Posted by R. John Howe on 09-15-2005 03:59 PM:

Yellow-ground Caucasian Wreck

Dear folks -

This is the second rug I bought in the last two weeks, this time in an antique coop in the Shenandoah valley a little southwest of where Steve Price lives.

As you can see this is a pretty tired yellow-ground Caucasian with a lattice field.

It has an attractive white-ground border

that may have some "Kufesque" traces in its intervening brackets.

The field devices are hexagonal cartouches arranged in lattices with barber-pole striping and have eight-pointed stars inside.

The drawing is fairly crude but some touches of detail, including outlining are retained.

The condition of this piece is pretty poor. The ends are missing, it has bare spots, some damage even of structure and as the image below shows, it has been subjected to some bad repair efforts.

Even a casual examination reveals areas of likely "Kashmiring" repair (look for areas where the columning of knots is lost).

The piece is quite dirty and I expect the color might improve a bit with a good washing.

It has white wool warps and ivory or tan wefts, the latter varying one to three picks between rows of knots.

The side selveges are damaged but appear originally to have been of three cords wrapped in brown wool.

I do not follow Caucasian weaving closely enough to give grounds for an attribution.

I do know that yellow grounds are often seen to be attractive and that many collectors cannot get enough of "Coke-bottle" green.

Filiberto suggested,as we were putting these images up,that the field design and color pallette are similar to Plate 155 in Ian Bennett's Caucasian book.

As usual, I invite frank comments on this humble piece.


R. John Howe

Posted by Amir Aharon on 09-16-2005 02:05 AM:

"A HUMBLE piece"

Hi John, everyone,

This is far from being a humble piece; maybe YOU are!

A very similar long rug was sold recently at Sotheby's New York
(June 3, 2005--lot # 19).

With all due respect to the Sotheby's lot, your rug has a much
more live color palette. Particularly the white ground main border
which you don't see in the Sotheby's nor the Ian Bennett's
mentioned in your post.

Despite the condition (outer guard missing on both ends and
bad reweaves), I think you were lucky Steve didn't visit Shanandoah valley that same day.

I have seen a Qashqai rug and a bagface with this lattice 8 pointed stars design, both on a madder background. There
are Turkmen (Ersari) pieces with a similar field too.

I suppose your rug is from the Gendje area (belonging to the
Kazak group weave structure). They usually have multicolored
sides. Do you see any remnants of different colors on the side

Anyway, I love yellow ground rugs in general, in fact I had a
very early turmeric ground Karapinar long rug with slanting 8-
pointed stars all over (without the lozenges) which I may post
in the future. As you see I maybe somewhat biased, yet I am sure you will be getting enthusiastic comments from dormant



Posted by Filiberto Boncompagni on 09-16-2005 03:34 AM:

Dear Amir,

I’m afraid the only enthusiasts about Caucasian stuff here on Turkotek (besides some occasional Italian visitors) are you and yours truly.

You are right, John’s one is a good find.
It seems that he took the pictures later in the day – or early in the morning. That gives a general yellowish tone. This is how the rug should look, boosting the Kelvin temperature at 7075.

Now, the colors seem very good.
This rug is very close to Bennett’s 155. The differences are in the border: #155 has a red ground main border and the secondary borders are both medachyl (blue on brown instead of red on brown, like in John’s rug inner border).
The outer border in John’s has “florets” instead.

Huh, I think that yellow ground Caucasians are quite sought after by collectors for their rarity… John, I think you should restore it.

Amir, do you still have the image of the one sold by Sotheby’s? It could be interesting to compare it.



Posted by Chuck Wagner on 09-16-2005 06:45 AM:

You say tomato...

John, Filiberto,

There are Caucasian enthusiasts, and then there are Caucasian enthusiasts who have been able to find affordable Caucasian rugs that fit their taste.... I'm still waiting. But then, I'm only recently back in the U.S., so the hunt has just begun.

It's the fortunate few who come across pieces with that pleasant yellow; I think John has made a great find. Scale and nature of field execution create a nicely balanced overall appearance, even with a coarse angular design. And, the white ground in the border complements the yellow quite nicely.

Now he has a quandry: wall or floor ??


Chuck Wagner

Posted by Filiberto Boncompagni on 09-16-2005 09:07 AM:

Well, Chuck...

Welcome to the club!

Posted by Sue Zimmerman on 09-16-2005 09:15 AM:

Hi John,

This rug is a harmonious, festive, and delightful work of art. It's simply bursting with radiating genuine innocence. Thank you for sharing this little gem which is now firmly in my mental collection. I have it flagged for heavy recall for the long cold Midwestern US winter ahead. Inspiring. Sue

Posted by David R.E. Hunt on 09-16-2005 09:21 AM:

Southwestern Caucasian = Anatolia?

Hi John

Nice find! But to be honest, my first impression, judging from the yellow and green colors, rustic drawing, and meandering selvage was that of a rural Turkish piece or Yoruk. The funky drawing of some of the design elements reminds of Plate 23, a Konya Yastik from "Pacific Collections", but then I am not exactly the cutting edge of caucasian carpet ascription (nor Turkish for that matter ) You must get us a better photo once you've had it cleaned. Out of curosity, what is the knot count?


Posted by R. John Howe on 09-16-2005 11:47 AM:

Dear folks -

Here are some images of the piece in Bennett's Caucasian book, plate 155 that Filiberto pointed out to me.

The ground color and the field design are quite similar.

Here's a little closer look. I think there is more detail in the drawing in the Bennett piece.

That's true even in the "kufesque" border.

This piece has also functioned to remind me how close Genje, Karabagh, and Kazak are, geographically, and how closely these rugs resemble one another.

There is a Kazak in Bennett (plate 42) the palette of which is surprising close to my piece here. In that piece the stars are in the border.

Some of the writers on Caucasians who reside in that part of the world use the category Kazak-Genje. Eiland indicates that the only real reason for retaining this distinction is that Wright and Wertime have indicated that there were 33,000 weavers in Genje at the turn of the 20th century. That's a lot of rugs.

All this will be "old hat" to those who follow Caucasians more closely but I had misplaced at least some of it.


R. John Howe

Posted by Amir_Aharon on 09-16-2005 02:38 PM:

Hello John, Filiberto, Chuck.........

Here is the Sotheby's piece. The caption says 'A Gendge long rug circa
1890--reduced in length, oxidized charcoals,
Minor repiling, small reweaves---size: 287 by 127 cms.'

It looks more like the one in Ian Bennetts' Nagel catalogue especially
the wide guard and the two medachyl
narrow borders. I think the Rosettes in John's rug outer guards are
attractive, so is the white wider border.

The red and white barber pole narrow guards appear in all the three
rugs , a
feature typical of many Gendje's I have

I agree with Filiberto; it's worth to professionally restore this piece
THEN hang it on the wall. Meanwhile it
can stay on the ground; no quandary

The second rug is an early 19th century Turkish Karapinar. I have
posted it
just for fun, since it also has
the 8-pointed stars and the rare turmeric yellow ground.

As for the few Caucasus rugs enthusiasts Turkotek members------A
thousand....always begins with one....(Chuck)


Posted by Tim Adam on 09-16-2005 04:25 PM:

Hi guys,

Unfortunately, I don't share your enthusiasm for John's new rug. It displays design deterioration on every level. The best piece is clearly the one in Bennett's book, followed by Sotheby's rug. Even the white border is not a happy choice. It is nice when viewed in isolation, but it gives John's rug an unbalanced appearance. Overall the rug is not ugly, but I don't think it is in any way remarkable and would warrant restoration. I'd expect that the cost of restoration would far exceed what John paid for it - hopefullly .



Posted by David R.E. Hunt on 09-16-2005 06:59 PM:

Precursor vs Deterioration

Hi Tim

Always the devil's advocate .

The white ground border does seem rather out of place, but not because of some detrimental effect upon the overall design, in my opinion. While there are numerous similarities between John's rug and the posted images, I think the answer as to the question as to provenance (at least for myself) might be answered by considering the relationship which seem to exist between the two designs. In essence, is the simplicity or "deterioration" which you cite, evidence of a rural interpretation of the more complex design of the Bennett piece, or of an early, precourser type relationship? It seems that early Caucasian rugs are greatly simplified in comparison to their latter progeny. Let's take a second look.

Here's the Bennett piece, and of course Johns.

The ever present questions of color representation not withstanding, these colors strike as quite different. Also, the border sequence and composition, preserved in the two caucasian pieces, differs in John's, yet mirrors (or at least is more like)that of the Turkish. My guess is that this carpet, owning to the variation found in the colors, use of color and it's lack of conformity to the tenants of border structure as seen in the caucasoid, says to me rural interpretation woven in a different weaving area or by a different weaving group. The others seem much darker too.

I think it's delightful, but don't believe it belongs with the other rugs. As always (at least with myself ) colors seem rather specialized a subject requiring more familiarity.


Posted by R. John Howe on 09-16-2005 08:00 PM:

Dear folks -

I'm glad that Tim Adams has dissented from the effusive praise of my piece.

I think he is clearly right that the Bennett and Sotheby's pieces are drawn with a steadier hand and show less conventionalization of design than does my rug. Even the diagonal use of color is more definite in the Bennett and Sotheby pieces.

Since this is the case, my rug is a candidate for asking whether we are liking it a bit for unsound reasons.

There is a tendency among folks attracted to tribal weavings to admire variation and to celebrate it as a sign of authenticity. This may be justified sometimes (no one celebrates the mechanical perfection of a high quality contemporary Indo-Bijar). But we may sometimes be admiring what might in fact but an instance of poor weaving. I don't think we should decide that quickly but I think it is something that those of us valuing "robust" tribal drawing are exposed to.

This problem seemed potentially widespread enough and serious enough to me that early in the days of Turkotek I designed and conducted a salon exporing the contours of this problem area. The salon was entitled "The Oops Thesis." Here is the link:

As is usual, we didn't settle much of anything but I think we did sensitize ourselves to a potential problem.

So while we might not ultimately accept Tim's assessment here (we might too), his statement alerts us to consider, self-consciously, whether we might be liking this piece for unsound reasons.

P.S.: FYI, no one would think I have paid too much for this piece. It was priced quite low and I bargained successfully for a substantially lower price than that.


R. John Howe

Posted by Filiberto Boncompagni on 09-17-2005 07:05 AM:

Dear John and all,

Unfortunately in the last few days I’m a bit busy and distracted from the discussion here.
For the moment I have to limit myself to an observation and a couple of questions.

Observation: Bennett’s #155 seems having a more sophisticated and mature drawing and execution compared to yours.
But… did you notice that, while in yours the lattice is organized in a pretty logical and obvious way (that is, starting from the top we see a row of four whole stars, then a row with two whole stars preceded and followed by an half star and so on…) on Bennett #155 the inner “window” framing the lattice is too narrow to show a row of four whole stars. It “truncates” irregularly the lattice on the left.
So the composition shows (from the top left): first row has ¾ of a star followed by 3 whole stars, the second has a fragment of a lattice, three stars and half star, and so on…
The question is:
Do you think the weaver did this “asymmetric framing” on purpose or it was a mistake?
This irregularity makes it more or less pleasing compared to yours?



Posted by Sue Zimmerman on 09-17-2005 08:20 AM:

Hi John,

I hope my perception is just too active about something that is concerning me. I hope you are not thinking that you can improve this rug's color by giving it a good washing in your bathtub. Please say it isn't so. Sue

Posted by R. John Howe on 09-17-2005 08:37 AM:

Hi Sue -

It is quite dirty and does need a thorough, gentle washing.

In my experience there is no way to predict in advance whether the colors will "come up" a bit, as the hopeful among us often say.

I have had this happen on occasion but it seems entirely experimental probably because I have washed only perhaps 50 rugs of various sizes in the last 10 years.

What are your thoughts and what is your experience in this regard?


R. John Howe

Posted by Steve Price on 09-17-2005 09:43 AM:

Originally posted by Sue Zimmerman
Hi John,

I hope my perception is just too active about something that is concerning me. I hope you are not thinking that you can improve this rug's color by giving it a good washing in your bathtub. Please say it isn't so. Sue

Hi Sue

I've only seen the color on my computer monitor. John has the rug in his hands, and says that it's dirty. A good washing will improve the colors of a dirty rug, and a bathtub is a handy place to wash it.

What is it that you are concerned about?


Steve Price

Posted by R. John Howe on 09-17-2005 10:51 AM:

Dear folks -

I think I may see what Sue Zimmerman is concerned about.

Sue tends sometimes to speak elliptically and apocalyptically, but she is not asleep. She often sees things.

I think her concern here is that she sees an area in the approximate center of this piece in which there is sign that the green has run into the sorrounding areas of yellow. This could signal that the green in this piece is unstable.

I don't know whether this piece has been washed before (I would guess several times) but if so this transfer does not appear to occur in any other areas (and there is lots of opportunity for it.

But Sue's concern is legitimate and I will test this area modestly before I entirely immerse this piece in cold water.

I have a similar concern about the areas that have been repaired. Probably chrome dyes were used there but who knows?

So these are two areas I need to test before I dunk this piece.


R. John Howe

Posted by Amir Aharon on 09-17-2005 12:18 PM:

To wash or not to wash...

Hi John,

If you are worried about color run, pour in a lot of vinegar (5%
concentrate). That should do it. Over here, for every glass of
textile powder shampoo you add 5 glasses of vinegar to the
water. Then you dunk the rug all night long.

After all this, if the color does run, then you better listen to Sue
when you clean your next rug.




Posted by R. John Howe on 09-17-2005 05:05 PM:

Dear folks -

I am washing this rug today and perhaps should document my experience for you.

Here's what I'm doing.

First, I thoroughly vacuumed the piece back first then front, making sure to vacuum the surface I was laying the rug on after each vacuuming.

Next, I prepared a salad bowl size portion of the detergent I planned to use, and collected some absorbent cloths. I then put a little of the mixture in the suspected areas and watched for any sign of running with my sponging cloths at alert. (To tell the truth I couldn't tell conclusively since the material darkens appreciably, especially on the back, when one moistens them. But I didn't see anything too alarming.) I have heard that professionals often work with eye droppers and strong vacuum that will take up liquid rapidly in such situations, but I have not been that cautious and have been more optimistic.

Next I put the piece in the bath tub and ran a full tub of coldish (actually closer to lukewarm) water. At first I did think I saw some signs of green coloration in the water (notice no soap yet in the tub washing) but the water was more a dirty brown when the tub filled.

Now I took a vegtable brush and worked the entire surface of the rug front and back. (This piece is just about the widest width that one could conveniently wash in a typical tub. It is about four widths of the tub long and so I have to turn in in folds about four times as I wash it.)

Once I had done this first washing, I drained out the tub sometimes holding up the rug to the shower to get things on its surface to run off more completely.

Now I piled the rug in a relatively small pile and turned on a luke-warm shower and let it rinse under the shower for over two hours, turning it occasionally to expose new surfaces. The water ran clean eventually.

Now I filled the tub again and put in enough of the detergent (I use Orvis) to get some suds in the water. Then I again worked the front and the back of the rug with the vegetable brush. The water was colored again but now more darkish gray suggesting dirt. I drained off this second tub of suds and again held the piece up to the shower to get as much of the soap off the pile as I could readily.

Then I piled the rug a second time and it is has been rinsing under the shower for perhaps 30 minutes as I write. In about an hour and a half, I will make another tub of Orvis suds and repeat the suds wash.

Then I will do an extended rinse. I have sometimes rinsed a rug up to eight hours (Orvis is reputed to stick and to act like a clear dye much like some hair conditioners. The Smithsonian folks have indicated that even if some is retained after washing that it is so PH neutral that it does not affect the wool adversely. Nevertheless I try seriously to rinse thoroughly.) Now I realize that not everyone could rinse in this way (Filiberto would likely be arrested there in Jordan.) Nor to I indulge when there are local water shortages. But I don't have a lawn to water. :-)

It may be too soon to celebrate, but I see no signs of further color runs in this rug and think that the color transfer that is there is likely the result of something having been spilled on it.

What my washing did uncover were several areas of white that I could not see previously. As if paint had be dripped in a minor way. Very minor, only two or three knots in about three different areas and I think I can get them off even if I have to clip the pile a little in these areas.


R. John Howe

Posted by R. John Howe on 09-17-2005 08:23 PM:

Dear folks -

As a kind of footnote to my description of how I wash and rinse a rug myself, it might be useful to some to describe roughly the specifics and results of an alternative method, namely vacuum extraction washing, that is being used in some instances nowadays.

One reason to describe this alternative method is to indicate how feeble our more usual efforts to wash antique pieces likely are.

One experienced collector here in DC recently told me that if one takes a piece that has been hand washed in more conventional terms (approximately what I have described here, although some continue to do detergent iterations of washing until the water after one seems rather clear) and subjects it to this vacuum extraction method, the results are dramatic.

Here, roughly is how the vacuum extraction proceeds as I understand it. One washes a rug in the usual way, but after the last conventional rinse in which the water appears to run clean, one places a fine mesh over the piece and then applies to it a vacuum that can take up water and that has a variable control for regulating strength of suction. The vacuum is then applied to the screen and additional moisture is extracted from the piece.

My experienced collector friend says that even on a just washed rug that seems quite clean, the moisture removed by this vacuum process will be jet black, signaling how feeble our usual washing methods are with regard to getting out the dirt that is there.


R. John Howe

Posted by R. John Howe on 09-17-2005 09:07 PM:

Dear folks -

I would now like to return to Filiberto's attempt to defend a little the aesthetics of my rug.

It should be apparent that I rather like it despite what seem to me its varied faults. But here I want to examine Filberto's indication above about the drawing of the field.

He said in part:

"Bennett’s #155 seems having a more sophisticated and mature drawing and execution compared to yours.

"But… did you notice that, while in yours the lattice is organized in a pretty logical and obvious way (that is, starting from the top we see a row of four whole stars, then a row with two whole stars preceded and followed by an half star and so on…) on Bennett #155 the inner “window” framing the lattice is too narrow to show a row of four whole stars. It “truncates” irregularly the lattice on the left.
So the composition shows (from the top left): first row has ¾ of a star followed by 3 whole stars, the second has a fragment of a lattice, three stars and half star, and so on…

"The question is:

"Do you think the weaver did this “asymmetric framing” on purpose or it was a mistake?"

My thoughts:

I don't know, but I agree that some aspects of the drawing of the field of my humble piece show some clear deliberation.

My piece started at the bottom of the image most directly above and the first horizontal row is of three full hexagons and two "half" ones. The two half hexagons are drawn so that slightly more of the star is seen in the right one that in the one on the left. As the rows proceed this left to right orientation seems to shift slightly to the right so that at the top more of the left hand star is visible.

As Filiberto says the Bennett piece is oriented so that the hexagons on the left side show more of themselves than do those on the right and so their stars are more complete.

I think you could argue that my weaver did a better job of estimating the real width of the field needed in order to keep the side half-hexagons in the three full hexagon rows about equal in width. She also effectively allowed horizontal space for four full hexgons in the four hexagon rows, something the Bennett weaver did not achieve.

Alternatively, you could argue that the Bennett piece weaver used the horizontal asymmetry of both the half hexagons and the less than full presentation of four full hexagons in four hexagon rows, to make clear that the borders merely surrounded (perhaps the position of the surround is by chance) the never-ending repeat character of the field design.

I think I would fault the weaver of my piece mostly for conventionalization of some details in the designs she used (although the simplicity of the main border devices may work to make it more graphically effective) and for failure to draw diagonals straight and crisply enough although I have immediately to say on this latter point that this variation is also one of the things I rather like about this piece.

I may be liking what Marla Mallett and Wendel Swan would call simply "bad weaving" but these irregularities function for me by giving some "movement" and "life" to the design and help it avoid the hard-edged, mechanical, to me, static regularity of the drawing in both the Bennett and the Sotheby pieces. But look at that last sentence, the conclusion is all in the adjectives rather than in any analysis and someone else can choose different ones to license a different result.

I think I'm still largely in the position I was in when I wrote the summary for "The Oops Thesis" salon. As will likely be apparent to those who see the sort of thing I put up here and read what I write about them, I am demonstrating a deplorable failure to "develop" as many experienced rug collectors would recommend.


R. John Howe

Posted by R. John Howe on 09-18-2005 08:20 AM:

Dear folks -

We are by now giving this piece more attention than is due, but
a few days ago David Hunt asked what the knot count is on it. I just remembered finally to do one.

There seem to be about seven knots horizontally and about five knots vertically, so it's about 35 knots per square inch.

The books say that Genje, Kazak and Karabagh pieces are often structurally indistinguishable and I noted one instance of this liklihood as I measured. The height of the knot nodes on this piece seem more than twice their width. Tall knot nodes are cited by Neff and Maggs as an indicator that a piece may be Kazak.

Also, one strong indicator that a piece may be Genje is that some of them have cotton wefts. The wefts in my piece seem to be wool.


R. John Howe

Posted by Amir_Aharon on 09-18-2005 02:46 PM:


As I said on the start of this thread, I like this rug. Since then many comments and queries have popped up about it's appeal, it's irregularities, age etc. I was wondering what makes a Caucasus rug attractive or special to collectors or even ordinary passers by. Howcome Bidjovs and Karabaghs are not as popular as Kazaks or Shirvans?

Most of us would agree that Caucasian rug collecting in the west started out with the love for their geometric designs and even so more for their bright profusion of colors. Naively executed, yet appealing to our senses. Sue Zimmerman who probably is not a Caucasian rugs collector was struck by John's Gendje(?) and right away expressed her feelings for it. Putting away the rarity and age of the piece for a second, it must be color blend in harmony with the geometric design.

I picked up three rugs from recent Sotheby's sales which fetched ENORMOUS prices, for your judgement. Of course rarity must have been an important factor here, but probably the crucial parameter was the use of COLORS in the right locations.

The use of yellows, greens and whites along with the ubiquitous red upgrades the attractiveness of these rugs. So be it with John's piece. The Karabaghs which are influenced by Persian CURVILINEAR designs and somewhat dull reddish colors are not popular even if they may have age or rarity.

I have also posted a black and white Karagashli to compare with the blue ground Sotheby's piece. As you may notice the field rectangular shapes are less curved than the Sotheby's Karagashli along with so many other differences!). This is what happens when a design is repeatedly copied from one rug to another. The iteration of copy eventually ends in a much simpler and more rectilinear design. That's an indication that the black and white piece is a later production. This is Jon Thompson's "design transmission theory" on comparing ages of rugs with similar designs. I believe John related to this theory when he mentioned the execution of the famous Ardabil carpet. Dave also touched on this theory when he talked about 'deterioration' of motifs going from tribal to sedentary and also vice versa.

Now back to John's (clean) Gendje(?), and the white wide border; one may say that the motifs there are much simpler than their parallels on the Bennett border; and so according to Thompson's
theory Bennett's Gendje should be earlier;which probably is. The field of John's rug, on the other hand is geometrically more perfect (courtesy of Filiberto's measurements), than that of Bennett's book. So where does that get us, if we apply the "design transmission theory" ?

The Wendel "OOPS thesis" is also applicable to John's white border. Could the woman have made a mistake when copying the motif from another rug? Maybe she purposely wanted a simpler and perhaps more appealing element from her point of view? Anyway, she came up with something pretty which most of us appreciate.

John, thanks for the washing instructions. Use VINEGAR next time!



Posted by Tim Adam on 09-18-2005 03:43 PM:

Do you think the weaver did this “asymmetric framing” on purpose or it was a mistake? This irregularity makes it more or less pleasing compared to yours?

Hi Filiberto,

Interesting questions. Maybe someone who has some weaving experience could let us know how easy/difficult it is to plan a rug layout on a simple loom. Since the grid layout in John's rug is perfect, I think it is more likely that it was planned that way, rather than that it came out this way by chance.

The weaver of Bennett's rug was, in my opinion, more skillfull than the weaver of John's rug. If John's weaver could accomplish a perfect grid, then I'd expect that Bennett's weaver could do the same. Now it didn't come out that way. So, I conclude that either it was planned to be asymmentrical or the weaver didn't care about it.

One could also argue that it really doesn't matter what the weaver intended. The only thing that matters is the final product. There are lots of weavers who created awful rugs. So, they can hardly be regarded as an authority in distinguishing between the good and the bad.

I think the 'irregularity' in Bennett's rug is a strength. In particular, the left side of the grid is better than the right side. The reason has to do with the concept of good proportions. This is hard for me to explain. But think about designing a house. It is typically not a good idea to have a window right below the roof, or right next to a corner. For a "good look" you need to have some space around the window. The same principles are at work in good rug design.



Posted by Filiberto Boncompagni on 09-19-2005 06:01 AM:

Hi John,

Regarding the attribution: I see you are still in doubt between Kazak or Ganja (I write Ganja in this way because that’s how they say it in Daghestan).
Your rug’s structure is perfectly compatible with a Kazak, as one should expect from a Ganja. If we look at the design of the border, it’s typical of Ganja. We have another example with almost identical design, identified as Ganja. You can call it Kazak, if you wish, but for me Ganja is a more satisfactory guess for the collector’s chronic anxiety for attribution. (Then, perhaps it was woven in Karabagh, who knows?)

Hi Amir,

I think that COLOR is the reason why those Sotheby’s pieces fetched such enormous prices – especially for the green field Bordjalou.
About Thompson’s theory on degeneration "design transmission theory", I roughly agree with it… with a pinch of salt.
Let’s not forget that there are – and were – more skilful and creative weavers working side by side – and often on the very same rug - with lesser-gifted ones. This could result in having two similar rugs of the same age but with a different quality. OR a rug perfect on one side but with defect on the other.

I don’ think the latter case (two weavers working on the same rug as explanation for its asymmetry) can apply to Thompson’s # 155, though. That one, either it was a mistake (bad planning) or it was on purpose.
Which brings me to Tim.

(Hi Tim) I also like more the extra “kick” given by the framing asymmetry and I think it was a deliberate choice. I’m not terribly sure about it, though. It could have been a mistake that turned out esthetically well.

Browsing on Sotheby’s I found this two examples of asymmetry.
These, in my opinion, are plain mistakes – or bad weaving, if you wish.

What do you think?



Posted by R. John Howe on 09-19-2005 06:55 AM:

Hi Filiberto -

I am not particularly suspecting that my rug is Kazak rather than Ganje. I have just been struck anew, as we searched for an attribution, with how similar (the literature acknowledges) Ganje, Kazak and Karabagh pieces are and can be.

And I have just been reporting various facts that might bear on this attribution collage as I have encountered them.

I don't have any real collector anxiety about placing this piece precisely in one of these categories, although I admit that we collectors give lots of indications that we suffer from that disease.


R. John Howe

Posted by Filiberto Boncompagni on 09-19-2005 07:44 AM:

Posted by Sue Zimmerman on 09-19-2005 08:21 AM:

Hi Amir and All,

Caucasian rugs, to me, are very accessible. I don't see them as "tribal" at all. Maybe my Caucasian genes have memories which resonate, or something. I also get a big kick out of children's art and have no trouble framing some of it regardless of it's having been drawn on such as unsuitable foundations as envelopes.
The motifs and design layout on John's rug are probably as old as the trade winds. Check out Jenny Housego's book, for instance.
For color usage, and symmetry, within this design family, check out some of the areas where traditional Latvian mittens were knit. No infinite repeat framing of motifs there, that I'm aware of.
To me it's unthinkable that the Caucasus lacked entry level "training schools" for the rug trade. Due to the Balkan amber trade I am sure some of those Latvian mittens ended up in the Caucasus. Just some thoughts. Sue

Posted by James Blanchard on 09-19-2005 11:44 AM:

Symmetry... which direction?

Hi all,

I always find that design symmetry, or perhaps more importantly design asymmetry, a key factor in the aesthetic appeal of a rug. In general, I find some amount of asymmetry to be pleasing, particularly in rugs with geometric repeating designs.

With respect to the two rugs being compared here (Bennett and John's), I like Filiberto's observation but it seems to me that the main axis for (a)symmetry is diagonal, not horizontal. That is the way that the colour scheme is oriented. It probably all comes to the same conclusion in the end, but seen in that light there are many more ways to look at asymmetry since one can reference the vertical and horizontal borders of the central field.

Just a thought....


Posted by Filiberto Boncompagni on 09-19-2005 12:03 PM:

Hi James,

You are right. Bennett’s rug is asymmetrical (not identical on both sides of a central line and considering the row of blue stars as the central line) also in diagonal.
Cannot say about John’s one, though.


Posted by Amir_Aharon on 09-19-2005 12:09 PM:

Hi Sue, Filiberto, Everyone,

Here are three more Ganjehs(that's how the Persians pronounce it to re-arouse resonance in Sue. By the way, Latvians probably make good mittens with interesting designs but no rug weaving tradition in Riga. The Ukrainians on the other hand, are famous for their kilims and pile rugs (Bessarabians).

Going back to Filiberto's last post, I also agree with Tim that assymetry in design is usually nice (whether deliberate or not); in a way it breaks the monotony. But not in the fashion of Filiberto's Moghan(?) Memling Gul rug was executed; as he said, "this is bad weaving" and has no aesthetic virtues whatsoever.

Please notice the ivory borders of the above three Ganjehs, remeniscent of John's piece. There is also the yellows and greens which have a positive effect on the overall appeal. The rug with the two columns of Memling guls is symmetric in comparison with Filiberto's similar piece which has "mistakes" and which seems less attractive.

Jon Thompson's "transmission theory" is not concerned with the above "mistake" phenomena but rather it's more about the change in form of an element (say from curvilinear to rectilinear) as a result of copy iteration throughout the years. He certainly was not talking about degeneration or deterioration of a motif, only the change in form; it comes handy when one compares ages of two similar designed rugs.

In fact, the later rug might even be more exciting than the earlier piece. A good example of this theory would be the typical Turkish Bergama design with 4 octagons, one in each spandrel (in the earlier rugs) and it's "degenerate" descendent where the 4 octagons change to ordinary rectangles eventually. Some design elements may totally disappear. The later piece may be naively drawn at one stage or another and mistaken for an earlier rug.



Posted by Tim Adam on 09-19-2005 02:19 PM:

Hi Filiberto,

I agree about the first of the two rugs you posted. An unfortunate composition. The second one I am not so sure. The asymmetry doesn't destroy the overall balance. There are other factors that would bother me more.



Posted by Walter Davison on 09-19-2005 04:16 PM:

Mr. Howe,

Would you mind putting up comparison shots of the “before washed” and “after washed” rug? Maybe that wouldn’t be very interesting for most of the folks here, but as a perhaps irredeemable novice, I would like to see the difference after all that well-described work you did (and maybe to allay the fears of some).



Posted by R. John Howe on 09-19-2005 06:58 PM:

Mr. Davison -

The piece is still drying in my storage room here in my condo apartment. I give it a day or two yet.

You have the only "before" images I can provide.

I can take after images, but I think they may function only to show that there were no additional dye runs.

One thing that does happen in a washing like this is that areas in which there are condition problem become more noticeable. Part of that is because even in a relatively gentle hand washing like this, some material is lost from the rug. What happens is analgous to what you see in the filter of your washer after running a load or two, but of course not that definite.

Since I talked a bit above about how I personally usually go about such a washing, it might be useful to mention one additional trick of the trade that an wise Afghan dealer here taught me early. It is a technique that can be important with larger pieces that are somewhat fragile, but one that is generally useful.

A rug immersed in water takes in a lot of it and becomes pretty heavy. If a larger, more fragile rug, is lifted wet directly out of the water, the weight can damage it. One way to prevent this, once the last rinse is complete, is to lay the rug out on the bottom of the tub and then gradually roll it up with the wefts horizontal. Once the rug is entirely rolled up but still laying in the water, take one end and drag it to the corner of the tub and stand it up on end there.

With this rug, I did the final rinse with the shower and so the tub was empty. That's better but not always feasible in all cases. Sometimes you want the rug out of the last rinse water quickly to minimize the rinse water settling in the rug as it drains out.

In six or seven hours (I often let a piece stand like this over night) the water will largely drain out of most of the upper three quarters of the roll. There will still be quite of bit of water in the lower quarter, but the overall weight of the water in the piece will have been greately reduced.

You can then with less danger lift the piece out of the (now empty) tub (reversing it momentarily so that it doesn't drip as you carry it through the house) and take it to an area where you can unroll it and spread it to dry completely.

I have two different places where I do this latter drying. If the piece is not large I hang it over the top of our laundry room door (the floor is tile). If the piece is larger, I take it to our storage room and spread it in a corner where the moisture will not harm anything. (Things, of course, would go faster if I could spread it on the lawn in good sun, but I don't usually want to sit out there with a good book most of the day even during those parts of the year that are warmer here.)

But to go back to your request, I will take some photos of the after condition of this piece and put them up.


R. John Howe

Posted by R. John Howe on 09-24-2005 12:05 PM:

After Washing Images

Dear folks -

This post is just to deliver on a promise to Mr. Davison to provide some images of this yellow ground fragment after my washing of it.

These images do not provide a good basis for detecting whether color has improved. My wife and I are "down" a bit with colds and so these photos were taken inside and in poor light.

They do, though, I think demonstrate that no further dye runs resulted from this washing. (I should note that Mr. Davison sounds as if he might be tempted to do some rug washing of his own. I would offer only this warning. Since he seems to be interested in younger Turkmen rugs he should be especially careful, since some of the red synthetics the Turkmen used do run readily into white areas.)

I'll put the after images up below in no particular order, FYI.

You will have noticed that more bare areas appear. This is the result of the loss of pile in washing that I mentioned previously.

There is one last thing I should mention. There is a tag sewn on one corner of the back of this rug.

It was sewn on after the rug lost its ends. The writing is not legible in this scan but says:

For expert cleaning and repair of your valuable rugs
S.S. Kaish Co.
Oriental Rugs
Syracuse, NY
(there was a telephone number now obliterated)

This is an interesting label. It indicates that there was once in the rather small city of Syracuse, NY a rug store other than that of the famous "Col." Charles W. Jacobsen, one of the real personalities in the oriental rug world in the approximate 1950s and after.

Jacobsen was perhaps the first person to sell oriental rugs by mail. He had a catalog and some slides strips that he would send you, if you expressed interest. Then, if you planned to purchase, say three oriental rugs, he would send you six on approval by mail. A flamboyant personality who likely knew his rugs pretty well and who wrote a couple of books you can still buy. Unfortunately, Jacobsen is so busy bragging and generally trying to "look good" to the reader that this overwhelms the quality of his writing. You don't get to see in his books what he likely knew about rugs.

It is a little remarkable to me that there might have been "room" in Syracuse for another oriental rug store. Kaish seems no longer to exist (at least at the Google level). The Col. is long dead, but Jacobsen's is still very much alive, with an excellent web site.

Just a little historical tidbit here at the end triggered by my remembering to mention this label.


R. John Howe

Posted by Walter_Davison on 09-24-2005 10:34 PM:

Mr. Howe,

Thank you for putting up those pictures. Very interesting. You didn't say if you were satisfied with the results. Are you? Seems like the low pile places show up more easily now. That’s OK for your situation – is that right?

I probably would have done as you did if I had a loosely woven rug, even with a nice rug like yours, though with less expertise, of course.

You ventured that I might be considering washing my newer Turkoman type rugs. I’ll take that as an invitation to join in.

The answer is "Not so, no way," at least in the same manner as you did. That is true not so much because of a fear that the red dyes might run, and it is not clear to me that that would be the case, so says Mr. Eiland in my library copy of one of his earlier texts, but because of the likelihood of uneven shrinkage of the warp and weft threads. Since Afghan Mauris are so tightly woven, many 200 to 300+ kpsi, on wool warp/weft foundations, uneven shrinkage of the foundation permanently ruining how the rug lies would be my greatest fear for my situation. Loosely woven all wool rugs are no problem, I think.

The most I would ever try to do would be to gently wipe the top ends of the pile in spots without getting any water on the warp/weft threads. Even then, I could very well fail - I have, of course, and that's how I know this. At the same time, I think lots of folks in Afghanistan have experienced this very fate with their rugs, and I once saw a cure for this uneven shrinkage. I haven't seen this procedure described in any texts, though I suppose it has been, and maybe you know all about it. On the off chance you have not seen it, here are the basics. What one needs for this cure is a thick wooden pole and other wood supports, a car jack, large metal spikes, some water, and ample sun. The rug is secured, backside up, to the bare ground by means of the fringe being tied around numerous spikes pounded into the ground. A thick wooden shaft is jacked up in the middle to stretch the rug taut. Then water is sprinkled onto the back of the rug, and it is
left to dry in the sun which causes some of the warp/weft threads to partially break, and I think the tautness is also tended to at the same time. The result is a very flexible rug – hardly any stiffness at all, none in fact, and the the rug will now lie perfectly flat. Some folks might say the rug is now really damaged goods, but what is a rug that wouldn't lie flat and is all crinkly. Would the condition after such stretching be like your 100+ year-old rugs, I wonder?

Below are pictures I took of some young Turkoman fellows in Jalalabad, Afghanistan, stretching rugs to fix the problem I fear if I attempted to wash an Afghan Mauri type rug. These are photos of photos, so the quality isn't very good. One shows the securing process, one the car jack being put in position, and the third a whole courtyard of rugs being stretched. So, though it is true that I may have at one time or another considered washing some and then stretching them in case it went badly, I have never had
sufficient courage to actually do it because, being a novice, of course, I would only make it worse.



Posted by Tim Adam on 09-25-2005 10:17 AM:

Hi Walt,

Very interesting report. How much water is sprinkled on the rug? Just a little to moisten it, or is the rug completely wet?


Posted by R. John Howe on 09-25-2005 10:22 AM:

Mr. Davison -

Am I satisfied with the results of my washing of this fragment?

I did so primarily because I intend to take it to a picnic held by our local rug club and I didn't want it to feel too dirty.

It now feels cleaner but in truth I have not had it out again in full sunlight since I washed it. But I think Steve's point will likely hold. The rug was very dirty. I've now washed it without experiencing dye runs. It is likely that the color in full sunlight will be at least slightly better than what it was when I purchased it.

On the other hand, the problem areas in it are now also likely more visible.

But, yes, for my main purpose, I am satisfied with this washing.

Different subject: I have not tried to remove "bubbles" from rugs using the method you describe which does seem to me to entail damaging the structure of the rug. I have seen others tack a rug down (stretched pretty tight) on a wooden floor and then sprinkle it with water and leave it overnight hoping for a similar flattening effect.

I know that one source of bubbles in Turkmen rugs is handspun warps that are somewhat irregular. I'm fairly close to one producer of them, who has in recent years moved to machine-spun wool warps (one of the few compromises he makes with traditional Turkmen weaving methods). This seems to result in rugs that predictably lay flat, something "westerners" want but that does not come in much in conversation in a yurt. :-)


R. John Howe

Posted by Walter Davison on 09-25-2005 11:00 AM:

Hi Tim,

I saw them put their hand in a bucket of water and then they repeatedly shook it off onto the back of the rug. So, it didn't look like they were soaking the rugs at all, rather it was just enough moisture for the desired effect. Too bad I didn't learn more about it.

Mr. Howe,

Your remark about the likely conversation in a yurt is very well taken, and I am sure you're right.


Posted by Horst Nitz on 09-26-2005 02:22 PM:

Hello Everybody,

please accept my apologies for entering the discussion at such a late stage to readdress the question of attribution.

I suggest John’s rug is Daghestan rather then Gendje. There are several indicators. Wool: after all the battering this rug has received, the superior quality wool still displays shine and spring. The lavish use of costly white wool (and this includes the yellow as white wool is prerequisite for yellow yarn) in a village rug suggests easy access. This is where the famous white highland sheep of the northern Caucasus may come in. Design: all-over lattice designs, a light yellow field and red-grounded medachyl borders are met more frequently in Daghestan rugs. Structure: In the absence of excessive weft shoots a 5 V x 7 H knot count suggests some warp compression, which apparently is a feature of many Daghestan rugs: the image of the back of the rug shows a slightly ribbed surface. A three-part selvedge is another typical element.


Horst Nitz

Posted by R. John Howe on 09-26-2005 08:25 PM:

Hi Horst -

Thanks for your considered views here. They seem a departure from the attributions most of us have suspected.

I'm tempted now to take the position that the height of the knot nodes indicates that this piece is likely Kazak.

Then we would have someone on every point of the attribution triangle that is acknowledged in the literature, plus your Dagestan indication.

Of course, the Kazak position would be less well defended than are the Ganje, the Karabaugh and Dagestan ones. :-)

One last thing: I got sickish with a cold, lost track of my week, and didn't attend the rug club picnic to which I intended to take this piece, so I have nothing to report about it from those who attended.


R. John Howe

Posted by Horst Nitz on 09-27-2005 03:20 AM:

Hi John

I feel a bit fluish myself and hope it will wear off in the course of a long working-day ahead; or as a patient of mine has just written in another e-mail: "in Brasil we say, a flue is an illness for the rich, the poor simply have to get on with their work."

Poor us. Good health to all.

Regards, Horst

Posted by David R.E. Hunt on 10-02-2005 11:12 AM:

Analog Alert

Hi John

I stopped at Borders Books in Germantown yeasterday morning on the way to the family estate and happened to pick up a copy of "Starting With Antique Oriental Rugs" or somesuch, a new rug book by Murry L. Eiland. For the book itself I'd say standard introductory text, but there was a plate of a rug labeled as Gendge which sported a dead ringer of the pattern and color of the lattice field in your rug. Seems some confusion with the Kazaks regarding attribution, owning to the variety of people in this area, up to and including Azeri Turks. The presence of cotton is diagnostic, if my memory serves


Posted by Filiberto Boncompagni on 10-03-2005 05:11 AM:

Hi John,

For your reference…
I stumbled upon this, browsing an old Hali (issue 78).

Actually, it has an ivory field and it’s said to be Shahsavan…



Posted by R. John Howe on 10-03-2005 06:44 AM:

Hi David -

Thanks for additional example. Yes, Genje (Ganje) pieces are one of the Caucasian types claimed to have one possible distinguishing feature. That is the presence of cotton in the structure.

The book you encountered is, I think, one by Murray Eiland III, the son of the Murray Eiland, Jr. who has written so extensively. The son is also co-author with his father on the most recent edition of the "comprehensive guide" which is comprehensive now since it includes Chinese, Tibetan and some other types not treated in the "guide" previously.

Filiberto -

Thanks for the citation, the photo and the further attribution.

Wendel Swan has been politely silent about my wreck, but your post may lure him yet to declare that my piece should under no conditions be associated with the Shasavan.


R. John Howe