which can be seen near the vertical centre close to the border. Below this human-animal couple, we see what may be interpreted as a snake—I don't know whether there is any Persian or Kurdish folklore or imagery from other sources (paintings, reliefs) that would support such interpretation.
Another attractive feature is the abrash in the camel field, from a ligher shade in the top half towards a medium brown in the lower half. (I don't think it is actual camel wool, at least the pile feels like the rest of the wool). On closer inspection, many smaller elements fill the gaps between the branches of the tree: quartered stars or fragments thereof; small blossoms; more florid / pointillistic blossoms reminding me of the complex botehs seen in some Afshar rugs; heraldic-looking shield or animal-head forms; diamonds on top of a flower or butterfly(?) shape; more snakes, or more prosaically, meandering lines. How many of these forms actually undertake to represent something physical is not easy to determine.
(1-A) A similar Bijar with tree design made available by a collector to another collector / rug friend who sent it to me. It is interesting to see that this variant, by aligning the central hooked diamond medallions and those of which just the half is visible, clearly indicates endless repeat design aspirations. This also explains the branch stubs which sem to belong to the adjcent trees to the left and right of the visible field. This has been lost in our rug, which could serve as rationale that it is a later design derived from this rug's repeat pattern.
(Click here or on image to enlarge comparable bijar (jpeg opens in a browser window))
(1-B) This bijar—guls instead of hooked diamonds, but the same overall design. The other example shows more wear but this does not necessaritly mean it is older (though I guess it is, looking at the design and what can be constructed as history of pattern degradation). The weaver, not longer knowing where the branch stugs belong, has simply flipped them to relate them to the central stem.
(1-C) Another rug with a 'tree of life' pattern (courtesy of Joseph Beck). The medallion seems a mix of a gul type and a hooked diamond form. Note the powerful border with large serrated botehs that, in keeping with the 'tree of life', remind of a Baluch or Caucasian variant of the boteh. The field is much narrower than in the other two rugs, and there is no hint of a repeat design.
(Click here or on image to enlarge NW Persian rug (jpeg opens in a browser window)).
(1-D) Close-up of the back side of the same rug, showing what loks like a typical NW Persian / Kurdish non-depressed weave.
(Click here or on image to enlarge NW Persian rug back view (jpeg opens in a browser window))
Compared to the interesting field, the border system is unremarkable. Instead of a main border there are three simple meander borders, a design often found in in Kurdish rugs. The dark brown background of the inner of these three borders adds a nice touch by framing the camel field and thus visually separating it from the 'noise' of the other outer borders.
The rug measures 6ft.8in. x 3ft.11in. (203 x 120 cm). It is a typical deeply depressed Bijar weave, though not fully depressed as in more modern production; I guess the angle is about 80°. Wool warps, heavy first weft followed by thin sinous weft. This makes the structure more rigid and a lot heavier than that of the non-depressed rugs I usually encounter. Given that most of the rug has also full long pile, this makes for a rug that is very heavy for its size (about 9 kg). The weave is coarse, should be symmetrical knots as usual for Bijar though I cannot verify that. The horizontal knot count is h.25/10, the vertical knot count is v.26/10, which means roughly 650 knots/dm2 (or, converted to to inches, approx. h.6.5,v.6.5 = approx. 42 kpsi).
The pile is mostly full and excellent, very dense, nice shiny wool (see images). Thinner near the borders, which show some foundation. There is a bit of bad or faded repiling near the left border and another area in the field ground near the top left corner. There are some areas of abraision visible at the back of the rug; so far the structure holds together nicely and the pile is firm, because it is very densely woven, but I would not recommend heavy use. Selvages are single cord wrapped in red and rose wool, some minor fraying. The twining at the top end right corner is unravelling a bit (see image).
A nice and very colourful palette of red, yellow, mid blue, mid green, rose and dark brown played out on the abrashed camel ground. To my eye all dyes seem natural, no nasty shades, no tip fading or difference in colour ot the back side.