Turkotek Discussion Forums

Turkotek Discussion Forums (http://www.turkotek.com/VB37/index.php)
-   Virtual Show and Tell (http://www.turkotek.com/VB37/forumdisplay.php?f=3)
-   -   An Ersari (Ali Eli?) chuval with personality (http://www.turkotek.com/VB37/showthread.php?t=3596)

David Katz June 12th, 2016 11:50 PM

An Ersari (Ali Eli?) chuval with personality
Hello Folks,

I just acquired this rather large chuval (3 feet 4 inches x 5 feet 7 inches (101.6 x 170.2 cm) which looks to me to be an Ersari/Ali Eli piece from around 1900. There is considerable use of silk, but only in the lower half or so of the chuval, which can be seen as a kind of highlighting in the two "floral" borders and also filling some of the double ram's horn motifs. I'm also struck by the beautiful (to me) banding on the sides. Would love to see what others think about this piece.

David Katz









Kay Dee June 13th, 2016 06:56 AM

Beautiful piece!:thumbsup:

Rich Larkin June 13th, 2016 02:17 PM

Hi David,

Very interesting! If I've seen those bands along the edge on a piece like this (i. e., Ersari juval) before, I've forgotten it.

I take it the (very pretty) gold (or amber?) silk is the one color in which that material is used. Right? Is it also used in any of the opposing gul quarters, where I see a similar color appears?


David Katz June 13th, 2016 02:24 PM

Hello Rich,

Yes, the shiny amber/gold is the silk, and I did not see any in the gul quarters with the light brown wool.

I too have never seen the stripes or banding on the sides of other juvals; has anyone else?


Chuck Wagner June 13th, 2016 03:57 PM

Hi David,

Congratulations are in order. It is not easy to find earlier Ali Eli pieces. I would judge that yours is early 1900's. I have a later (and equally enormous) complete Ali Eli piece; I would guess WW-II-ish in age; images are below. Note also the black & white image from Mackie & Thompson which was taken in an Ali Eli dwelling.

I think the stripes on the side, and the brocade on the reverse, are distinctive. Note the the Russian tradecloth patchwork on the seam. As I recall, Dave Hunt owns (or owned) a similar chuval.

Chuck Wagner






David Katz June 13th, 2016 07:42 PM

Thanks Chuck,

If this banding is truly unique to the Ali Eli, it makes one wonder why other tribes did not take advantage of the lateral reaches of their chuvals as an opportunity for this or other kinds of embellishments. Maybe there were rules about not writing in the margins?;)



Rich Larkin June 14th, 2016 03:30 AM

Good find on the Mackie and Thompson illustration, Chuck! It features a less emphatic version of the infiltrating horizontal stripes from the edges. Your piece does, too.

David, I was referring to the color in the quarters in, for example, the gul in the upper right corner. It also appears in the kotshanak design of the main border, also in the upper right corner. Nevertheless, I see the silk is essentially a different color. I wasn't sure how close those colors were to one another, considering the much greater reflective quality of the silk as compared with the wool.


David Katz June 14th, 2016 04:38 AM


There is indeed a light brown wool in the main guls and the kotshanak border that has a similar tone to the silk, especially in full sunlight as in my photos. However, one of the double ram's horn motifs in the kotshanak border, shown in my fourth image, actually has both. If you look carefully, you can see that the silk does not fill the upper ram's horns; in fact, the point at which this transition between silk (below) and light brown wool (above) corresponds to a discontinuity in the weaving as a whole. Look, for example, at the vertical brown lines one either side of the border. I will post a higher magnification image of this area.



David Katz June 14th, 2016 04:53 AM


Here is the enlarged image, in which I've indicated the transition from silk to wool (dotted line) as well as the vertical discontinuity (arrow).



Rich Larkin June 14th, 2016 01:44 PM

Hi David,

That is an odd little set of weaving circumstances. Shifts to the right or left are not uncommon in weavings of this ilk. We usually attribute it to some felt need on the part of the weaver to adjust in order to get the overall layout of the design right, or something like that. What's odd is that she (or her aunt, etc.,...) went to the wool at that point. Certainly, whoever was on the loom at that juncture was aware they were in the middle of the silk part. Go figure.

Were they running out of the silk? You mention that it is confined to the lower part of the piece. I take it the "two 'floral' borders" you mention are the ones made up of back to back small boteh-like devices in a meander (called, I believe, the badam border, among other esoteric names). Apparently, the silk also appears in those areas. How far above the kotshanak with the material shift does silk appear elsewhere in the bag? I'm striving to spot it in some of your images, but it is elusive.


David Katz June 14th, 2016 02:52 PM

Hi Rich,

There is no silk in the kotashank border above the transition point indicated in the annotated image. There is, however, some silk highlighting that extends a little higher up in the badem borders, and then stops. I wonder if the weaver's original intention was to use silk throughout, and then, as you suggest, the supply ran out. I'll also point out that the motifs inside of the kotashank elements are different in the upper and lower halves of the chuval; stars in the upper half and circles in the lower. Might this strengthen the hypothesis that we're looking at the work of more than one weaver?



Rich Larkin June 14th, 2016 06:12 PM



I'll also point out that the motifs inside of the kotashank elements are different in the upper and lower halves of the chuval; stars in the upper half and circles in the lower. Might this strengthen the hypothesis that we're looking at the work of more than one weaver?
Sounds very plausible to me. Anybody who has been chasing after this kind of material for any appreciable time knows there are endless opportunities for asking the question, "What was she doing here?" I wonder whether your piece reflects a little intra-family one-upmanship? The sister takes over the weaving and decides to show everybody how her sister wasn't doing it right. And by the way, let's not use up all that silk on her juval.

I shouldn't be so cynical. As the French used to say, "Honi soit qui mal y pense."


David Katz June 14th, 2016 11:39 PM


On the other hand, perhaps one of the sisters was thinking "Honi soie que mal on voit!" Anyways, I think you're onto an exciting plot for an Ersari soap opera...

Another interesting feature of this juval is the different wrappings of the selvages (see below): the white is cotton, the gray appears to be wool, and the dark brown is definitely a coarse animal hair, e.g., goat.


For Chuck I've got some images of another Ersari juval that is obviously much more recent than the Ali Eli and yet has embroidery on the back that is similar to the juval you posted. Interestingly, in both yours and mine, the front of the embroidery is facing the inside of the bag:





Rich Larkin June 15th, 2016 12:09 PM

Hi David,

On my screen, the red in most of your initial images is on the tomato side. In your last two detail shots, it is more to the liver persuasion. Then, a couple are in the middle. Which best represent the true color of the piece in your estimation?


David Katz June 15th, 2016 01:28 PM


Are you referring to the 3 photos of the 1930s/40s-ish juval? The last, showing the detail of the embroidery on the back, is very true to the actual colors. There is a "bad" tomato red dye that was used for details throughout (you can see where it ran on the back of the bag); the ground color is more on the liver end of the spectrum.


Rich Larkin June 15th, 2016 10:25 PM


I was referring to most of the images in frame #1 (tomato red), the second last in frame #1 showing the horizontal stripes in a vertical position (middle shade), and frames 9 and 13 (top shot) for the liver tone. I like all of them; I'm just wondering what the effect is in real life.


David R E Hunt June 15th, 2016 11:48 PM

Hi Guys

Not to divert from the present line of inquiry, find links to the fore mentioned chuvals here and here


David Katz June 16th, 2016 12:11 AM

Thank you Dave; beautiful piece. It looks like the sides might have been reduced; do you know if this juval had the stripes we've been discussing?



David Katz June 16th, 2016 12:59 AM


Originally Posted by Rich Larkin (Post 20650)

I was referring to most of the images in frame #1 (tomato red), the second last in frame #1 showing the horizontal stripes in a vertical position (middle shade), and frames 9 and 13 (top shot) for the liver tone. I like all of them; I'm just wondering what the effect is in real life.


Hi Rich,

There is no tomato red in the Ali Eli juval; it must be a monitor issue. The ground color is the cinnamon brown typical of many older Turkmen pieces (some of the close-up images were taken with artificial lighting, which may have produced some distortion).



Rich Larkin June 16th, 2016 03:56 AM

Hi Dave,

Of course, like rugs, tomatoes do not constitute a monolithic block of color. :fez:

I was referring mostly to the central field in those opening images. That color seems more intense and saturated than, say, the strips along the outside of the outer borders, next to the horizontal stripes. Anyway, there seems to me to be a significant difference between the color in those images and the shade manifest in frames 9 and 13. I mentioned it because I consider there to be a recognizable palette typical of certain older Ersari rugs that features a liver-colored madder. I wondered whether your bag partook of it.


David R E Hunt June 16th, 2016 10:41 PM

Hi David

No, both these chuval demonstrate this "piano key" motive on the lateral sides. It's just that they are stuffed as if pillows,
so the edges are bent out of camera view.Cameras can do strange things to objects within images,especially carpets, as I have come to learn : ).




I agree Rich, maybe it's my monitor, but the ground color of this chuval,seen at distance, sure does look tomato or some such lighter rich red.
I googled "cinnamon brown" and the results yielded a much darker color than on the ground of this chuval.

This said, cameras and light can do strange things to carpets, and in my experience their can be quite a discrepancy
between the apparent color when viewed at a distance VS viewed/photographed at close range.A lot of factors at play
yet I suspect light, with all it's attendant varieties, the primary culprit.

I have seen a fair number of, or at least what appear to be, these over sized bags and bag faces of some common affinity.

None especially old but I think there are a few around that date to the pre-synthetic age. All large and used as storage in a yurt,
pegged to a lattice and with this band of embroidery facing outward for obvious reasons.Or were they? There is record of these
large bags being used as storage for grain, not among the Ali Eli (who ever they are aside from Ersari?) but among Turkmen in general.

I have seen a few piled Tekke variants of such gigantic size, but few and far between.

This 12 triangle or such archetypal gul defines the format standard, in combination with the over sized chemche minor gul which stands
as a dimensional equal to the accompanying major gul. In short they are both the same size. The almond or badam border and the kochack
border seem to define the border repertoire of this group, and wheather or not the "dice" minor border should be deemed a requisite is anyones call.

Mine are both asymetric open left with a fair amount of warp depression, and I read many are of this construct.

I have seen some interesting examples of this class, and these two stuffed chuvals number among the most
beautiful objects I own, even with the minimal but glaring synthetic orange accents. The wool quality is very high, as with the pile.


Chuck Wagner June 17th, 2016 12:50 AM


Long time no chit-chat. Good to hear from you.

Gigantic indeed - note that the floor tiles in my image are 1 ft x 1 ft. The bag is 6 feet long and 3-1/4 feet deep. Strap it to a camel, toss the kids in, and you're on your way...:thumbsup:
Chuck Wagner

David R E Hunt June 18th, 2016 01:02 AM

Hi David

These two aren't the perfect match but there are many similarities.



In regards to the use of silk in your chuval,
I have come to suspect some may be nothing but substitution of materials, in this instance the upgrade from wool to silk.
Telling that it's introduction commences at what seems a pause in the weaving process?

What is the knot count of this bag face? What type of knot, and open left or right?
I suspect it might fall within the lower ranges of say Tekke or Kizil Ayak weaving...


Chuck: What Up?

Rich Larkin June 20th, 2016 01:46 PM



I googled "cinnamon brown" and the results yielded a much darker color than on the ground of this chuval.
You are a brave man. I can barely tell cinammon brown when I'm at the spice rack...nevermind when I'm looking at rugs! :errormonkey:


David R E Hunt June 21st, 2016 10:36 AM

Hi Rich

I Googled "Pumpkin Brown" and came up with a range of colors, some of which seem fairly close...:fez: Maybe "Cinnamon Pumpkin"?


David Katz June 21st, 2016 08:06 PM


Call me a Luddite, but my reference for the cinnamon color is, well, real cinnamon. Its not a perfect match, but in the ballpark. The juval is a tad more reddish than what I found in our kitchen spice rack.

The knot count is in the range of 120 - 130 kpsi; see image below (thanks to Rich and Steve for offline verification that I was doing this correctly).


Regarding knot structure, that's a bit more complicated. I found two different kinds of knots at the bottom and top margins of the juval, respectively (these were simply the areas where it was easiest to unequivocally view the knots). Along the bottom I found symmetrical knots; two strands of pile emerging between paired warps, with a collar underneath. However, along the top margin of the juval, where it had been cut, one could see the knot structure in cross-section. The image below shows what clearly looks like an asymmetrical knot.


(BTW, kudos to Rich for his lucid explanation of knot identification in the recent Turkotek post, "Unusual Turkmen Rug").


David R E Hunt June 22nd, 2016 02:47 AM

Hi David

No worries, was just funnin about the color. Between the selection of words we use to describe them, the type of light,
it's intensity, direction, distance and angle, to say nothing of the specs of the camera itself, color is a dicey subject.

To judge from the photos the nap is low on this bag face. Is it evenly worn or is there much damage to the pile?
I suspect these two bag faces to be the product of differing weaving groups. My chuval is in new condition and the
pile is high, so unless the pile has been worn away on your piece I would think it of another weave.
Ditto for the knot count, approx 100 VS 120-130. Colors differ as well, but there is considerable difference in age here.

It's not unusual to see the different types of knotting, especially symmetrical knots along edges in Turkmen weaving,
but the primary knot, the type in which the vast majority of the field is executed, is the type used to divine tribal provenance.
And knowing if the pile is open left or open right (my chuval has asymmetrical knots with considerable warp depression,
and open left, in example) might aid in pinning down the tribal origin of your weaving...


Chuck Wagner June 22nd, 2016 03:40 AM

Hi all,

I was able to get hold of a copy of HALI that contains an article relevant to this discussion.

If you all already have it and this is old news, fine, then I will stop with the images below. Steve has agreed to let me post scans of the entire article, if you're interested.

The reference is: S. Peter Poullada, Kizil Ayak & Ali Eli Chuvals, Turkmen Weavings of the Middle Amu Darya, HALI 148, September-October 2006, pp. 66-73.

It's an interesting article, delving into structural and design details in an effort to discriminate between Kizil Ayak and Ali Eli work. I have included the text specific to Ali Eli materials, as well as larger images of the chuval mentioned therein.






Chuck Wagner

David Katz June 22nd, 2016 03:50 AM


Fantastic! I've been looking for that article, which is referenced quite a bit by other writers. I would love to see the whole thing if possible.



David Katz June 22nd, 2016 04:11 AM


Originally Posted by David R E Hunt (Post 20666)

And knowing if the pile is open left or open right (my chuval has asymmetrical knots with considerable warp depression, and open left, in example) might aid in pinning down the tribal origin of your weaving...



In my image of the asymmetrical knot in Frame #26, the pile is facing up, so this knot would be open to the left. Thus, with respect to at least this knot structure, the major and minor gul designs and the extensive use of silk, my juval shares several characteristics with those described in the article by Poullada that Chuck just posted. Of course, mine and yours are 16-gul, and the Poullada article makes no mention of the "piano keyboard" banding on the sides. In answer to your question, the pile on mine is uniformly low.


Steve Price June 22nd, 2016 01:02 PM


Originally Posted by Chuck Wagner (Post 20667)
Hi all,

I was able to get hold of a copy of HALI that contains an article relevant to this discussion.

If you all already have it and this is old news, fine, then I will stop with the images below. Steve has agreed to let me post scans of the entire article, if you're interested.

Hi Chuck

I'm sorry, I misunderstood the question. I thought you were asking if it would be OK to post all the images in the article, not images of the entire article. The "fair use doctrine" allows others to reproduce excerpts, and HALI has never objected to our doing that. But putting up an entire article is way beyond fair use. The article is their intellectual property, and I don't think it would be OK for us to publish it without the express permission of their editor.

My apologies for having been my addle-headed self when I replied to your inquiry.

Steve Price

David R E Hunt June 22nd, 2016 10:36 PM

Hi Chuck

I remember when this article came out, have been wanting to read it for some time, but not yet.
Will have to track down a copy of HALI 148.
Thanks for the images, and the reminder.

Hi David

You had said that

"In my image of the asymmetrical knot in Frame #26, the pile is facing up, so this knot would be open to the left"

but it is my understanding that when a weaver places a section of pile yarn around the warp threads in the course
of the weaving process, the ends of the yarn are pulled back toward the weaver and hence down. If the pile ends
are directed up the weaving is inverted. This is common, carpets are woven upside down in relation to the
design with some frequency. When properly oriented in relation to the direction in which it was woven the pile of a
carpet should face downward. In short if the pile in your weaving is facing up (toward the viewer)the knots are
asymmetric open to the right and not to the left, but I could be misreading something...



Steve Wallace June 22nd, 2016 11:51 PM

Digital previous Hali editions
Hi all

At the risk of saying the obvious ... here goes .... if you currently subscribe to Hali, you can see all past issues digitally via the Exact Editions site. I can't provide a link as you need to be logged in to the site.

Thanks for the discussion

Chuck Wagner June 23rd, 2016 12:28 AM

Hi Steve(s),

I'm not surprised that there may be some concern so I will restrict any further extracts from the article to images of the pieces and their captions. These are scans from a hard copy that I own.

I have some additional reference material found on the internet; I'll add that tomorrow. Interestingly, on my 20th century chuval, the knots are asymmetrical open right.

Chuck Wagner

Rich Larkin June 23rd, 2016 01:10 AM


As I look at the cut cross-section of David's piece, I think those knots are open left.


David Katz June 23rd, 2016 02:51 AM


Originally Posted by David R E Hunt (Post 20672)
Hi Chuck

In short if the pile in your weaving is facing up (toward the viewer)the knots are asymmetric open to the right and not to the left, but I could be misreading something...



Just to be clear, when I say the pile is "facing up," I mean facing towards the top of the image, not coming out of the plane of the image towards the viewer. The viewer is looking down on the cut ends of the warps in cross-section. Does that help?


Chuck Wagner June 23rd, 2016 04:19 AM

David (Katz)

I think it's more definitive to fold the rug back, pile side out, along a horizontal row of knots and look at them closely, as in the following image:


Coming up from the bottom, the first row of ivory wool knots can be seen to have a tuft of wool (the pile) coming out of the right side of the knot node (the clusters of more horizontal fibers) and descending to the right. This is asymmetric open right.

Chuck Wagner

David Katz June 23rd, 2016 02:07 PM


Thank you for that helpful image; here is a similar view of the knots in my juval showing that they are asymmetric, open to the left.


Any further thoughts as to where you think this piece fits in relation to the Ali Eli juvals in the Poullada article, e.g., later (and larger) pieces of the same group, or something more distinct?

Best regards,


David R E Hunt June 24th, 2016 01:55 AM

Hi Chuck

You wouldn't have any Kizil Ayak images from the Poullouda article you could post by any chance?

Hi David

I wanted to be sure so I went back and checked my definitions in regards to open left and open right assym knots
in Eiland's "Oriental Carpets A Complete Guide" and compared the listed diagrams and found, assuming the fabric
properly oriented, your rug to be assymetric open right. The "open" space is to the right of the emerging pile tufts.


My guess in regard to provenance is Kizil Ayak...


Chuck Wagner June 24th, 2016 03:14 AM


That's inconsistent with what I use as a guide, which is Marla's book. John Howe (haven't dueled with him in years...) has these online with Marla's permission:



Also, yes, I do have images of the Kizil Ayak examples; I'll post some tomorrow. I have a longer post coming tonight, already. It will have a small example of a couple of them

Chuck Wagner

David R E Hunt June 24th, 2016 03:24 AM

Hi Chuck

Interesting. The diagrams in "Oriental Carpets" and Marla's book are clearly opposed.
I pulled the book out and compared. Hmmm...


Chuck Wagner June 24th, 2016 04:01 AM

David (et al.),

....Get a beverage, this is a long one....

Regarding your question about where I think your piece fits, the short answer is - uncertain but OK with my initial guess, late 19th to early 20th century.

I haven't researched any responses to Poullada's article, or anything else he has written. Indeed, on the internet, information on the Ali Eli is sparse at best. I'm sure that people with access to academic material, particularly Russian stuff, have a more complete picture than I. After I discuss your piece, I will include several images of text hoisted from online scanned copies of old British and French field reports.

Poullada apparently has a pretty nice collection of Kizil Ayak and Ali Eli chuvals but like any analysis, it's based on the data he had, and not on data that he didn't have. Other examples are probably out there that might negate some of his definitive criteria. Asymmetry direction is one that I suspect would not survive a full disclosure of all Ali Eli goods - maybe the earlier ones, but probably not the later ones.

On to a few details. First, the minor guls on your piece do not appear in his Ali Eli examples. They do appear in several of his Kizil Ayak examples:


But, your piece is consistent with his Ali Eli criteria - AsymL, major gul characteristics, etc.

Now, in the image from Mackie and Thompson, we can see several features that are identical to those on mine and Dave's newer bags (oddly, I get a better image from my cell phone than from my scanner - go figure):

First image


- At right we can see the double-ended latch hook border element with the diagonal line inside the interior square. Ditto on mine, although it alternates with one with a floral device inside the square on mine.

- At right again, just inside the short edge bars at the edge of the bag, we can see (sometimes) a white triangle-bar motif filling the space between the edge bars and the guard border. Ditto on mine.

- The badam (almond) gul guard borders are consistent with both your and our my piece.

- We see the bars at the edge. Shorter and fatter than on mine, but there nevertheless. Ditto on yours but yours are much narrower. Note that these edge bars occur on none of Poullada's Ali Eli or Kizil Ayak pieces. But some have been cut back and may have originally had them.

Second image:


- The double motif at the center of the major gul is the same as my newer piece, as well as yours, consistent with his Ali Eli criteria.

- The minor guls are very similar to yours; mine does not have the more pronounced diamond at the center.

- Knot direction is unknown but I would bet AsymR. Mine is open right, not open left.

So, what do I think ?

- I think the bars at the sides may be a more recent phenomenon, which is why I am inclined to place yours a little later than some shown in his article.

- I think the presence of the badam gul guard border is a more recent phenomenon. None of Poullada's Kizil Ayak examples have badam gul borders, and only one of his Ali Eli examples has them.

- I think knot direction evolved. Like Mae West said: 'I used to be Snow White, but I drifted..."

So, that's what I think, and why. Also, when I say "my piece" I think that we can say the same about Dave's bag.

INTERMISSION :dancer: :dancer: :dancer:

Now, as for Ali Eli research, here are the historical references I have found. You can all read them and we can discuss later if desired. I will only note that a) the Ali Eli had a few thieves amongst them and b) apparently the nature, and then later Russians, re-routed the Amu Darya at some point, and I wonder what (if anything) that does to Poullada's working hypothesis regarding which bank these folks lived on.













Chuck Wagner

David R E Hunt June 25th, 2016 03:03 PM

Hi Chuck

Thanks for the Kizil Ayak images and accompanying historical data.

There seems to be a lot of transposition, between the two tribes, Kizil Ayak and Ali Eli, of these design elements
which point toward some trans-tribal migration of said design elements.


http://adamanddavid1.homestead.com/elemen_4_red.jpg http://adamanddavid1.homestead.com/element3red.jpg

Compare a rustic Kazil Ayak from my collection with it's major and minor guls


to those of the Kizil Ayak in Peter Poullada's collection.
The minor guls are basically the same and the major guls are both of the archetypal variety,
but the internal drawing of the major gul represents the variable of interest, for this
twelve triangle banner gul found within is used as the major gul of David's bagface.


This twelve trangle banner gul also represents the signature major gul of the Ali Eli. The Kizil Ayak is the antecedent




Chuck Wagner June 25th, 2016 07:14 PM

Hi Dave, et al,

To support our discussion, I'll post a couple more composite images from the HALI article. But that's where I will stop; I have discussed this with Steve and we agree that we do not want to exceed reasonable use of HALI's material. I have tried to include attributions for all the material posted above as well, for similar reasons: credit where credit is due.

Let me note that Poullada's structural cafeteria for Kizil Ayak work include:
- Vertical-to-horizontal knot ratio 1.1 : 1, and as high as 2 : 1
- Very spare use of silk, largely in oder pieces
- Evenly spaced vertical rows of square knots
- No cotton in the foundation, ivory to light brown warps
- Asymmetric open right knots

I'm going to start with a chuval of my own, which I have for years called Ersari but - this thread having rekindled my energy - I am inclined to give it further consideration and wonder if any of you have opinions regarding a more narrow attribution.


Based on the dye job I would judge late 19th century for my piece, below. The borders are consistent with the diamond-in-a-rectangle motif so common on Kizil Ayak work, but that is also found on Ali Eli pieces.

The problem is, this one is asymmetric open left (see below). So, are his criteria less reliable on later pieces, or is this something else ? He says Ali Eli are open left, but makes no mention of dyrnak minor guls.:




I haven't found many chuval examples attributed as Kizil Ayak with dyrnak minor guls, but there is at least one out there (this one is very similar to mine) on RugRabbit, listed as sold, btw:


I have another likely Ersari piece I want to discuss, but I'll wait until we've run out of steam on the Poullada-style pieces. Here are the two composite sets of Kizil Ayak examples from his article, for design comparisons.



Chuck Wagner

David R E Hunt June 26th, 2016 09:37 PM

Hi Chuck

When our words fail us, so our definitions. I'm familiar with MM's treatment of textile structure
and of her propensity for self interpretation of the weaving process, which is all well and good
being qualified to do so, but if Marla and Peter Poullada are not on the same page in regard to labeling
knots we have a problem. Or do we?

Dichotomy abounds here, knot count, color, pile height, tribes. This is a trend, a tale of two tribes,
Kizil Ayak and Ali Eli, intertwined in both history and their textiles, the latter morphing into their
present day representatives. In the bigger picture I guess it really doesn't matter whether the knotting
is open left or open right, and the bagface Kizil Ayak or Ali Eli. But as a collector I would want to know,
and am perplexed by this knotting conondrum.

I consulted three resources, two books and the internet and found BOTH knots described asymmetric left.


Chuck Wagner June 26th, 2016 10:08 PM


I wonder if the Eilands provide rewards (in the form of 19th century carpet of your choice) for finding typos in their book.

Because I think you found one. I just looked at their description, for the first time, and I think it's backwards. Peter Stone's and Parviz Tanavoli show opposite relationships from Eiland & Eiland, but the same as Marla's; I'll scan and post later. But basically, all the text agrees: if the pile opens to the right of the knot and the wrapped side is to the left, it is open to the right.

Go figure.

Chuck Wagner

Steve Price June 26th, 2016 10:14 PM

Hi Dave

I'm more than a little puzzled when people who aren't novices differ in their report of whether a piece is asymmetrically knotted open left or open right. Making the determination isn't mysterious; it involves little more than determining which end was woven first and knowing which wrist you're wearing your watch on.

Is it possible that the knot type is non-uniform in some of these pieces? Perhaps people are accurately reporting the knot type in the small area in which they looked.

Steve Price

Chuck Wagner June 26th, 2016 11:19 PM


Here are the drawings from Parviz Tanavoli's "Kings, Heroes & Lovers - Pictorial Rugs from the Tribes and Villages of Iran". He has a pretty good system for coding knot geometry:



Chuck Wagner

David Katz June 27th, 2016 05:49 PM

Thank you Chuck for all of the technical and historical references.

I found Peter Poullada's email address online (he is president of the San Francisco Bay Area Rug Society) and sent him the images of my presumed Ali Eli juval. Here, with his permission, is his response:

"Hi David,
this juval appears to show many of the very characteristic elements of an Ali Eli although I have never seen a twenty gul version and the size seems much bigger than usual. Some of the elements that suggest this tribal attribution are: ASL knotting, (unlike Kizil Ayak and 75-80% of Ersari that are ASR), the very characteristic main gul with its little " c" forms in quarters ( whenever you see these you get ASL knotting), the chemche minor guls have typically symmetric layout ie no elongation or width, this form is shared with the Kizil Ayak. The palette seems a bit more brick red than other AE I have seen, at least in this photo, and the use of silk is normal ( almost never found in KA or Ersari) It seems to have more borders than usual but the comparison sample is probably just too small. Nothing out of the ordinary there.

So apart from the number of guls (and I don't really think that is a marker of tribal identity for any of the Turkmen tribes) I think an AE attribution is fine. Appears to have white wool warps which is more like the KA but again is not unknown with Ali Eli, but I can't tell the wefts. Often they are not the same wool color ( again unlike the Ersari who tend to use the same ).
So based on the photos you sent I would have no problem at all attributing it to the Ali Eli.

Hard to tell what age, second half of 19th ? Third quarter ? And most likely Ali Ali of Andkhoy ( Northern Afghanistan ) not Abivard, Southeast of Ashkhabad in Turkmenistan today, the other home of the AE post 1750's.
A nice piece !
all the best


Chuck Wagner June 27th, 2016 07:20 PM

Hi David,

A very helpful response. As you have his attention, would you mind asking him to take a look at the images I posted for mine (above) and provide any insight he might have ? It has a vertical to horizontal knot ratio of 1.8 to 1 (5 H, 9 V, on the average, counting on the back), which is one of his criteria for Kizil Ayak but, it is Asym open left, which he says is inconsistent with Kizil Ayak. And then there's the dyrnak guls.

The V to H ratio is well above the 1 : 1 he expects for Ersari pieces.

Chuck Wagner

David R E Hunt June 28th, 2016 12:29 AM

Hi Guys

Very good, end of the mystery :)...


Chuck Wagner June 30th, 2016 02:48 PM

Greetings all,

Hat tip to David and Steve :cheers: for helping me contact Peter Poullada, enabling an interesting exchange of emails. Peter reviewed and remarked on several detailed images of the pieces I posted earlier, as well as a third (mentioned, but first posted below).

A) Regarding the first , this mid-20th century Ali Eli piece:
These remarks may extend to Dave's bag as well; the two seem to be related.

Peter's remarks:


This one is almost certainly a 20th century Northern Afghan version of the old Ali Eli format.( See Loges) In my day in the 1970's in Afghanistan they would call this Labijar or Taghan, both towns near Shibergan in N. Afghanistan. The weavers themselves were probably sedentarized Ersari and almost all these are ASR, but a lot of the weaving villages were mixed up Turkmen tribal groups who had fled from the Lebab in the 1930's and had to some extent lost a lot of their " tribal " cultural identities, or at least the women didn't weave so " endogamously ". What is confusing is the extent that even in the mid-20th century some of these Northern Afghan weavers managed to continue to use vegetal dyes even though the synthetics started coming in full use in the 1890's.
I am intrigued and fascinated to see that the backs of your pieces are so decorated and wonder if we just aren't aware of the extent of back decoration in older " Ersari " pieces because they have been cut off and discarded for so long !
There was some additional discussion on the photo from Mackie & Thompson; Peter recalled that Pfeiffer's photos were taken near Andkhoy in northwest Afghanistan.

B) Regarding the second piece posted earlier
Peter's remarks:


Let me take on this dyrnak one. It is a very nice example. Sight unseen and without handling it I would say 1880's. The combination of the precise drawing of the guls which are archetypal to the Saryk,but appear in the late 19th in some of the Marv Tekke pieces ( see below for why) the use of the dyrnak-claw minor gul, that main border which is really quite rare, and that they almost always seem to use ASL knotting convinces me that they constitute a separate Lebab-region tribal group, i.e. Not to be lumped into the " Ersari " basket. And I am pretty certain they are not Ali Eli.

Just to restate the thesis and the message: from the 1650's on to the end of the 19th century there were between 25 and 30 different and well-defined
( self-identified) tribal groupings of Turkmen inhabiting this region along the Amu Darya, known from the 17th century onwards as the " Lebab" and now within the Lebap Province of modern Turkmenistan and into Northern Afghanistan.

Karpov's 1926 survey of the region, confirmed by Vinnikov in the 1950's gives us their tribal names, as well as detailing their relative populations and the locations of their weaving villages ( qishlaq). These two sources clearly and unambiguously identity and differentiate for us territories of the four main Ersari sub-tribes ( Bekaul, Kara, Ulugh Tepe and Gunesh), and confirm that there were many many tribes along the river who were NOT Ersari.
When I drove the length of the Lebab on both sides of the river in 2014, all 250 kms from Charjui to Kerki and back, we visited with my guide who was himself a Bekaul Ersari, more than 20 separate tribally self-identified Turkmen groups, including Kizil Ayak ( a silk weaving area and the location of an important sufi shrine ).

But in terms of linking tribes to weavings, we have in current use only four or five " labels " used by the trade to attach to all these Turkmen: Ersari, Kizil Ayak, " Beshir " ( which is a place not a people) and Burdalyk ( also a place ).And I am trying to add Ali Eli to the list. So unfortunately we are stuck with speculations in trying to attach more precise names to weavings even if we can be pretty sure they constitute products of a separate ( as opposed to " Ersari ") tribal group.

But we can derive some hints: in some cases we know precisely which qishlaq a weaving came from, thanks to some of the labels and acquisition cards Moshkova left us. In some cases we have hints: for example the Khalach/Khaladj are from a region north of Kerki on the west bank that has a very high volume of silk production so we can speculatively link use of lots of silk to their weavings. In some cases ( like juvals from the Khojambas region on the East Bank, north of Kerki, inhabited by the Salors) , we do have evidence of ASL knotting, depressed warps and use of " Salor-like " motifs to suggest a connection or speculative attribution. etc etc.

I know this is a long-winded comment, but the point is with regard to this juval: we know from the Persian and Bukharan chronicles that the Saryk came to the Marv Oasis in the second half of the 18th century, and mostly had concentrated there by 1785. But most of our catalogues etc forget to mention where did they come from? Most likely from the Lebab, or perhaps some from the southernmost areas of Khorezm ( Khanate of Khiva) that lies along the river north of the Lebab. And we are told by the Khivan Chronicles (principally the Firdaus al Iqbal) that they were closely associated with a large and important tribal group called the Saqar, who still live on the west bank south of Charjui. So I propose to speculate that there are weavings from this region out there in our collections that are either residual Saryks ( let us call them MAD or Lebab Saryks), who did not migrate to Marv or they could be products of the associated Saqars. But for some reason the Saqars unlike the more familiar Marv Saryks used the ASl knot rather than the symmetric ??
So to summarize, No question in my mind this juval constitutes a separate tribal group, but I have seen and handled only a half dozen of them. So the more the merrier !


Here is my version, a redder pallette and very Saryk style elem, and with yellow and pink silk highlights but I would suggest the same tribal group.
He graciously allowed his piece to be shown; a very handsome piece, and less abused:

C) Moving on to my third piece (images below), let me note a discussion back in 2006, initiated by Jack Williams, featuring an exceptionally similar chuval. Jack hypothesized a Saryk attribution; some other members suggested Ersari at the time.

Here's the link:


Peter's remarks:


On the final one, let me throw out a speculation: an early 20th century weaving of the Lebab region Khojambas Salors. ASL with depressed warps.
Well, that got my attention and I inquired further, and after some additional discussion (not argument) he added:


I am reluctant to accept arguments about attribution only on the basis of design and aesthetics, particularly for 20th century pieces. No doubt there are some very " Saryk-like " features, but then we have trained ourselves to define what is a Saryk based on Russian evidence collected only from Marv, Yolatan and Pende Saryks. And this was after their 100 year interaction sharing the oasis of Marv with the Salors of Marv. Hmm, makes me wonder. And the symmetric versus ASR/ASL technique is a powerful signal, I think.
Now, on to images of the piece in question, followed by some of my remarks and supporting images from the discussion with Peter.

The images; note conspicuous use of silk. Knots are asymmetric open left:








Now some of my remarks to Peter, with images:


CW: I would argue that the crispness of the design execution should count for something, here is mine at left, Ersari center, and Saryk right, from Mackie & Thompson


CW: The trouble with a Saryk attribution is that most (but I think, not all) Saryk work is built with symmetrical knots. If you’re right about my dyrnak gul piece, that’s another argument in favor of non-Ersari. Frankly, I think it is far too nicely done to be Ersari. Snobbish, I know, but I’ve never personally seen an Ersari so well or consistently executed. So Saryk is something I would consider; still, a lot of silk for a Saryk piece. Salor work often has copious amounts of silk.

Additionally, both the pubished Saryk and Ersari analog examples are more cluttered than mine. Salor pieces always seem to have an elegant balance of space and geometrics, as does mine.

Saryk left, Ersari right:

Kizil Ayak (left), Saryk (right):


CW: So I am satisfied with a quite late Salor hypothesis – if “feels” right, but I would love to have enough personal insight to be able to justify it to skeptics.
There seems to be a slowly growing body of evidence and opinion that the Salor were present and weaving in the very early 20th C and that they did not just disappear overnight (sneaking away, like the Baltimore Colts). Rather, they gradually dispersed and redistributed over time.

Chuck Wagner

Chuck Wagner July 6th, 2016 06:06 AM

HI all,

To wrap this up, I'll add a little more content from some ongoing email discussions I've had with Peter about late 19th and 20h century Salor demographics. His remarks:

There is brief mention of Salors in the MAD in the section of William Woods article " Turkmen Ethnohistory " in the Amstey Collection catalogue Vanishing Jewels: Central Asian Tribal Weavings: A Catalog of an Exhibition at the Rochester Museum,1990. See the extensive bibliography by Wood who was a Central Asian History PhD at University of Indiana and now teaches at a university in Southern Ca;ifornia. Wood refers to Salors in Sayat ( on the Left Bank even though Vinnikov makes it clear the Sayat are another tribe).He adds that the number of Salors along the river are twice the number found in Sarakhs, which most rug books show as the center of Salors in the 19th and 20th century. The map that accompanies Vinnikov clearly shows the main concentrations based on the 1926 census as being along the river north of Charjui and south on the Right bank around Khojambas.In the 1926 census after the four Ersari subtribes, the MAD Salors were the largest population of the tribes in the Lebab Region, and that is still true today in Turkmenistan's Lebap province.
As for the Salors of Northern Afghanistan, it is true there are a few mixed in with the Saryks in Maruchak, as well as with some of the Tekke north of Herat ( where they are all just called " Mauri " ) and just over the border in the Pende oasis, but most of the Marv Salors ended up in Sarakhs, fleeing the Tekke invasion in 1830
Chuck Wagner

Unregistered July 7th, 2016 11:40 AM

Open left or Right?
Dear Mr Wagner:

I did look at the closeup picture you posted of your chuval. The one where you are trying to show the type of knot.

Is your chuval woven upside down? Or did the weaving begin at the bottom? If it is woven upside down, top to bottom, the it appears to me to be open to the right. And converse if woven from the bottom to the top then open to the left. Perhaps this question has already be considered in the much comments this discussion has drawn. If so then it is my fault for missing them.

Best wishes to all,

Hans Kraklemeyer

Chuck Wagner July 7th, 2016 01:59 PM

Greetings Hans,

I assume you are referring to the last of the three chuvals.

Terminology: Top and bottom on a vertical loom; far and near, on a horizontal loom.

The term "upside-down weaving" is somewhat misleading. Working on a horizontal loom and starting at the far end, a weaver would have to sit on unknotted warps, making tension management almost impossible. On a vertical loom, starting at the top and working down would cause the newest wefts to tend to fall, and again make tension management very difficult.

So, when we see a bag that appears to have been woven upside-down, a more likely explanation is that we are looking at the second of two bag faces, Starting at the bottom and working upward, the order is typically:

-the first kilim back
-the first bag face
-the middle kilim saddle
-the second bag face
-the second kilim back

(authors note: Duh; wrong; see following posts)

The design on the second bag face is upside-down relative to the first so that when draped over a pack animal, the bags appear the same from both sides. It is a more likely explanation even with chuvals intended to be hung on tent frames; it would be a circumstance where the loom was large enough to accommodate two bags and the weaver had the intent to produce two chuvals on the same warp structure. If you are looking at a bag "woven upside-down" you are looking at the second bag face, likely cut away from the complete weaving.

When weaving, the weaver always pulls the knots tightly toward the completed knot/weft structure - again, because to do otherwise would result in a loose and structurally unsound mess - so the pile points toward the bottom/near end of the weaving.

So, this is definitely open left.

Chuck Wagner

Steve Price July 7th, 2016 02:46 PM

Hi Chuck

I don't know whether chuvals were always, sometimes or never woven in the same sequence as khorjins, but the sequence of weaving a pair of khorjins is:
1. First face (beginning at what will be the upper lip of one bag)
2. First back
3. Bridge
4. Second back
5. Second face (beginning at what will be the bottom of the second bag)

In this sequence, one face is woven from it's open end, the other from its closed end.

Somewhere in our Archives is a discussion of this question. My recollection is that it arose because the numbers of extant chuvals woven from the bottom is very different than the number woven from the top.

Steve Price

Chuck Wagner July 7th, 2016 03:19 PM



I knew I would get something reversed, and you are right.

Serves me right for typing before my first coffee of the day.

Here is an example of what Steve describes:



Chuck Wagner July 7th, 2016 04:24 PM


That said, there are exceptions to that rule, like this Sistani bag which would have to have all the back & bridge kilim work done at one end:


Chuck Wagner

Rich Larkin July 8th, 2016 05:25 PM

There were six men of Indostan, to learning much inclined....
Hi Chuck,

Great posting of all that material referencing, inter alia, the Ali-Eli. Seems like a good time to make my septo-annual citation of that venerable old chestnut "The Six Blind Men and the Elephant," that so efficiently reflects the traditional approach to the study of traditional weaving from the greater Middle East. As you aptly pointed out a while back (as usual, I'm late to this parade), in regard to Peter Poullada's purported take on the distinctions between the work of the Ali Eli and the Kizil Ajak, "...like any analysis, it's based on the data he had, and not on data that he didn't have." What I appreciate about Poullada, though, is that he at least acknowledges the great complexity of the tribal situation among the Turkomans in particular in any given period. Indeed, it has to be speculative how much the putative weaving practices from an identifiable group in one period can be used in the analysis of material from a different period; and whether it is legitimate to limit the candidates for a handful of more or less similar weavings (e. g., the juvals above) to, say, the Ali Eli and the Kizil Ajak. How about those folks in the various published lists that most of us have never heard of? And where were they in 1830? Or 1873? Or whenever? And what did their weavings look like? I'm guessing we don't know.

I don't advocate ignoring what we do know. Bravo to Poullada! :cheers: Just keep in mind the analyses only take us so far, and must be considered tentative, a fact Poullada understands as well as anyone.


Chuck Wagner July 9th, 2016 02:50 AM

Hi Rich,

Well noted; all attributions based on supposition are inherently uncertain. While little can be "known" about Turkmen pieces outside those items properly documented at the time of original accession, I think that studied inference is worth a lot. And as you note, Peter has put some serious effort into his work, and he has not allowed "mission creep" to distract him away from his target demographic. From a separate email discussion on a different piece I own:

I will have to go back and check my notes on Erik Risman's and my comprehensive knot count calculations for all known, cataloged " Ersaris" and KA.
In the case of Ali Eli bags, the fact that David's piece meets Peter's Ali Eli structural and design criteria AND also has the very rare hash lines on the sides - allows a tenuous link to be made between that bag and the Ali Eli pieces in Mackie and Thompson (and Dave and my bags). It is an indirect link, but the feature is sufficiently rare to be notable.

In the case of my last piece, I am satisfied with his suggestion of a possible turn-of-the-century Amu Darya centric Salor attribution. Witness this image (the camera strap is 1/4 inch wide):


While anything is theoretically possible, I would be interested to see someone produce a known Ersari, Kizil Ayak, or Saryk piece with a back that looks like that, with a lot of silk, and with a maximally depressed AsymL knot.

I have two Tekke pieces that have a similar appearance due to the fine un-depressed knotting (the chuval is more than 400 knots / in sq). But I have never seen an Ersari piece that looks anything remotely like this.

Not being argumentative, your point is well taken. But also, these are not SWAGs.

Chuck Wagner

Rich Larkin July 9th, 2016 04:28 AM

Hi Chuck,

Of course, it isn't our fault we can't identify the weavings of the Uch El, or the Adaqli, or the Tivechi, if they wove anything. We could probably do some mischief by planting a statement someplace saying their rugs were indistinguishable from those of the Ali Eli. Seriously, I have no doubt Poullada's work establishes a solid basis for identifying a group of weavings with which the Ali Eli were thoroughly involved. It's just that the rug appreciating community seems regularly to forget that the reality of provenance must be more complicated than our familiar and, in truth, simplistic models; though, as I mentioned, Poullada himself is largely not guilty in these regards in my opinion.

I've mentioned before that I heard Murray Eiland, Jr., M. D., say at the Textile Museum (quite a number of years ago) that he looked everywhere around Central Asia, unsuccessfully, for someone who would admit to being Ersari. He said people whom he asked didn't seem to understand the question. I suspect if many weavers could be clued into the vocabulary we use regulary to analyze and discuss rugs, there would be a good deal of head scratching. Not that it's such a bad thing. :confused:


Chuck Wagner July 9th, 2016 03:22 PM


Interesting remark from Eiland, I wouldn't have guessed that from his writings. Dick Parsons lived in Afghanistan for many years and didn't seem to have any trouble finding people who identified as Ersari. His book includes his own version of the clan and sub-clan structure. Even so, I would posit that people would be more inclined to identify at the clan and sub-clan level first, and maybe that's what Eiland encountered ...(just guessing).

The Russians don't seem to have that problem either; there are recent books and articles available online that discuss staffing choices and power sharing in Turkmenistani government structures based on tribal affiliations and inter-tribal rivalry, including passing over one individual (General Akmurad Rejepov) who was Ersari and considered unacceptable to the Ahkal Tekke elders.

It is easy to think that the number of persons who keep any interest in their own tribal affiliations has been dwindling rapidly for decades, thanks in particular to the Soviet Union's shake-n-bake approach to population redistribution.

Nevertheless, your larger point is still taken; there are many things we cannot know - after all, there is no formal written Turkoman historical record. In the meantime we organize by what we do know, or hypothesize, and then engage in interesting discussions... :cheers:

Chuck Wagner

Rich Larkin July 9th, 2016 04:24 PM

Hi Chuck,

Your comments about the reported invisibility of the Ersari reflect my own thoughts at the time when I heard Dr. Eiland on the subject. I believe the thing boiled down as you suggest to sub-tribes and clans as having been the more familiar context for his informants. They didn't think of themselves as Ersari, but rather, as XXXX. Also, the occasionally encountered experience in the Middle East of not getting a straight answer to a question may have been a factor. Parsons' extensive experience in Afghanistan probably served him well.


Marla Mallett July 11th, 2016 12:08 AM

Knots open left or right?
Iíve just now tuned in to this discussion and have seen the confusion concerning the construction of asymmetrical knots open left or right. David Hunt has pointed out what he sees as a discrepancy between the diagrams in my WOVEN STRUCTURES book and those in Murray Eliandís book.

Actually the constructions illustrated in the two publications are exactly the same. The yarns are wrapped and the knots formed in precisely the same way. Murray has simply drawn his knots with the loose yarn ends pointing upward in an unnatural manner. That can be confusing. Iíve drawn the yarn ends pointing downward as they look to a weaver who must pull them down tightly toward the already woven part of the rug. Thatís the way Tanavoliís are drawn as well.

In any case, as someone has mentioned, the knots should be described as ďopenĒ on the side in which the yarn ends emerge. It is of course necessary to have the rug oriented as it was on the loom when examining the structure.

Marla Mallett

Patrick Weiler July 13th, 2016 05:15 AM


You hit the nail on the head with "It is of course necessary to have the rug oriented as it was on the loom when examining the structure." Often what is remaining of a bag face does not include enough to confirm which end was up or down when it was on the loom. Asymmetric open left is asymmetric open right when a piece is turned upside down. The pull of the knot pile can often be the arbiter.

Patrick Weiler

Dinie Gootjes July 16th, 2016 02:58 AM

Patrick, I think you can always find the knot collar above the tufts of yarn? Especially clear where two colours meet vertically. That would indicate that you are looking at the rug the right way. I remember one rug where the pile was lying two ways in different parts of the rug. The knot collars made it possible to decide how it was woven, and so what type the knots were.


David R E Hunt July 17th, 2016 01:15 AM

Hi Marla

Thanks for helping out us structurally challenged folks : (.

But my issue with this was more at metaphysical. I had somehow seemed to recalled an early discussion,
in one of the old rug books, in which the "open" portion of the terms "Open Left" or "Open Right" referred
to the space or grove in the pile. Thus in "Open Left" the pile emerges on the right side of the knot, and in
"Open Right" the reverse is true. Not to worry, it's just me. I'm confused and a bit out of the loop on this.

Note to self: if you ever do any serious publishing, it's a good idea to have the structural analysis done by
a qualified professional...


All times are GMT +1. The time now is 01:18 PM.

Powered by vBulletin® Version 3.8.10
Copyright ©2000 - 2017, vBulletin Solutions, Inc.