Posted by Filiberto Boncompagni on 02-11-2006 05:22 AM:

Wise decision

Hi John,

Thank you for your report; I would have reacted earlier if I weren’t busy.
That Bidjov is really nice, bold design, interesting borders. Bennett’s plate 97 shows an even more spaced and bold design. I didn’t scan it, because it is a b&w plate… And because I’m lazy.

I remember that the design of the second Shirvan you posted

was discussed at length a couple of years ago. Unfortunately the discussion degenerated… I think that whoever took the decision of presenting this rug simply as Shirvan made a wise choice because the design has at least two denominations for Azerbaijan: Gymyl or Kabala (Gabala) and one for Daghestan: Djakul
In the discussion many other similar examples of this kind of design from Persia were presented.

I still have the images. Eleven, actually, of which seven Persian - if anybody cares I could post them again here…


Posted by Tim Adam on 02-11-2006 06:56 AM:

Hi Filiberto,

This second Shirvan is the most interesting piece to me, and I'd be very interested in those additional images. It first reminded me of a Konya rug, but on second thoughts this may be too much a leap.

However, the general layout is reminiscent of some classical Turkish rugs I think.


Posted by R. John Howe on 02-11-2006 07:32 AM:

Hi Filiberto -

I expect Jerry Thompson himself might be interested in seeing your images.

I didn't ask him for his specific attribution indicators on this piece, but expect that they include:

white cotton selveges (these are likely restored, since Jerry is very particular about the condition of selveges in his pieces, but would have put back the sort that were there),

the tan wool warps (two tone warps would have been even more classically Shirvan)

and the color palette (especially that visible in the minor borders.


R. John Howe

Posted by Filiberto Boncompagni on 02-11-2006 10:19 AM:

No, John, the Shirvan attribution is perfectly fine for me.

This is an extract of what it was discussed. I started with the following:

The most common label for this type of rug, according to the design, is Gymyl. The term was introduced by Lyatif Kerimov in the book “Azerbeijan Carpets”. He mentioned that this type of carpets is produced in the village of Gymyl, (a.k.a. Gimil, Qýmýl, Gimil', Kymyl) in the district of Kuba. I was told that it’s at 18 km from Kuba.

First scan from E. Gans-Ruedin, in “Le Tapis du Caucase” shows a “ Kouba rug” (pages 266/7):

The text says, “according to the Azerbeijan (sic) Carpet of Kerimov this kind of rug should be a Kouba from the Gymyl region.”

Same book, pages 302/3. This time the title for the rug says “GYMYL”:

But the text says that “Kerimov in the Azerbeijan Carpet, page 67, attributes a similar rug to Kabala (Shirvan).” Note: Kabala is also spelled Qabala and Gabala.

In Bennett and Bassoul’s “Rugs of the Caucasus” pages 254/5 it is shown a KOUBA “GYMYL”:

Here is the commentary written by Mr. Aziz Bassoul:

“The ornamental composition gives to this late 19th- early 20th century rug its name of "Gymyl". The midnight-blue central field shows an elongated honcha medallion and four corner designs of identical pattern. According to Kerimov, op. cit., N. 32, honcha means a "present served on a tray". The central red medallion, referred to as turanj (a kind of citrus fruit), contains two concentric medallions, the outer of which is in blue and the innermost in red.
The Gymyl composition could have originated in Tabriz as early as the 16th and 17th centuries, but this remains doubtful. The relationship with the Melayer rugs of Iran is, however, unquestionable”

D. Chirkov’s book “Daghestan decorative art” (Moscow 1971) shows another interesting item

which it is said to be a “Djakul napped carpet. Wool weaving. 1870-1890. Made by Tatima.”
Here is the related commentary:

“Djakul (doll), one other specimen of the napped carpet variety, is no less popular in Tabasaran. The ornamental motif in the carpets's centre recalls dimly a human figure and derives perhaps from the anthropomorphic variety of art objects represented in good many decorative art forms throughout Daghestan. In turn, the central pattern is set against the background of a light chequered ornamental motif which, as the legend goes, represents a sheep-skin. The carpet is noted for the overall symmetry in relation to the vertical and horizontal axes. The intense colour of the central figures penetrates far into the meshy background by virtue of the edges being deeply serrated. Deep intercurrent of blue permits to take advantage of the mutual penetration of colours, not withstanding their contrasting shades. In most Tabasaran carpets the colouring consists of dark-red and blue colours in ornamentation set against the light-golden background. Black colour shows up if only in the outlines. Djakul enters into the group of carpets from the villages of the Rubas-Chai valley remarkable for their brightness and festal iridescence of colour range.”

Note: the Tabasaran are one of the indigenous people of Caucasus. Their presence is documented since the 5th cen. AD.

Vugar Dadashov, an Azerbaijani carpet dealer said that

The real name of this design is "Gadim Gabala" which means "ancient Gabala". Historically this type of carpets firstly appeared in Gabala Region (one of the northern regions of Azerbaijan) and it spread from here to the east - Guba region later in XVIII century. (including Gymyl village).
The same design was also called "Mammad Aghaly…
The medallion of the central field is another form of ‘lachak-turunj’, which was used in the decoration of books and architectural buildings and engravings of gravestones since 9th century. I strongly believe that lachak turunj firstly appeared in the capital of ancient Azerbaijan – city of Tabriz. But the formation of today’s Gymyl carpets was realized in Gabala. From this point of view, I prefer to name this design as Gadim Gabala (Ancient Gabala), as old carpet makers named it.

If we carefully compare both Gabala and Gymyl carpets, we will witness that exactly Gabala carpets have got simpler design,

(above image provided by Mr. Dadashov) while Gymyl carpets have got more complicated but the same design. And this is additional proof for authenticity of this idea.
The in-question carpet type was started to be produced in Gymyl village in 18th century. At the same time, Azerbaijan carpets was started to appear in western markets. So production of these carpets in Gymyl village and highly demand for Caucasian Azerbaijan carpets in western markets came across at the same time. Therefore, Gymyl type of carpets became famous.

Additionally, I want to mention that in his later book “Azerbaijan Carpets”, famous Azerbaijan scholar Latif Kerimov pointed out that Gymyl type of carpets has descended from Gabala.
He added that he believed that the Daghestan variant was copied from the Azerbaijani one.

About the Persian origin of the design, P.R.J Ford, in his “Oriental Carpet design” attributes these two North-West Persian rugs to the Kurds… (see pages 242-244) implying the Shasavans, the “north-west Persian dragon designs of the sixteenth or seventeenth centuries” and “early Heriz carpets”.

Wendel Swan sent an image

with the following comment: By its caption, you will see that I first thought it was Afshar, but some knowledgeable Persian dealers have said it is more like a kind of
Bakhshaish from Persian Azerbaijan. It is not Caucasian, having cotton warps and wefts.

The design is shared by a variety Persian medallion rugs, some of which are called Feraghan, while others have different attributions. In any event, these Gymyl rugs, as well as mine, simply focus in on the center of the medallion and cut off the anchors.

I first thought my rug was an Afshar because the Afshar use this "close-up" approach to design quite frequently. Also, the handle is a bit stiff, as also is found in Afshar village production as are these rather distinctive "spots." But I don't think it is Afshar any longer.

Vincent Keers contributed with an image of a Malayer

And I ended the sequence of images with this photo of a modern Shirvan kilim:

Best regards,


Posted by Tim Adam on 02-12-2006 09:49 AM:

Hi Filiberto,

If we carefully compare both Gabala and Gymyl carpets, we will witness that exactly Gabala carpets have got simpler design, while Gymyl carpets have got more complicated but the same design. And this is additional proof for authenticity of this idea (that the design originated in Gabala rather than Gymyl).

I don't really follow this argument, do you? Also, if we wanted to make a point about design origin, would we not need to see some older pieces? All of the rugs in your post seem to be late genre pieces.

The following rug may seem a big leap, but it has the same design layout and is perhaps 200 years older than any of the rugs presented so far.


Posted by Filiberto Boncompagni on 02-15-2006 03:08 AM:

Hi Tim,

What is this rug exactly? The cloudband border made me have a look at Stone’s “Guide”. Interestingly, in the comment to image A-16 showing the transformations of that border he notes that “this evolution tends toward elaboration rather than simplification, the usual tendency”.
The same could apply to Vugar’s argument.


Posted by Tim Adam on 02-15-2006 07:28 AM:

Hi Filiberto,

The rug in my previous post is a 16-17th century village rug with Medallion design from west Turkey (285x173cm), according to Spuhler. It is from his book "Oriental Carpets" Plate 23.

The Cloudband design looked curious to me too. Maybe the weaver simply had a hard time to weave it vertically?

Coming back to the Shirvan rug that stated this thread, the design may be a variant of this, quite common, Medallion design, rather than a new invention.

Do you think it is very old? I find the main field very well executed, but the border is extremely degenerate.


Posted by R. John Howe on 02-15-2006 07:54 AM:

Hi Tim -

Jerry Thompson estimates this piece as "19th century."

About your "degenerate" comment: I know a rug analyst here who claims that we should not use such terms as "degenerate," although there are a lot of such references in rug literature.

Her argument is two-fold.

First, she says, we do not usually know enough about a given feature to be able to identify it's origin accurately. "Degenerate" is comparative and requires knowledge of an "original" which she doubts we usually have.

Secondly, as your conversation with Filiberto has also noted, some times "development" (another term she doubts we have earned the right to use) seems to occur by "elaboration" rather than the "simplification" we usually expect. More, some simplifications are seen by some as aesthetic improvements (some old Kurdish renditions of even older Persian designs are often cited as examples) and that the term "degenerate" seems unavoidably to refer to loss. The piece below (that we saw recently in the "color" salon) might be seen as an example aesthetic improvement resulting in part from a degree of conventionalization.

She advises use of descriptions that indicate variation without such unearned evaluations.

She's not having much impact on the way people talk about rugs and rug design, but it's an interesting argument.

To go back to your evaluation of this particular border, it does seem to be a simplification of some more articulated design, but, for me, it's degree of abstraction "fits" the field, the drawing of the field that you admire. I'm not arguing for the correctness of my view here, only that I think it's possible to see and to evaluate it differently than you have without committing demonstrable aesthetic error.


R. John Howe

Posted by Filiberto Boncompagni on 02-15-2006 08:20 AM:

Hi Tim,

The Cloudband design looked curious to me too. Maybe the weaver simply had a hard time to weave it vertically?

No, I think she was mocking us, weaving in the same rug a correct version of the cloudband border and its “development”.

As you ask me, my un-expert opinion for the age of Jerry’s rug is end of 19th.
In any case and for my taste, it’s the best of the Caucasians in this thread, closely followed by Vugar’s one.
the design may be a variant of this, quite common, Medallion design, rather than a new invention.

Yes, I agree.


Posted by Tim Adam on 02-15-2006 09:47 PM:

Hi John,

I agree. Using the term 'degenerate' can be unfortunate because it has a negative connotation. For example, if indeed the Bidjov design can be traced back to the Caucasian palmette (shield) rugs, then one should really not talk about a degeneration of the design.

In the Shirvan rug I don't mind talking about a degenerate border design, however, cause its quite clear what the original design was (to address your rug analyst's point) and I really don't like what the weaver made of it.