Archive for September, 2010

David Zahirpour: Collector’s Potpourri

Posted in Uncategorized on September 15, 2010 by rjohn

On August 28, 2010, David Zahirpour

gave a Rug and Textile Appreciation Morning program on the topic “Collector’s Potpourri”  here at The Textile Museum in Washington, D. C.

David’s version of this frequent type session invited participants to bring Caucasian, Northwest Persian and South Persian rugs and textiles for comparison and discussion.  David brought a number of pieces of this sort himself.

As usual, I arrived early to claim my front row seat.  David was just hanging the first of his pieces

a gabbeh.

In awhile, David’s array was in place,

and he was ready for the audience.

Potpourri-type sessions draw an audience, but also lots of material.  And this session was no exception.

We could see,  before David started, that this session would run past our usual target ending time of 12 noon…

and we did.

Tom Goehner,

the TM’s Curator of Education, introduced David, saying

that David was a local oriental rug and textile dealer with a long relationship with the TM.  He often does RTAM programs and has been active for years in providing rug repair demonstrations in various TM programs.

David said that he would follow a general geographic sequence, beginning with the Caucasus.

He started with a Caucasian pile rug of a familiar type.

This is a Shirvan rug, from the southeastern Caucasian province of that name, with a “prayer” or “niche” design, although such rugs were not necessarily used in prayer.  David estimated it to have been woven about 1880.

The lattice field design is described as “Marsali,” and has a good range of color.  David called attention to the fact that the primary device used in its niche border

is an abstracted version of the “boteh” design.

Here, for comparison, is a more curvilinear boteh progression from the writings of John Irving,

as well as a, mostly, more abstracted array of Persian boteh forms by Peter Stone.

David said that Shirvans were one of the more numerous Caucasian pile varieties woven, another being the Kazaks, from the southwest.

He had brought as his Kazak example, a more recent piece, also in a niche format.  He said that there was a huge production of Kazak rugs and that many of them tend to be among the largest sizes produced in the Caucasus

David joked, but also not quite, that this could seen to be a rather pedagogical prayer rug, providing the novice Muslim explicit indications about how to arrange himself or herself for prayer.

The “hand” images indicate where the hands are to go, and the square inside the mirhab indicates where the head is to touch.

Even the horizontal lines moving out from the mirhab can be seen to indicate where the shoulders should be, while the worshiper is bowing in prayer.

David’s next Caucasian piece was the large Shirvin kilim below.

Many kilims were used with the long side on the horizontal, but the rug books often show them with the long side on the vertical, so here is this piece in the latter orientation.

David said that the weave of this piece was “slit kilim” (some use the term “slit tapestry”).  The wefts move over and under alternate wefts,

turning back at vertical color changes.  This turning creates the slits.

David says that he, personally, reserves the term “kilim” for pieces that are woven entirely of this over-under structure.

He called attention to the fact that there are no separate “structural” wefts in textiles woven entirely with a kilim weave.

The strands (in the case of slit kilim: wefts) that create the pattern are also part of the basic structural level of the piece.

He contrasted this Shirvan kilim with a second piece,

to which he said he would not apply the term “kilim,” since it is woven in part in a sumac weave.

The plain-colored, horizontal striped areas are “kilim” (i.e. “weft-faced tapestry) because the structure in those areas is a matter of wefts moving over and under alternate warps.

But, he said, it is not correct to refer to the overall piece as “kilim” because of the narrow bands of patterned sumac.

Remembering possible less experienced members of his audience, he distinguished “sumac” from “kilim.”

In kilim, he said, as he had noted with the first piece, the wefts move over and under alternate warps and there are no “structural” wefts.

“Sumac” is different in several respects.

First, the wefts that create the designs in sumac move over more than one warp before moving under one.

(In the image of sumac weave in the drawing below, the weft moves over four wefts and then back under two.)

More, when this move “under” is made, it is a movement back that wraps the second of the two warps just moved over.  This feature is worth being made explicit.  The decorating wefts in sumac do not “interlace” the warps, but instead “wrap” them.

And although there is a structure called “weftless sumac,”

the most usual variety of sumac has, not just wrapping wefts that create the designs, but also (in between the rows of the latter) separate “structural” wefts.

“Kilims” lack these structural wefts.

The structural wefts, in the varieties of sumac that have them, are analogous to those in between the rows of knots in pile rugs.

Note:  David’s “kilim” vs “sumac” distinction is similar to Parviz Tanavoli’s contrast of “gelim” vs “palas” in the latter’s book “Persian Flatweaves. ” Although Tanavoli more explicitly acknowledges varieties of “tapestries,” on the “gelim” side, and does not consider the presence of some structures other than tapestry (for example, “brocade”) disqualifying for a “gelim” classification, he does seem to place “extra weft wrappings” (which appears to be his term for “sumac”) firmly on the “palas” side.

David’s next piece was described as likely Caucasian and entirely in sumac.

He said that this piece was too small to be a side panel of a mafrash transport bag, and was instead the face of an “envelope” type bag that would have had a kilim back.

He called attention to the larger designs in the central band of the field of this piece,

which he described as  “crab” devices.

Someone in the audience asked about the weaver’s rendition of, what is sometimes informally referred to as the “bird-on-a-pole” design, that serves as the main border for this piece.

It does appear that this weaver had difficulty rotating this device for use as the mail borders on the ends of the field.

But I have just been reading a conversation in Turkotek’s archives about similar occurrences with this same device.

There, it was noted that it seems that weavers very often “do something” other than a strict turning of a perfect rendition of this device, 90 degrees in orientation, AND that even folks as careful about analysis as Marla Mallett, are often unconvinced that the seeming failures to render the device perfectly are a sign of weaver inability or error.

And on a different tack, notice that the “bird” forms have what seem like “Greek crosses” at their center.

Sometimes folks see this usage as a sign that the weaver may have been Armenian, but this “bird-on-a-pole” border was widely used, even by Turkmen weavers.  There the “cruciform” device is abstracted to small sets of  squares.

It is true that the broad Islamic design tradition was often purely “geometric,” without any intent toward “representation, and geometry alone could account for what we experience as “cruciform” devices of various sorts.  Still, the presence of these seeming “Greek crosses” in textiles often likely woven by Islamic weavers, does draw one’s eye.

Someone in the audience had brought in a quite small sumac bag that David treated next in comparison.

He admired this little piece, saying that it was part of a pair.  He drew attention to the unusual “arrowhead” or “chevron” borders at its top and bottom.

He said that the sumac is very fine, despite the fact that the plain-woven back seems coarser.

Notice that the single “birds” on the side borders are rendered more successfully here, and that the “Greek cross” devices are even more visible despite the much smaller size of the bag face itself.

It is interesting to see that, sometimes, when weavers use devices on smaller formats, they do not reduce the device size proportionally.

David moved next to another pile Caucasian rug.  This rug does not have great age.  David estimated that it was woven about 1900.  Its color palette is narrow and of pale colors.  There is transfer of the pinkish red to some parts of the warps.  It seems likely that some of the dyes in this rug are “chrome” or of other synthetic varieties.

David described this piece as a Kuba, with a weave that is more city-like and designs that are geometric, but unlike those of other Caucasians.

He described the designs on this rug as of the “dragon” variety.

The notion of Caucasian “dragon” carpets has been a focus of much debate and disagreement among serious rug students.  Among the things contested  are: 1) what their defining characterics are, and 2) where they were woven.

A few years ago I heard Paul Ramsey, the Denver rug dealer and serious rug and textile student, give a paper on Caucasian “dragon” carpets in which he indicated (although this was not the main thrust of his argument) that most rugs seen to be of the “dragon” variety have lattice designs and that there are “dragon” carpets without “dragons.”

Here are three areas of this rug in which there may be “dragon” forms visible.

It may be that one or more of these examples can be read as abstracted “dragon” forms, but it may still be permissable to describe this rug as of the “dragon” variety, if the three design elements above turn out to be most appropriately read as “lattices.”

Here, again, is a comprehensive image of this rug.

It seems to me, looking at the image above, that the “lattice” reading is plausible.

Regardless, it is an odd field design for a Caucasian rug.

The next piece was a Caucasian pile rug with a lot more color.

The piece, brought in by someone in the audience, was described in the room as possibly Seychour.

Here are some closer details.

The “cruciform-like” character of the larger field medallions was noted.

Someone said that the color palette looked like that of some Bijov rugs.

The consensus was that it was definitely east Caucasus and a very attractive piece.

Although he did not treat it extensively, David next called attention to a khorjin face with a diagonal design that was hanging from one corner on the board.

This design is seen in both sumac woven bags and in pile rugs and is usually attributed to Karabagh.

I did not examine it closely, but think it is a sumac khorjin face.

David next treated a piece someone had brought in.  It was half of a sumac khorjin.

David said that it was very classic and called attention to its having another version of the “bird” border we saw earlier.

He said that the small floral border surrounding the device in the center of this piece makes him call it “northwest Persian,” rather than “Shahsavan.”

David’s next piece was a complete khorjin that he had opened up for pedagogical purposes.

He demonstrated how the two more decorated ends would be folded toward the center,

to form a complete khorjin set of bags from a single woven strip.

Below is a closer look at on of these khorjin faces.

Although nicely composed, David said this piece exhibits some of the irregularities that are the result of the fact that this weaver worked without close reference to a cartoon.

With his next piece David moved to south Persia.

This is a Bakhtiari salt bag with some nice reds.

David noted that multiple techniques were employed to make it, including reverse sumak, embroidry and pile.

He called attention to the location of the pile strip at the bottom,

saying that loaded bags were often dragged and the pile worked to reduce wear in that area.

He then turned it completely over to reveal the very different flatwoven design on its back.

Next, David treated two Kurd bag faces.

He said the one pinned to the board (behind him in this photo) was woven by Jaff Kurds and the one under his arm by other Kurdish weavers.

He said that the Jaff Kurd piece was coarser and that this coarseness was the reason why many Jaff Kurds have offset knotting.

That the larger distance created in the coarser weave required the use of offset knotting (in which pile knots move horizontally, from one row to the next above it, by only one warp, rather than the usual two of ordinary pile weaving).

He used the Jaff Kurd bag face as an accessible example of the use of offset knotting to make diagonals in designs, steeper.

The image above is the only one I have of the piece under David’s arm.  I have not much better additional images of the one hanging by a corner on the board behind him.

Here is the most comprehensive image of the one on the board.

And here is a closer detail of it.

David passed this piece in the audience, so folks could examine the offset knotting up close, and I took it once more in my lap as it passed.

This piece did not have the range of color that some Jaff Kurd bag faces exhibit.  Perhaps its most noteworthy feature was the white ground of the central diamond.

Next David moved to treat a Senneh piece.

He said that this Senneh is entirely in a kilim weave and in perfect condition.

He said that weaving is very fine and tight.  Herati, floral and boteh designs are used.

The dyes are natural and chrome.

David’s next piece was also a Senneh, this time an altered horse blanket.  Below, is an overall view of this piece (ignore the band in front of it; that is a separate piece just placed for momentary convenience).

Good color.

Structure, again, is kilim.

David said the wool is wonderful.

Stripes in design are heavily instrumented.

Lovely, despite the alteration.

David now moved to South Persian pieces.

He began by rehearsing some of the major designations and groups used to distinguish South Persian rugs and textiles.

Shiraz, the major marketing center, that includes some workshops; the Afshar, Qashqa’i, Khamseh, Luri tribal references; Kashkuli, a designation that some see as a tribal reference and others see, mostly, as an indication that a given piece is from the best of South Persian production.

He began with a gabbeh.

Here is a comprehensive look at this rug.

David said that this piece was largely woven from undyed wool.  The black, white and grey are all undyed.

Goat hair was used for warps in this piece.

Gabbehs are still being woven, but the modern production has a very different structure.  David said that this piece was likely woven about 1940, before the modern production began.

He pointed to an odd feature in this piece.

There are occasional small areas of naturally dyed red and/or blue, sometimes visible,

sometimes partly buried, and sometimes completely so.  They may have been added as “nazarlik,” protections against supernatural evil.

David dealt next with the tent band he had brought.

As you can see this is a lengthy piece.

David said that this piece had been sold to him as a “gabbeh” band, a term I had not heard previously.

He was told that a ground level is woven,

and is, subsequently,, embellished with embroidery.

In the case of this band it would appear that the black and white wider area is the woven ground and that the two narrow red-yellow-black strips are the “embroidery.”

David and I did a little subsequent research on this piece with folks experienced with bands, and found it was likely woven by other, perhaps not even Persian, tribal weavers, and that it was done entirely on the loom.  There is no extra embroidery added.

Marla Mallett often warns of claims of what can look like embroidery (some brocade usages can look like this).  She says that weavers will strive to do everything they can on the loom itself.  So, apparently, with David’s band.

Now David held up a Qashqa’i bag face.

Her it is again, flat on the board, and in a little better light.

David described this as a “classic” Qashqa’i rendition.

It has the expected birds, and animals in its field and Memling guls in the corners.

The top corner, above, shows the colorful chevon designs on its opening edge.

The next piece was Khamseh.

Geometric designs predominate, and the expected chickens are visible.

David next moved to an interesting Afshar bag face.

One might be tempted to call this a mat, but its narro,w bottom elem,

says that it is a bag face.

David says that its central field device,

is of reminiscent Karabaugh or Bijar.

He pointed to its “animal” border.

The next piece was another Qashqa’i,

David thought is 1930-ish.  It was more coarsely woven.

Its red, probably a chrome dye.

David held up the next piece.  A small, complete, flatwoven Qashqa’i, with leather straps.

Here is a closer comprehensive look.

And an even closer half of one of its faces.

David said that he was sure that this piece was made for use as a child’s school bag.

The next piece was a lovely, sumak bag face.

David said the the “openness” of its design might indicate an early date.

David said its warps are coarser and that it could be either southern Caucasus or Shahsavan.

He said that the drawing is very nice and

the use of outlining is very effective.

On balance, he estimated it to have been woven in the later part of the fourth quarter of the 19th century.

The next piece was the one below.

A closer look at this face design.

There were suggestions in the room that this could be another Qashqa’i piece.

David said that the cotton warps suggested that this was unlikely.

He then looked at the back, found that the piece is has only one pick of wefts between rows of knots.

He said that he felt that this piece was woven in NW Iran.

The next piece was the embroidered one below.

It seemed clear that this was not a complete piece, that it was part of something that had been cut off and then rebordered on the the left hand side.  In fact, all of the outside borders have been added.

David said that he could not tell for sure whether the embroidery material was silk.

Those in the room said the the calligraphy on this piece was in Arabic letters, but they could not tell what the specific language was.  Some thought perhaps, Turkish.  They said that the style of writing is not Caucasian.

Note: If some reader has some ideas about this calligraphy or can read it, please write me at

David had looked at the back of this piece before the session began.

This is the best image I have of this back.

There was speculation about whether the material of the back of this piece might be Russian, but there was none, in my notes, or that I can recall, about whether the piece, itself, might be Central Asian.

It is clearly an altered piece,  made from some questionable materials, is perhaps not very old, with odd field designs, in which calligraphy is prominently featured, and has some features that are similar to some Central Asian embroidery.  The main border seems to contain some pretty traditional drawing.

The next piece was described as “South Persian,” with no attempt at further specificity.

David says that, although this weaving is in a Qashqa’i style, he is reluctant to call it that.

It has a classical use of the “running dog,” a Turkish knot, with two picks of weft between rows of knots, and is entirely of wool.

He estimates that it was woven in the 1920s and thinks that it may not be a bag face.

The next piece may have been the most interesting of the day.

It might be best to begin with what this piece is not.

This piece is not old…David estimated that it was likely woven in the 1950s, 1960s.

It seems very likely that at least some of its dyes are not natural.

It has pile fronts and some flatwoven areas in the parts of the connecting bridge that are visible in the images above.  That much is usual.

But, it has some unusual and very interesting features.

You have already seen in the images above, that this piece has pile panels below those that mark the usual bag area of a khorjin set.

These pile panels continue onto the back of the piece before the flatwoven area of its back begins.

The image below shows what the back looks like comprehensively.

But, of course, the feature that attracts the eye is that there are small triangular, tasseled pieces on the sides of the connecting area of this bag at about its halfway point.

Let’s look at this feature, first, from the front.

Here is a closer look at one of these tab devices from the front/top.

We can see clearly that it is a separate piece, added to the basic bag assembly.  But, after examining these tabs carefully front

and back,

David determined that they are integral to the khorjin set because they are made from the same materials and with the same dye colors.

But this raises the question their function.

The fact that they occur at the midway point on the bridge would seem to make them very exposed to damage, wear and loss, if the bag is carried by this midpoint.

After examining it carefully David concluded that this piece was not designed to be carried in the way most khorjins usually are.

Instead, the room provided that its extra pile panels, make it possible for for the bag to stand on its own, with these extra panel areas serving as flat bottoms for both bags.

David believes that this piece was possibly made for use in the market and that the triangular, tasseled tabs were amulets meant to function protectively for the bag contents…nazarlik, if you will.

This, in turn, David said, suggests that this bag set should not be here.  It was, he believes, definitely not made for sale, but instead for the protective holding of  valued family possessions.  Somewhere in NW Iran there was a family that was moved, somehow, to give up what was likely a prized textile, made entirely for their own use.

The next piece is one of those that I brought.

This piece is about 9 inches wide and nearly 24 inches long.  It is flatwoven and the first question I always ask of those who have not seen it before is “What is it?”

One reason that it is puzzling is that, while it seems likely to have been a part of a larger assembly of some sort, it is complete as it came off the loom.   It has not be cut down.

Once a dealer came to a TM “rug morning” program with a similar piece and claimed that it was a unique kind of bag face.  David, who has not seen it before, pretty quickly wondered whether it might not be one of the front tabs on a horse cover.

That is precisely what it is.

There has also been some disagreement about its attribution.  Here are closer looks at its top third,

its middle,

and its bottom third.

Conjectures about its proper attribution have varied, with one experience person even suggesting that it might be Heriz.

David thought its probable assignment less problematic, saying that he saw it as either Lori or Kurdish.

I’ve owned this piece for about ten years and bought it here in D. C.  Sometime in the past year, I was looking through a stack of small pieces in another dealer’s shop here, and encountered what could well be the other front tab for this horsecover assembly.

Since both of these tabs have surfaced in shops in Washington,   D. C., a further intriguing question is: “Could the back of this horse cover also be here somewhere?”

The next piece was a side panel from a small Shahsavan cargo-bag type “mafrash.”

The colors are mild and the decoration is comprised of five Memling guls, bracketed top and bottom by borders of a much smaller scale.

David said that it has a cotton foundation and that the milder colors are from dyes that are partly not natural.

If that is so, the synthetics used seem benign, since there is not much difference in color front to back.

The owner of the next piece described it as a Shahsavan wallet.

It is just short of five inches wide and eight inches long.  Its face is done in sumak and its back looks like this.

David asked the owner how he thought the piece was carried, and the owner replied that he thought that it was just tucked inside one’s belt, waistband or sash.

David indicated that he thought it might have been carried around the neck on a cord attached to its corners (David is undoubtedly right that “a cord around the neck” is a traditional way to carry such a wallet, but there is no trace of the remnants of such a cord on the upper edge or corners of this piece).

David said that it was also possible that this piece was used as a tobacco bag.

David did not think this little bag is very old (it’s in perfect condition).

He estimated that it could have been woven as late as the 1960s.

The next piece was much larger.  It was a NW Persian jajim and has a warp-faced weave.

David said that such pieces are woven in narrow strips, 200 – 300 feet long, then cut into sections and sewn together.

He said such pieces were woven near Rescht, famous for its flatweaves and opulent embroideries.

The next piece was a Baluch pile rug.

David said that this piece is reputed to be Iranian and is very unusual.

He said that the rugs we call Baluch made in a number of locations, but many are made in NE Iran.  Baluch production has been huge, and much of it is commercial. He said the best Baluch rugs are from near Meshed and are very fine.

Below is a closer corner of the piece above.

And here is a closer image of much of its field.

The next piece was a salt bag with a pile front.  David said that this is a Persian Kurdish weaving.

He said that is it a dowry piece, but simple.  The graphic effect is very nice.

Here is a closer look at a detail of its front.

Here, below, is an overall look at its interesting back.

And again, a closer detail focused on the animals.

Lovely natural colors and recognizable quadrapeds.  David saw them as goats and a reflection of weaver interest.

The next piece was an Afshar rug with a square-ish shape, something David said we see with fair frequency in Afshars. He also noted that there are extra, narrow, “skirt” borders both top and bottom.

Referring to the red-ground corner pieces in the image above, he noted that they often can be combined to form the the same design as that seen here in the central blue-ground “medallion.” A comparison of how the corners in this piece would look when combined, shows that not quite to be the case for this rug.

Below is a closer, lower corner.  David  said the colors are classically Afshar.  He noted some restoration, but said that it has been done well.

Notice the bottom skirt border in the image above.

Below is a detail of much of the center of the field.

And, here, a close view of the central device in the field.

David moved next to a large flatwoven piece below.

Here is a comprehensive view of it.

David said that this is likely a fairly new “soufreh,” an eating cloth (think “picnic”), but that its square shape suggests that it could also have been used as a ‘ru-korsi,” the covering used with a “korsi,” the traditional Iranian heating “table.”

Many readers will be familiar with this heating system, but it might be useful  to some to give the wikipedia link that describes both the “korsi” heating unit and (briefly) the “ru-korsi” covering.

Below are some closer detail of this piece.  First, a corner.

Next, its field.

Strong graphics.

The next piece was the city rug below.

David said that this was a better quality Kashan, likely woven about 1920-30

It has damage in much of its lower half.

Here is a closer view of an upper corner.

The next piece was another decorative rug.  It was heavy and difficult to get up flat on the board.

David said that is was a later type Kashan, probably 1950-60.

Here are some closer details of it.

An upper right corner.

Here is a view of one side.

A lower corner.

A center of the field.

This piece has attractive colors.

The next piece was a large Caucasian kilim.

The holders could not get it all up onto the board.

Here are some closer views of details of it.

Below, is an upper third.  Notice a little calligraphy between the two medallions in the top row left.

There is definite lettering.  It is not entirely clear what it says.

Two more details of other parts of this piece.

David spoke to questions,

and the program came to an end.

The audience began its migration.

I want to thank David for permitting me to fashion this virtual version of his program and for his help with the editing work we did on it, afterward.

I hope you have enjoyed this in our continuing attempt to convey some of these interesting and valuable sessions to a larger audience.


R. John Howe