“Potpourri” with Mark Keshishian

On November 3, 2012, Mark Keshishian

gave a “potpourri” session at The Textile Museum, here in Washington, D. C.

Mark is active in the family rug business, is a accredited appraiser, and used to be successfully active in competitive  “strong man” events.

He had brought several pieces, some from his family’s collection, and he started with these.


Mark said that this is a fragment of a Caucasian rug woven in Karabagh in the late 17th century.  It is from a rug of the “dragon” type that is long and relatively narrow.  He said that there are a number of such fragments in the family collection.

Here are some detail images of P1.

Mark’s second piece was this one.


This rug has a field dominated by Lesghi stars.  It was published in Harold Keshishian’s (Mark’s uncle) book.  It is attributed to Armenia in 1903.

Here are some detail images of P2.

The next piece was the one below.


Here is the most comprehensive image I can manage.

This is, of course, what is popularly known as a “cloud band” or  “Khondzoresk” Kazak.  Mark said that “Khondzoresk” translates to the worm in the apple.”

As the detail below shows, this rug has some anatomically correct human figures in its designs and is dated 1833.  It is thought to have been woven in Armenia and was likely published, Mark thought, in his father’s book on Armenian rugs.

Here are close-ups of two of its medallions.

The next piece was the one below.


This piece is a contemporary rendition of a familiar Kazak type (Fachralo Kazak design), woven in Turkey by Woven Legends.  Its dyes are alleged to be from natural sources.  Mark said this piece just needs to get “a little dirty” to look better.

Some detail images of P4.

Mark’s fifth piece was another contemporary one, this time a Bijar replica also by Woven Legends


Details of P5.

With the next rug, below, Mark went back to older material.


This mat is a “Manchester” Kashan, with a design seen with fair frequency here in the U.S. for obvious reasons.  A finely woven rug with a good range of color and precise drawing.  The pinkish ground of its field is somewhat unusual.

Mark said this is one of his favorite rugs.

Details of P6.

The next piece was another Manchester Kashan, this time with a niche and “tree of life” design.


It has a wider color palette than an initial look at it suggests.

Here are some closer detail images of P7.

Mark drew attention to the back of this piece, noting its lighter colors. 

He said that more definite reds, especially wine-reds, were preferred, in the U.S., to these light shades, and so the market arranged for such rugs to be “painted” in more desirable ones.

Mark said that such painting started in the early 1920s and ended by the mid 1930s.

The next piece was this one.


My notes indicate that this is an Indian rug done with “Manchester” type wool.  Mark said  that the front of  this piece looks different from a typical Manchester Kashan, but that the wool feels better than the Persian one.

Details of P8.

Here is a closer detail of the back of P8.

The next piece was a Ferahan Sarouk.


Some details of P9.

The next piece was another Ferahan Sarouk.


There was talk in the room about,  since the warps of this piece are parallel to its short side, whether it could have been the face of a chuval.  The consensus was that while there are, on rare occasion, pile chuvals in NW Persia (some Bijar instances were remembered) no one knew of a Ferahan Sarouk example. These pieces tended to be seen in pairs, with the pair to this piece actually now owned by a member in the audience.

Some details of P10.

The next rug had a very different look.


Mark said that this rug is a contemporary piece, woven in Turkey with Ushak designs.

Wool quality is great.

Here are some closer detail images of P11.

The next rug was the one below.


This is a silk Kashan, woven, just before or after WWI.

Mark used this rug to talk about some distinctive characteristics of silk, particularly with regard to washing it. 

He said that silk doesn’t have a “memory,” like wool does, that lets its individual fibers return, after washing, to their previous, “separated” posture.  That is, when washed, silk’s fibers often tend to stick together. 

He said, that ideally one should avoid washing a silk rug, but that there are some solutions (not available at the retail level) that can help prevent this “sticking together” tendency.  It is also necessary to comb a washed silk rug frequently until it is dry.

Here are some closer detail images of P12.

The next piece was the small, Tabriz mat below.


Here are some details of P13.

The next piece was Caucasian


This ivory-ground rug is a small Kazak.  Its design is described by Bennett as a hooked polygon.  He groups it with some “miscellaneous” Kazaks of no great age.  Its ivory ground is attractive and its restricted color palette is graphically effective.

Here are some detail images of P14.

Mark called attention to the nice, enriching, touch of “barber pole” centers of the “leaf” forms.

The next piece was another silk rug.  Its owner has given me a detailed description of it after the fact. 

Modern Qum.  Silk pile on silk warp and weft.  Knot is asymmetric open on the right.  Relatively finely woven (600+ kpi).  Signed top and bottom.  Lobed medallion with center flower, surrounded by red carnations, on bed of leaves, with outer row of flowers.  More flowers in lobes, with leaves in field.  Flowers surrounding center field. Asymmetric borders, with flowers in cartouches in main border.  (There are good detail images on all of this below.)


Mark said that such Persian silk rugs are “drying up” in the market, due to the U.S. trade sanctions aimed at Iran. 

He estimated that it could be as much as 25 years old.  Its owner volunteered from the audience that it was woven quite recently.

Here are some detail images of P15.

Mark attributed the next rug, another “tree of life” design, to Kerman.


Asked whether this rug has the distinctive, three-weft between rows of knots, Kerman structure, Mark said that he bases his attributions mostly on such things as design, color palette and character of the wool, and doesn’t usually delve deeply into structure.

He said this size is called a “meditation” carpet.

Mark said that he wondered whether this piece might not have been painted.

I do not have a comprehensive, unobstructed shot, but here are some details of P15.

The next rug was Caucasian.


Its central medallion could be an abstracted version of some seen in rugs woven in Karabagh. 

Its owner said that it was purchased in Baku.  He said that this is an old rug, and was woven in the 1880’s.

Its border was seen to be unusual.

Its owner said that this piece had at one time been cut up into four quarters, but was subsequently restored in Istanbul.  Mark, examined it, more closely and noted that the ends were too perfect for the wear this rug would have had had it been used on the floor.

The next rug was this one.


Its colors were said to suggest that it is an Azeri weaving.  Its owner said that it was purchased in Istanbul.

Here are some details of P17.

Design devices like the one below, are of the sort that invite fanciful names, like “space-shuttle.”

The are six such devices in the rug.  The four like this one are oriented at the side.  I have turned it 45 degrees here, just to cause trouble.

The next rug has the look of a classic “Seichur.”


As you can see, it features a great green an unusual coloration for this type.

This rug has the classic “Seichur” running dog border, but its owner also drew attention to its inner “cabbage” border which he said is a Kuba usage.

The field of this rug features three so-called “St. Andrew’s cross” devices.

Its owner estimated its age as 1880-90’s.  Nice corrosion.

Here are some additional detail images of P17.

Its owner said that its structure is closer to other Kuba pieces, since its alternate warps are depressed.  Shirvans usually have warps on a single level.

The next piece was another Caucasian, this time with a white ground and monumental-sized field devices.



The owner of this rug, collects mostly eastern Caucasian rugs and so this piece seems likely to be from that area. 

The borders and collect seem similar to Shirvan usages, but the white ground of the field and the large scale of the devices in it, are unusual.

Here are some detail images of P19.

The next rug was Anatolian.


This rug has good color and a nice complexity of design.  It is estimated to have been woven about 1875.

Its owner said that when we think of Anatolian rugs woven in Milas, we tend to think of “niche” designs, but this Milas is not of that sort.

Here are some details of this piece.

This rug has a good purple.

The next rug was another woven with silk.


Its owner said this it was woven in the Kayseri area of Anatolia.  He said that it was woven properly for a silk rug, that is, with the pile pointing toward the niche-end, but that this makes it “upside-down” when hung as most prefer the the pile pointing down.

This rug has cross-panels, a design usage rare in Persian rugs, but that we see in Turkmen ensis and a few varieties of Anatolian rugs.  They occur with some frequency in the once, very desirable, but now largely neglected, Ghiodes rugs.

Here are some details of P21.

The next piece was a compartmented design.


It owner said that the dealer called it “Baluch,” but that others have said it’s “Middle Amu Dyra.” 

Mark said that is an unusual and interesting piece.

It has a wider range of color in the lower parts of its field, as if the weaver had run out of some colors.

Here are some closer details of P22.

Here is a look at the back of this piece.

The next piece was this bag, attributed to the Luri-Bakhtiari .


This khojin half has some interesting features that are worthy of closer examination.

The motif on the front face of the bag clearly comes down from the device that appears on the famous Berlin phoenix-dragon rug.


Courtesy Peter Stone

This motif appears in both Persian and Anatolian weaving.  There is a comment on it on p.75 of John Wertime’s sumak bag book. 

One question that remains is whether the motif on this bag (below) is the dragon or the phoenix?  To me, the top horizontal strip of the Berlin device is the more plausible candidate for the halves from which the “gul-like” device,” below, is formed.  Maybe it’s the phoenix being used.

As a former Turkmen collector, my initial reaction was less elevated, and based on what I “see.”  The field of this piece contains six devices that seem like “fractured” guls.

Their horizontal division, is “clean” enough, although the shapes of the top and bottom halves are such a half rotated 180 degrees.  The result is that the outside perimeter outline of the comprehensive “gul” form seems slightly askew. 

This “broken” character is strongly reinforced by jagged vertical division of the two halves. 

This drawing (and a rich use of attractive colors) heightens the aesthetic interest this piece projects.  If, as seems to be the case, these field devices are not gul forms at all, but rather come down from the “dragon and phoenix” rug in Berlin, so much the better.

There are also quadruped devices in the field of this bag’s front,

and in one stripe on its back there are camels.

Here are a couple more details of this interesting piece.

An attractive yellow-ground pile panel at the lower edge of the bag’s front, a frequent Bakhtiari usage.

The next piece was Baluch.  An unusual camel-ground design.


Its owner said that it contains lots of actual camel hair.

Some additional detail images of P24.

Mark looked at its back.

A little closer look for you, too.

Aha! Quadrupeds in this one, too.

The next piece seemed a “face” of some sort.


Mark said that it looked to him like something is missing.  He praised its colors, especially its good green.

I wondered if this could be a fragment of an Anatolian yastik. Yastiks are almost always rectangular and would be taller in this orientation.  If this is a yastik fragment , what might be missing are lappets on both ends.

Morehouse shows at least six Central Anatolian yastiks with a center medallion surrounded by four “armature” forms.  Oddly, the one  with the most similar field design to this one has cross-panels at its ends, but no lappets.

As luck would have it, I’ve been working on a couple of posts, going back and forth and by chance found an instance that will let us test our suspicion that P25,  above, could be a yastik fragment.

This is an older yastik in very poor condition, but with a very similar field.

If the lappets at the ends of the piece above were removed it would look like the truncated image of it below right.



P25 looks very much like a former yastik that has lost its lappets.

Here are some closer detail images of P25.

The next rug was a long, classical Talish.


This rug has great color and the classic Talish, main border.

Talish rugs, its owner said, are woven in the Southeast Azerbaijan part of the Caucasus.

Fields in such Talish long rugs are often plain.  This one has an attractive, colorful drawing of a lattice design, with stars in compartments.

Some Talish rugs have blue wefts, visible on their backs, moving in occasionally from the side selveges.  Some claim that this feature is definitive for Talish pieces.  Below is a look at the back of this piece.  It has these interpenetrating blue wefts.

The next piece was a flat weave in slit tapestry.


It was attributed by its owner to Karabaugh.  He said that it had been cut in half and has now been sewn back together.

Here are details of P27.

The next textile was another Caucasian flat weave: this time a Shirvan kilim.


Mark said that this piece has great color, great wool and a nice, tight weave.  He said that late 19th century would be a safe age estimate.

Here are some closer details of P228.

The next piece was an “envelope” bag of some sort, complete with a seeming triangular flap.


My notes say only that this is an “end panel” from a Bakhtiari piece, presumably a cargo-bag type mafrash.

Here are some closer details of P29.

The next piece was this pile panel with good graphic punch.


This is one front panel from an Anatolian saddle bag set, Mark said, from the Bergama area.  They are called “heybes.”

 This seems a frequent heybe face design.  I found one in a local flea market.  But it may, also, be a quite old design.  This is what the front of a complete heybe with this field design, likely looked like, if it was reinforced with leather (some were).

Because the back of this analogous piece was likely covered with leather, we do not have an image of it.

But perusal of the only book on heybes (by Bieber, Pinkwart and Steiner) allows the conjecture that its back might look like this.

The striped backs of heybes vary widely, and to some extent, systematically.  They are, more than the fronts, used for determining attribution.

Here are two details of P30.

The next piece was a small Turkman rug.


This piece is the size often called “wedding rugs,” although this one features “chuval” guls, rather than the more frequent small version of the main carpet gul.

The owner called attention to the variation in the size of the guls.  This is something seen in more than a few Turkmen pieces.  It has led to the fanciful suggestion that such size differences are often deliberate and meant to represent birds in swooping flight.

Its owner said that this rug is Tekke.  I didn’t look at it closely, but the main border made me wonder whether it might not be Yomut.

Its owner said that its “chalky” white is from the use of cotton as pile.

Here are some more detail images of P31.

The next piece was a small Anatolian kilim.


It is in slit tapestry and was attributed to the Konya area.

Its green was admired.

Here are some additional details of P32.

The next piece was one I had brought.  It is an Uzbek needlepoint in silk


Mark as whether I knew how old it is and I jokingly said that I thought you “could look at your watch,” that such things are still being made.

Seemingly a little suspicious of my joke, Mark looked around on this piece a bit, and found that it is dated, something I had not noticed at all.

So it’s a little less “fresh” than I would have guessed.

Here are two more details of it.

Although is just a function of the fine silk and the digital needlepoint technique, I like the crispness of the drawing, especially of the Memling guls.

The next piece was a “double-niche” Turkish rug in silk.


Notice the architectural features on the sides of the niches that become floral at the top.  The details of the lamp designs are articulated, made possible by the fineness of the silk.

Here are some additional detail images of P34.

The next rug was a small Malayer.

Mark said that it has a really soft handle for a Malayer.

He noted that it is single-wefted.  There is just one pick of weft between rows of knots (warps are visible whenever the weft passes beneath them) and the wefts are “fat,” and make the warps seem to “undulate.”

Here are a couple more detail images of P35.

The next rug was a fairly large Kazak with three over-sized “Lesghi guls” dominating its field.


Mark noted that there is an orange used, and said that this is unusual for a Kazak.

He admired the greens.

He said that the weft is medium brown and could suggest that it is Karabagh (ed.:although Tschebull found tans, and even dark brown warps in some of the pieces in his catalog “Kazak”).

Here are a couple of additional details of P36.

The next rug was Chinese.


Mark placed it just after WWI.

My notes include an indication that some thought this rug was likely woven by non-Hans in China.  They said something about “Persian” women weaving in the north of China, where the Great Wall meets Mongolia.

Here are some details of P37.

The next piece was the kilim below.


It owners said this piece was woven in either northwest Bulgaria (Chiprovtsy) or southeast Serbia (Pirot).

It has eccentric wefts that follow the curves of it designs.  In the market, these are all called Sarkoy kilims.

Has a border executed in the Serbian style.. 

Recently purchased in Istanbul.

The field is primarily very dark blue green with some small areas of blue. The purple red in this example is particularly nice and may be cochineal or Balkan kermes.

Mid-to late 19th Century. Owner: “We aren’t certain of the origin of the motifs covering the field-perhaps they are a local version of Ottoman floral style.

“Depending on its age, the kilim may have been woven by Orthodox Christian folk or by Orthodox Christians and Muslims living and working side-by-side.  

“Everything changed in this region after about 1868.

Here are two more details of P38.

The next piece was is a Manastir prayer-format kilim from northeast Bulgaria, likely from the Delioman region.


Its owner said that the ivory ground is unusual.

End borders are typical for a particular sub-group of Manastir kilims. 

Attention was also called to the pronounced use of outlining on the sides.  The width of the outlining is varied to produce a three-dimensional effect.

Third quarter, 19th century.

Its owner said Manastir kilims were frequently woven by the women of heterodox (Alevi or Bektashi) Muslim families in Bulgaria for domestic religious use, to provide a protective function against bad luck,  and to serve a devotional function.

There is some debate as to how many of these textiles were actually used for prayer.  However, Ottoman records for the 19th Century in Bulgaria document the practice of requiring Muslim conscripts to the Turkish Army to bring their own “prayer” kilim.  Some Manastir kilims have features that document this practice.

Here are some details of P39.

The next piece was large and delicate.


Its owner said that this piece was woven in Syria in or near the city of Aleppo. 

It may have originally been a wall, window or door hanging (a portiere) which was subsequently been repurposed as a bed or table cover.

The piece consists of two panels sewn together longitudinally. 

Although the piece is relatively delicate, it is also quite heavy due to the use of a large amount of metallic wrapped thread.

This cover dates before 1910 and was reported to have been collected in Adana, Turkey.

Its design reflect a typically Syrian interpretation of the ottoman floral style with lavish use of metallic thread: fine metal wire wrapped around the yellow silk fibers.  This technique results in the textile having a shimmering gold tone and texture

Here are some closer details of P40 and its decorations.

I have rotated the devices below to give you larger close-ups.

The next piece was the one below.


This piece is fragmented in the sense that all of its edges have been bound.

It owner said that it was woven by her grandmother in an Armenian area of Anatolia.  The grandmother wove three of these.  The was a suggestion that it could have been a yastik (it is yastik-size).

Here are two details of P41.

The next piece was a framed Chinese textile.


This is embroidery on silk.

Metallic thread used.

Needle embroidery rather than brocade.

Here are some closer details of P42.

Made in two panels, then sewn together.

The next piece was a familiar type Yomut Turkmen flat woven bag.


I own one that it nearly identical and have a story to tell about mine. 

I bought it from a Selcuk dealer in 2007, in partial recognition of an exception job of hosting he did for us during a several day stay in his town. 

But the interesting thing is that he claimed it was a specialized type of bag and was used to hold seed as planting occurred.  I was skeptical, but he persisted, saying that he had seen a photograph of a Turkman farmer using this kind of bag in this way.

I didn’t debate, but Turkmen were very practical folks, and many bags could be, and likely were, used for a variety of purposes.  This bag is of a size that would suggest that it was not strictly reserved for use in planting.

Here is one more detailsof P43.

The bands with designs on this piece are woven in sumac.  Turkmen uses of sumac are relatively rare, and so noteworthy.

The next piece was a small Anatolian bag in species of brocade.


A relatively coarse, but attractive piece. 

Here are two closer details of it. 

I did not manage a photo of its back.

A potpourri RTAM will always attract at least one recent Turkman piece, and this was the one for today.


Mark said that this is Tekke Turkmen Soviet era production.

I have heard that the white ground stripes at the ends of this piece are nearly a signature indicator of Soviet production.

The last piece of the day’s session was was large and heavy.


It is Qashqa’i.  Southwest Persia.

Here are some detail images of P46.

Odd, “chickens” in the center of the field.

Dark warps.

Here is its back.

Wefts seem less dark than the warps.

Mark took questions,

and adjourned the session.

After-session conversations and explorations started up.

I hope you have enjoyed this virtual version of an, admittedly, long, “potpourri” RTAM.

My thanks to Mark for permitting this virtual version to be produced, and for his editorial work on my draft.

‘Til next time,

R. John Howe

Comments are closed.

%d bloggers like this: