Posted by R. John Howe on 01-27-2003 08:22 AM:

Beautiful Bidjar

Dear folks -

As I mentioned in the introductory essay, I was very attracted to the smallish Bidjar piece that Joe Fell presented.

I especially liked its large scale field and the way the smaller scale border complements it. And, of course, the color is quite wonderful.

I am inviting John Collins, who looks in on our proceedings here from time to time with a careful observation or two, to comment on this piece, indicating any particular features of it that we should be noticing.

John, please note, that as part of my inducement here I have gone back, researched and adopted your preferred spelling of "Bidjar," something about which I know you are a partisan.


R. John Howe

Posted by M. Wendorf on 01-27-2003 10:57 AM:


Dear John:

In English I believe the most appropriate spelling is Bijar. I think that Bidjar is from the French.

Joe Fell's Bijar is well known and of a known type. It was exhibited last year at Minasian's Kurdish rug exhibition just outside of Chicago.

Thanks, michael wendorf

Posted by John Collins on 01-27-2003 03:08 PM:

Yes, this is certainly a beautiful Bijar. I will be in the office tomorrow and will post a few references. I wrote an article once for ORR about Arabesque themes in Bijar rugs which refered to this particlar design. Also, Rippon Boswell sold another wagireh from this group some years ago. This example is a wagireh from the Garrous area. The border is employed as a guard border in large format Garrous carpets. I'll round up a few pictures.

As for the spelling, I would guess that the "d" was a Germanism which rooted itself in the trade and in early rug books
because the Germans were some of the first modern Westerners to write about rugs as art. I reluctantly stopped using the "d" about ten years ago at the urging of John and Susan Wertime, who have attempted to regularize our transliterations of Persian words. Lo these many years, I have been

Cheers, John Collins

Posted by R. John Howe on 01-28-2003 08:40 AM:

Dear folks -

John Collins indicates that he wrote an article of a particular group of Bijar rugs in the Oriental Rug Review.

This was, in fact, a stellar issue with wonderful pictures of some truly beautiful rugs, plus some fine writing. This issue, many of us know, can still be purchased. Here's the link:


R. John Howe

Posted by R. John Howe on 02-02-2003 03:51 AM:

Dear folks -

In his post above John Collins seems to say that Joe Fell's piece is a wagireh (a sampler) rather than just a small Bijar rug. If so, I'm not sure that Joe was/is aware of that. At least, he didn't mention it in his presentation.

I'm guessing that John's indicators are the smallish overall size and the outsize scale of the field devices. He also indicates that the single border on the Fell piece is a "guard" border on some large Bijar rugs.

John my yet get a chance to post some of the additional comments he indicated would be forthcoming. (He actually has a business to conduct and that may distracted him from the more leisurely activities that we pursue here.)

I will give him a few days before I begin to "mine" his ORR article myself.


R. John Howe

Posted by Tracy Davis on 02-02-2003 01:57 PM:

John, I agree with you that this rug's a stunner. And let me take the time here to thank you for posting your photos and comments about these TM morning sessions-- it's a treat, and the next best thing to be able to see these rugs in person. I appreciate your taking the time to do this.

My question involves the Bijar-- has it been reduced? It's hard to tell from the photos.

Posted by M. Wendorf on 02-02-2003 05:23 PM:

Bijar Arabesque Rug

Dear Readers:

The size of this Garrus Bijar arabesque rug is 5'2" x 7'3". I do not think the rug is reduced. I have never thought of this rug as a wagireh. The size could be consistent which such label, but I would expect to see more design variation in a wagireh. There are several related examples that have been published or exhibited in this size and with similar drawing. A great example was at Sotheby's London in December 1989. A long carpet formerly owned by J. McMullan and now at the MET may be the granddaddy of them all.
Met 1970.302.6

John's article in ORR is great reading and I hope he posts further on this group.

Thank you, michael wendorf

Posted by John Collins on 02-03-2003 04:41 PM:

Dear folks,

I have not had much time this week to reply, and I don't seem to have a copy of the ORR issue. However. I do owe you an explaination for the "wagireh" comment.

Well, about wagirehs…

There were paper cartoons used in many Persian workshops. There were also “sample” rugs made. These were used in many areas, but the Bijar production was more prolific than most and often quite artistic. Some of the wagirehs were as small as 2’ square. At the other end of the spectrum, I recall a 10’ square Bijar sample of the “split-arabesque” design that was in a Massachussetts dealer’s collection.
The small formats usually focus on a crucial point of a design which will be repeated in the full carpet. Medium size examples will often show a corner and medallion section, with several border options. Many of these are mere rug curiosities and some are actually unattractive. I feel the most successful Bijar samplers are those which have an internal logic, a composition which is aesthetically pleasing on its own, aside from the “sample” function. Such an example is shown in my Bijar medallion article in HALI 111, page 74. I think that some designs, especially the large-scale themes of the “split-arabesque” and its related designs needed to be worked out in full scale, perhaps for the benefit of the person who was ordering the carpets to be made. Remember, this was an atelier situation where carpets up to 6 or 8 meters might be ordered. The 10’ square sampler in Massachussetts was a reasonable look at the design in full scale for such a customer. Now, the Fell example is obviously a normal 5’ x 7’ proportion. However, I interpret it as a sampler of a complex field design, framed by a classic Garrous guard border. See the guards in the abovementioned Hali article Ill. 6. Framing this field so minimally would be an odd thing to do if this was meant to be a typical rug. Such eccentricity would not be expected in an atelier situation. Bijar weavers, however, often created wagirehs with a pleasing symmetrical design. None of us know positively if this was a sampler but that is my guess, based on decades of handling these Bijar products. Nothing radical here, just an observation.

Best Wishes, John Collins

Posted by Tracy Davis on 02-03-2003 06:36 PM:

Thank you for the additional comments, John C. If the rug hasn't been reduced, then I can understand the conclusion that this may have been a wagireh. The rug as it is looks incomplete to me. Have you seen other wagirehs (is that the plural?), maybe smaller in size, that have just one border, i.e. they consist of a field design and major border with nothing else? In that case, they wouldn't truly be "samplers" in the way I've always understood them, as in a replacement for cartoon or talim. Their modern analogue might be what we call a "strikeoff", which is done to show a customer the actual colors or motifs in the rug to be woven. Has a distinction like this ever been made?

Posted by Michael_Wendorf on 02-04-2003 03:38 PM:

Dear Folks:

Tracey's strike off suggestion is interesting and seems to be consistent with what John calls a sample rug as distinct from a sampler or wagireh. It seems to me the distinction between sample rug or strike off and sampler or wagireh is a small but relevant point when considering this arabesque rug. In the end, I am not persuaded that this is either but agree that it is possible. The size of this rug and other related pieces, including an example with silk warps and cotton wefts, is too standard to make me think this was a strike off or wagireh. Either way it is quite beautiful.

I remember well John's contribution in ORR XII, number 4. His article "Arabesque Themes in 19th Century Bijar Carpets" credited the carpet designers of Bijar with being the keepers of a classical carpet tradition during the 19th century, keeping alive the various "vase, palmette, and strapwork traditions of the central Persian carpets of the 16th and 17th centuries." The article then went on to describe two types, Type A and Type B, of split arabesque rugs and to demonstrate how they are the synthesis of 16th century Shah Abbas Isfahan palmette rugs and 17th century Isfahan tree rugs. This seems to me to be pretty much right on. Indeed, though not specifically mentioned by John, beautiful arabesque designs may be found on 16th century Timurid inspired faience mosaics in the Jami Mosque in Isfahan. For those of you who have Oriental Rugs in the Metropolitan Museum of Art by Dimand and Malley, the arabesque model is described in some detail at pp. 85-86.

Regards, michael wendorf

Posted by John Collins on 02-04-2003 03:49 PM:

Dear Tracy,

That is precisely what I mean.


for an even larger wagireh with a "split-arabesque" field in full scale which is surrounded by the same Garrous guard border on three sides. It is only a short step (and a much more aesthetically pleasing one) to the Fell example.

Cheers, John Collins

Posted by M. Wendorf on 02-05-2003 12:43 PM:


Dear Folks:

A. Cecil Edwards says this:

"The designs of Bijar have always been few, simple and generally rectilinear. The professional designer is unknown in Bijar ... Formerally a design was associated with a particular village ... The introduction of the wagireh system - whereby a mat was woven which showed one repeat of the design and was used as a pattern by the weaver - enabled the merchants to distribute the better designs more widely. The more recent introduction of scale-paper patterns has still further simplified the distribution." p. 125 The Persian Carpet. Edwards also states the Ziegler used wagirehs.

Kurt Erdmann in 700 Years of Oriental Carpets writes:

During the preparations for the carpet exhibition in the Kunst und Gewerbe Museum in Hamburg a curious piece turned up; it resembled embroidered sample cloths known from many districts. I had never seen knotted models and, therefore,
sent a photograph of the piece to Heinrich Jacoby ... Jacoby identified as Ziegler wagireh.

Erdmann then summarizes the literature as of 1970: In Lewis's Practical Book of Oriental Rugs a Bijar wagireh is illustrated and Lewis says: "Sample corners are mats about two feet square and are woven for the purpose of showing the variation of border, colour, and esign to some wealthy ruler who wishes a carpet woven. They are afterwards used in the weaver's family and seldon reach the market."

W. Grote-Hasenbalg illustrates a Herez wagireh and says: "Wagirehs, which occasionally reach Europe, serve as models for the manufacture of large carpets in those districts of the Orient which work for the European market. Our illustration shows the design for the medallion and corner, the filling between these, the main border stripe and guards."

Dilly: regarding a Bijar sampler: "Among the most interesting and delightful of small rugs is the Bijar Wagireh or Orinak, which displays sections of numerous patterns artistically combined. the purpose of the weaving was to produce models of craftsmanship and color combination for use in the creation of carpets. Some Wagirehs contain as many as five incipent carpet designs. An ulterior purpose undoubtedly was the preservation of pattern and color, generation after generation. "

Jacoby: In Persia before a new design is woven a small sample piece, a wagireh, is made from which the final appearance of the border and field colours in wool can be judged. It is sufficient if such wagirehs show a piece of the border and a section of the design without reproducing the whole."


Best, michael wendorf

Posted by R. John Howe on 02-13-2003 06:09 PM:

Dear folks -

John Collins has not had a chance to post further in this thread and indicates that he no longer has a copy of the ORR issue in which his Bidjar article appears.

I want to present and to describe very briefly some of the pieces in this article and issue, simply because folks ought to have a chance to enjoy them. I will not attempt to summarize Mr. Collins' article or to comment on these pieces myself. But perhaps the images will provide a basis for others to do so.

First, Collins presented a piece, I think, in its first publication, of what I think is the most magnificent Bijdar rug of which I know.

Collins labels this piece as Type A of the "split arabesque" variety. In ORR this piece appears with the long side on the vertical. I have rotated it left here to provide a larger image. You can get the ORR effect by printing it off and looking at it in that orientation.

Collins described a second very beautiful piece, also rotated left here, as a Type B of the "split arabesque" group.

Again, a most impressive rug.

Then, Collins presents a smaller (6' X9') piece with a white ground border.

He describes this as a "rural variant of the Type B archetype." He comments on the very effective border.

Collins moved in later portions of this article to other Bijar design groups. One such is the Bidjars that feature birds and animals with strapwork.

Here is just one of three rugs he presents in this latter group.

Last, I want to move outside the article itself to indicate that this was a Bidjar issue with a number of dealers presenting fine Bidjar examples in their ads. Mr. Collins had the back page of this issue and presented this rug on it. Once again, I have rotated this piece left to give you a larger image.

This is an issue worth having for the sheer beauty of the pieces included in it.

As I indicate above, perhaps others will be stimulated to comment on these pieces and/or this article.


R. John Howe

Posted by Patrick Weiler on 02-14-2003 01:26 AM:



Only the first of the Collins pieces appears to have resolved corners, where the major border design is fluidly continued around the corners unbroken.
What might this say about this piece, the other pieces and Bijar weavings in general?

Michael Wendorf quotes Edwards:

"The designs of Bijar have always been few, simple and generally rectilinear. The professional designer is unknown in Bijar ... Formerally a design was associated with a particular village ... The introduction of the wagireh system - whereby a mat was woven which showed one repeat of the design and was used as a pattern by the weaver - enabled the merchants to distribute the better designs more widely. The more recent introduction of scale-paper patterns has still further simplified the distribution."

Does this indicate that the first piece is later than the others?

My take on the Fell piece is that it follows the pattern found in later 19th century Caucasian rugs. They magnified and simplified a portion of a design from an earlier, larger carpet and produced marketable, more readily made smaller rugs in workshop settings. The Fell Bijar rug is just not as geometricized as many of the Caucasian interpretations of their classical designs.

Patrick Weiler

Posted by R. John Howe on 02-14-2003 05:11 AM:

Hi Pat -

Collins estimates that the Type A piece, the first one was woven in the "first half" of the 19th century. He places the second piece in "third quarter" of the 19th century and the piece with the white ground border as likely to have been made in the "last quarter" of the 19th century.

I think his age estimates are based in part on the detail of articulation of the designs used. He uses the word "archetype" with regard to the second piece. And there is some seeming conventionalization of design in the rugs estimated to be later. He may have other indicators of age as well, since he likely had these pieces in his hand at one point.

About the "butted" versus "resolved" border, I have seen butted borders in many rather sophisticated rugs. I think resolved borders are something more like the exception, excepting perhaps, in the instances of some groups of city carpets. I have not even noticed what Steve plausibly claimed: that often weavers manage corner resolution on the bottom of the rug as they begin and that the "butting" is more frequent at the top. I don't think "butted" borders is an age indicator, but others may find it so sometimes.

(I have encountered, one oddity, with regard to "resolved borders" during my trips to one local flea market. Some Turkmen chuval designs are now being produced in not entirely unattractive, but very coarse and inexpensive sumak. I notice that the borders on these pieces are invariably resolved for some reason. Also there are some borders used on Chinese and Tibetan pieces that seem almost always to be resolved.)

It seems very likely to me that the first two rugs I've presented above, were woven following a cartoon, but it is remarkable what weavers can do out of their heads. Edwards says that if you ask a Heriz weaver for her "pattern," she will often take out a small handkerchief-like item with a two-color curvilinear design. But she weaves this design in 13 or 15 colors in a rectilinear mode. Seems like magic.


R. John Howe

Posted by M. Wendorf on 02-14-2003 08:36 AM:

Dear Folks:

I think the age estimates of John Collins are likely to be based also in part on overall drawing, handle and color. Collins wrote concerning the A group that "The drawing is very curvilinear and naturalistic. The handle is very soft and pliable." ORR p. 13. The soft and pliable handle is a probable identifier of earlier all wool Bijars. The later all wool pieces tend to be much heavier and stiffer in the hand. Still later, cotton is more common in the foundation.

In addition, he had the ready comparison of the other known A type, the famous McMullan Arabesque carpet at the MET. That carpet is inscribed 1794. The McMullan carpet also has no main border. I am not sure whether two carpets make a group or not.

It is important when considering these pieces to consider the size of these carpets. The type A and type B rugs illustrated are very large carpets, over 18 feet long. These are very likely commissioned pieces for wealthy families. Finally, these rugs are most likely a synthesis of two distinct floral traditions as discussed in John's article. This remains, some ten years later, an excellent read.

Best, michael wendorf