Archive for September, 2009

Persian “City” Rugs, Old and New

Posted in Uncategorized on September 24, 2009 by rjohn

On July 19, 2009, David Zahirpour


gave a Rug and Textile Appreciation Morning program at The Textile Museum, here in Washington, D.C., on the subject of “Persian ‘City’ Rugs, Old and New.”

He did not define “city rug” rigorously, but drew on the understandings in the widespread use of this designation, both in the market and in the literature.

Jon Thompson provides an example of the latter in the organization of his basic book, “Oriental Carpets.”  Thompson offered a four-part typology for distinguishing oriental rugs and carpets.

1. Tribal weavings

2. Products of cottage industries

3. Carpets manufactured in town or city workshops

4. Court carpets.

David was concerned with the rugs described in item 3 above.

Thompson gives an archetypal example of a “city rug”

IshfahanSarrafMamouryworkshop1970sand his caption for it provides additional, concrete, defining details.

He says, “Workshop rugs were produced for commerce by an organized team of specialists.  Artists design the patterns, dyers match the colours, and weavers work from detailed cartoons which make possible the precise execution of large curvilinear designs…”

David provided a handout that listed the Persian cities (or areas) within which most “city rugs” have traditionally been woven.

They are:

Ishfahan, Sarouk, Kashan, Tabriz, Meshed, Kirman and Bijar

He also listed some noteworthy workshops.

In Kashan: Mohtashem, Dabir, Shad Sar,

In Tabriz: Hajji Jalili

In Meshed: Amogli

In Ishafahan: Seirafian, Sarraf Mamoury

In Sarouk: Mehajeran, Ferahan

In Kerman: Ravar (really a village)

Because rugs from an identifiable workshop are usually seen to be more valuable than those for which only a “city” designation can be given, a claim that a given rug was woven in a particular workshop is sometimes looked on with suspicion.  Some argue that many workshop designations are marketing ploys, like the use of “Serapi” to signal a high-quality Heriz.

As I prepared this virtual version of this “rug morning,” I looked at a lot of auction catalogs for the years 1975 to about 1995 and did find that a great many Kashans offered were claimed to have been woven in the Mohtashem workshop and that many Tabriz’ sellers tried to hang onto a possible workshop origin with the words “possibly Hajji Jalili.”

Still others say that many workshop attributions are based on real indicators.  For example, it is claimed that there are real, concrete indicators on which a correct “Mohtashem” attribution can be based.  And other diagnotic features are claimed for other particular workshops as well.

David and I talked after about the problem of recognizing accurately whether a given rug was, in fact, woven by a particular workshop.

He is of the view that it is dangerous to cite particular indicators, since they may be open to manipulation.  He says that there are particular rugs that were identifiably woven in particular workshops, but that the recognitions are not of sort that can be put readily into words.  They are more on the order of the answer to the question “How do you recognize your children?”

This may well be correct, but it puts the buyer at a distinct disadvantage, if all of the skills needed to recognize whether a given rug was woven in a particular workshop, reside only in the experience of dealers.  It has the additional disadvantage of not making the recognitions used open to scrutiny and critique.  It is a kind of elitism.   At a minimum it suggests that, if one undertakes to buy a rug that was in fact woven in a particular workshop, one needs to engage one’s own, independent expert resource on the workshop of interest.

Different subject: It is interesting to note what rugs are usually left out of most listings of Persian “city” rugs.

First, Heriz’ are not included, even those seen remarkable enough to earn the praise of the “Serapi” designation.  This omission might be justified on the basis of the proud practice of many Heriz weavers of using a guiding pattern that is a simple, often a two or three-color, picture or printed handkerchief, rather than a knot-by-knot digital cartoon.

And Heriz’ tend to move toward the rectilinear even when the pattern is curvilinear.

But look at what sometimes results.

Herizsilklate19thThis Heriz is in silk and is estimated to have been woven in the late 19th century.

Rugs from south Persia, other than Kermans, are also usually excluded from the “city” rug category.  To some extent this may be the result of the “tribal” names many of them carry, but here are two southwest Persian examples that show what these rugs can be as well.

The first is the Khamseh below.

SWPersiaKhamsehdated1865This piece puts the lie to the seeming occasional impression in the literature that Khamseh rugs are on the cruder side.  This piece is dated 1865.

A second southwest Persian example is the classic design below.

SWPersiaQashquaiworkshopOpie explicitly designates this a “Qashqa’i workshop rug,”  demonstrating that not all observers see a tribal designation as disabling for inclusion in the “workshop” category.

“Sennehs” are not usually seen as “city” rugs, although they can exhibit great precision and sophistication.   But David includes the nearby “Bijars” in his “city” listing.

So while the designations of “city” and “workshop” rugs operate pretty well to pick out some particular types for focus, there seems, sometimes, some inconsistency in the standards used to define these groups of rugs.

In his “Oriental Rug Lexicon,” Peter Stone provides these additional observations about “workshop” rugs.

“…Workshop rugs have been faulted because of their lack of design spontaneity.  Some designs, however, have achieved great artistic merit.  Many of these rugs are woven of the finest materials according to the highest technical standards…”

Stone then goes on with some thoughts that bear on the “old-new” aspect of David’s topic.

He says , “The modern workshop or factory system of rug production began in Persia in the late nineteenth century.  Most contemporary oriental rugs are woven through this system…”

My sense is that, in David’s session, his “old” city rugs  likely include those woven, either in the 19th century, or in the 20th until about 1930.  “New” city rugs are younger than that, might be quite contemporary, and, nowadays, include Persian designs woven in other countries.

We move now to the rugs that David had brought.  David walked us through the geographic types, supplementing that, on occasion, with workshop examples he also had.

We will follow David’s lead here, but it would have been very difficult to bring into this session good examples of all of the types (particularly since many of them would be room-size or even larger).

And one of the advantages of a virtual presentation is that we can supplement examples, where we need them, without having to lift, carry, put up, take down and carry again, at all.

So we are going to show David’s examples of his categories, but will supplement them, where needed, with images of pieces that were not in the room on July 19, 2009.

As the distinctions in his handout suggest, David prepared  carefully.


And, even as he is ready to speak, considers, closely, what he wants to say.


David had arrayed the pieces he had brought on the front board.



I am not treating the various types and workshops here in precisely the same sequence David did in his session.  Instead I will follow a sequence of types that came back to me as David confirmed afterward the descriptions of the pieces he had brought in.

The first “city” on this latter listing was “Tabriz.”  In some senses this is very appropriate, since Tabriz merchants were hugely influential in establishing rug weaving workshops in many parts of Iran.

Tabriz, most will know, is located in Iran’s far northwest.

We start with two examples of older Tabriz rugs.


Tabriz rugs seem to be quite varied in both design and color palette.


Almost anything is possible.

The example above was woven in 1925 and, so, is just inside the older group we described above.

But Tabriz rugs are fairly easy to identify by structure, since they have symmetric knots tied with hooks, rather than by hand, and warps that are deeply depressed.

The use of hooks has at least two effects.

First, jufti knots, a problem in many areas, cannot be tied with a hook and so are not a problem in Tabriz rugs.

Second, the use of the hook gives the back a very uniform appearance, almost like something machine-made.  Tabriz rugs are recognizable on the basis of type of knot and what Neff and Maggs call “weave pattern.” After one has seen a few, a quick glance at the back is often sufficient to recognize a Tabriz.

Some Tabriz rugs were woven with silk pile.  Here is an example with a very unusual design.  Cross-panels occur in pieces like Turkomen engsis and in some Anatolian types, like Ghiordes niche designs, but are rare in Persian rugs.


Rugs are still being woven in Tabriz.  Here is a contemporary example.


Tabriz rugs are sometimes woven with pile that is wool in some areas and silk in others.  David had a new example in the room.


Since silk stands abrasion less well than wool, as such a rug wears, an attractive “embossed” effect appears.

One Tabriz workshop, Hajji Jalili, is frequently cited or suggested.

Here are some examples the labeling of which says “possibly Hajji Jalili.”


The following one seems very large.


A third Hajji Jalili “possible” is the rug below.


Here, below, is one rug attributed to the Hajji Jalili workshop without equivocation.


Bijar is south of Tabriz, but still quite far north in the Persian northwest and we treat its rugs next.

Despite the fact that Bijar weavers are acknowledged to have woven some of the most impressive rugs known, the inclusion of this city and area in a listing of makers of “city” rugs might strike some as a departure.  Bijar designs often seem less formal and border corners are not usually resolved, this latter often seen as evidence that a cartoon was not followed closely.

Nevertheless, David included them in his listing


(he was, in fact, pressed by someone in the audience about whether he saw resolved corners as a defining characteristic of a “city” rug and he said not).

I have not included some of the readily available “Garrus” Bijars with their dramatic strapwork.  (The sort of thing that John Collins showed in his famous ORR article on Bijars.)  But I have chosen a couple of examples that seem to me both typical and attractive.

Here, for example, is a rather typical Bijar medallion design.


The piece above is attributed to the 19th century.

And here is one with a field of repeating devices, nested between one another, in alternate rows.

Bijar1900Bijar weaving often projects a Kurdish-like flair for color usage (for me, the use of blue in the rug above is inspired), although many say the best of Bijar weaving is, in fact, done by Afshars.

David had brought a new Bijar of the sort that you will encounter frequently, nowadays, if you go looking for one.


A quality item, woven in Iran, with a dense “herati” design and a traditional Bijar structure, such a rug is still difficult to beat for wear.  A few Persian producers of Bijars have begun, again, to use natural dyes.  The colors of these natural-dye, Bijar pieces, seem a little bright at first, but are jewel-like.

Rugs with a Bijar structure and with similar “herati” designs are woven in great numbers in India and Pakistan.

Sarouk rugs are also woven in Iran’s northwest and that is the group we treat next.


Rugs designated “Sarouks” are woven in a variety of villages in Arak.   The literature seems to suggest a kind of sequence of related, but distinctive types in this Arak area.  The first is “Ferahan.”  A second is “Sarouk.”  A third is “Ferahan Sarouk,” a term that seems, as one reads its defining components, more a sub-group of “Ferahans,” than it is of “Sarouks.”

There are a number of additional related types in this area.  There are “Josan” Sarouks (one of which we will see), there are “Mahals,” “Lilihans,” “Sarabands,” and the earlier “Mir Sarabands.”

We will treat only the “Sarouk,” Ferahan Sarouk, and Josan Sarouk” in our review here of “city” rugs from the Arak area.

“Ferahans” seem to have been an early Arak type.  They  have symmetric knots on a cotton foundation.  The handle of Ferahans is moderately flexible because alternate warps are not deeply depressed.  Designs tend to be “classic repeating” patterns like the “herati.”  The palette is often lighter and can feature lots of white.

“Sarouks” seem to have emerged as a clear type about 1890 and are distinguished from “Ferahans” by a structure with much more deeply depressed warps and the adoption, frequently, of medallion designs.  The Sarouk color palette tends to be darker.  Blue and red often dominate.

“Ferahan Sarouks” seem to be rugs that resemble “Ferahans,” but which have an alternate warp depression that moves them closer to “Sarouks.”  They often display some “Ferahan” color usages, including a distinctive green and expanses of white.

Here are two older Sarouks that David had in the room.


The rug above is the kind of old Saruk that most would readily identify.

David had another in the room with rewoven ends.


We did not have any “Ferahans” in the room, but there were three “Ferahan Sarouks” present.

Here is the first one.


David described this piece as an “antique Ferahan Sarouk.”

Someone had brought in a very different looking “Ferahan Sarouk.”


Here are some closer detail images of this piece.



A third “Ferahan Sarouk” in the room moved in a very different direction again.


This is a piece that requires closer details.

First a corner.


And another detail of field and border, closer yet.


The “Ferahan” heritage of this piece is visible in its lighter coloration and the small repeat devices that cover its field .

Here are some additional “Feraghan Sarouk” rugs not in the room.

SaroukFerahan19thAnd here is a second example with a lighter palette and a niche design.


Sarouks are still being made.  We had a new Josan Sarouk in the room.


The Mehajeran workshop is perhaps the most frequently mentioned  one making Sarouk rugs.

David had brought one Mehajeran Sarouk.


And here is another, larger one with a deeply saturated red field.



The next “city rug” type we treat here is the “Kashan.”

Kashan has a deep weaving tradition and was for a long time famous for its shawls.  But as the shawl market receded in the late 19th century, weavers turned to carpet weaving for which demand was rising sharply.

One possibly apocryphal story about how that happened involves a Kashan dealer who had a lot of Merino wool that had been processed in Manchester, England.  His wife was a skilled weaver.  He asked her to weave a rug using this “Manchester” wool and the results pleased lots of Kashan ladies and the market and before long “Manchester Kashans” became a valued rug product.  Kashans made from wool processed in Manchester are still marketed as “Manchester Kashans” and are seen by some to be very collectible.

Here is one example of a “Manchester Kashan.”

KeshanManchesternotdatedBut Kashans of other sorts and qualities are made and it is one city famous for particular workshops.

David had an old Kashan in the room made from “kurk” wool.  It is the larger piece in the image below.


Here are some closer details of this rug.  First a corner.


Then its somewhat unusual open field.


As we noted above, Kashans are one variety of “city” rug in which specific designers are alleged and noted.  This despite the claim of Edwards that there were no factories in Kashan and that all weaving is carried out in homes (see David’s indication below about this Edwards claim).  So it may be that “workshops” in Kashan were quite small family operations.

The most famous Kashan workshop is “Mohtashem” and, although Edwards does not mention it, you will, nowadays, rather frequently hear claims that a given rug was made there.

Here are some Kashans that are also described as “Mohtashems.”


The piece above is attributed to the 19th century.

Here is another, this one woven in 1900.


Silk rugs were woven in Kashan and here is one below with a niche design that “may be a Mohtashem.”


The rugs woven in the Mohtasham workshop themselves vary in type.  David had a smaller one in the room.


He said that this small piece was antique and in silk.

Here are two closer details of it.



As we mentioned above, the claim that a given rug was woven in the Mohtashem workshop is a claim that it is of very high quality and some feel that the designation has become mostly a marketing ploy.

Some even doubt that we can identify “Mohtashems,” reliably, using a stable set of indicators.  Still others, say “no,” there are real Mohtashems, and indicators.

But if Edwards is right about “no rug factories” in Kashan, one wonders where all the rugs claimed to be “Mohtashems” could have been woven.  (David, by the way, says that he has personally seen rug “factories” in Kashan and that Edwards’ indication is not correct.)

The strength of the urge to hold onto some sort of Mohtashem connection is demonstrated by the description of the rug below.


This is a new Chinese rug described as “Mohtashem-inspired.”

Two other Kashan “workshops” whose work we can illustrate are termed “Shad Sar” and “Dabir.”

Here is an alleged “Shad Sar” example.


And here is a rug claimed to be a “Dabir.”


This rug is estimated to have been woven early in the 20th century.

Ishfahans, as we said, as we showed one at the beginning, is often seen to be a near archetype of the “city” rug.  This beautiful city is only a little further south of Kashan.

Here is a typical Ishafahan carpet.

IshfahanMEilandIt is the example that appears in the most recent edition of the Eiland’s comprehensive guide.  Notice the great detail and the seeming perfection of the drawing.

David had brought an Ishfahan, with wool pile, but a silk foundation.


Since this rug was in the room we can show you some partial images that exhibit the details of the drawing in Ishafahan rugs.





Some Ishafans are entirely of silk.  The rug below is one such.


Some Ishafans are attributed to specific workshops.

The one below is attributed to the Ahmad workshop in about 1900.


The one below is referred to as a “Seirafinian” Ishfahan.


Here, again, below is the Ishfahan we saw at the beginning.  It is attributed to the “Sarraf Mamoury” workshop, about 1970, in Jon Thompson’s “Carpet Magic.”


I once heard George Jevremovic say in a presentation that he was sending  his “city rug designs” to Chinese weavers because the Chinese want to get designs right.

David had a contemporary piece in the room, with an Isfahan-like design, that he thought had likely been woven in China.


Again, some closer partial images are possible.




Meshed is not an area that is often listed when “city rugs” are discussed, but David includes both Meshed, and the Meshed “Amoghli” workshop, in his city and workshop listings.

Here is an example of a Amoghli Meshed.


Below is a signed Amoghli Meshed, with an elaborate design, woven in 1925.


And here is another attributed to the 19th century.


It is interesting, given the seeming scarcity of Meshed rugs, in general, how frequently the references to a “Agmohli” connection can be encountered.

A scan for “Agmohli” on eBay will often produce something like this.


Looks like an Ishfahan.

For our last geographic center for city rugs, we move to the south center of Iran and to the Kerman area.

Kerman rugs are seen by some to project the acme of Persian rug design and color usage.  Cecil Edwards was so wowed by them that some complain that he devoted too much of his classic book on Persian carpets to them.  But they clearly deserves serious attention when one examines city rugs.

David had a Kerman piece in the room.


He described this as a “Raver” Kerman.  “Raver” most will know, is a village and area close to Kerman city where the best Kerman rugs are said to have been woven.

Here are some other Kerman examples I collected from various sources.  The piece below is a large 19th century Kerman that projects both the sophistication of design and the range of color for which Kerman carpets are famous.


And here are two additional Raver Kermans.  The first is another 19th century rug.  To me it seems glorious.


The second example is a striking medallion design for which no date is given.


We have completed our virtual treatment of the types David discussed in this session.

David answered questions.


My own was that since the Tabriz merchants are seen to have been so influential in establishing rug production in so many parts of Iran, why do they seem not to have mandated a Tabriz structure or the Tabriz hook-based methods?

It seems remarkable, given the seeming influence of the Tabriz organizers, how varied the structures and other usages are in the different areas in which both city rugs and others were produced.

The session came to an end and folks began their after-program conversations and their further examination of the pieces in the room.





I want to thank David for permitting me to fashion this virtual version of his session and for his very considerable help in preparing and editing it.

‘Til next time,

R. John Howe

Afshar Rugs and Textiles, Part 1

Posted in Seidman, Michael on September 9, 2009 by rjohn

Dear folks –

On May 9, 2009, Austin Doyle and Michael Seidman


gave a Rug and Textile Appreciation Morning program on Afshar rugs and textiles.

This post is one of three that provides a virtual version of this program.

Part 1 is Austin’s lecture.

Part 2 presents the pieces that Austin and Michael brought in to illustrate particular aspects of Afshar weaving.

Part 3 is devoted to the pieces that members of the audience brought to this session.

The Myers Room was full, again, for this program.



Tom Goehner,


the TM’s Curator of Education, introduced Austin and Michael, saying that Austin is a medical doctor specializing in oncology. He is also the president of the Washington area rug club, the International Hajji Babas.

Michael is a molecular biologist, employed in research, and is a member of The Textile Museum board.

Both have presented previous “rug mornings,” and both of them collect Afshar weavings.

Austin began


with a lecture on the Afshars and their weavings.


I have presented a virtual version of it below in some detail. Because we do not have permission to include the images of examples used to illustrate particular points, this lecture is presented entirely in text.   Still, I would suggest that it is worth “plowing through” a bit, since Austin has summarized a lot of the current literature on Afshar weaving conveniently.

Austin first said that Afshari weaving has not been studied much because few westerners have visited the Kerman area in south central Iran where the largest numbers of Afshars have been located.

iranmap Here is a closer look at Kerman and its surroundings taken from Edwards’ The Persian Carpet.


Notice that Edwards seems here to place Afshars to the south of Kerman. Opie, in his book Tribal Rugs of Southern Persia, places them more to the southwest.

Centuries before the arrival of the Afshars in Kerman province, the area was inhabited by a variety of Persian, Turkic and Arabic-speaking tribes. Among the most important were three Baluch tribes, two Lori tribes, as well as some Lak tribes.

This diverse background, Austin said, is reflected in the immense complexity of various pastoral nomads that moved about in the Kerman area, among which Afshars were numbered.

A sense of this complexity can be seen in the following brief description. Sirjan (see at extreme lower left in the Kerman map above) is usually seen as the major collecting point for Afshar rugs, but also for those of the Buchakchis (a tribe less well known to rug collectors). In his respected book, The Persian Carpet, Cecil Edwards indicates that he thinks that Bam was actually the Afshari trading center, and that the Doragahis, another tribe, greatly outnumber the Afshars in the Sirjan area. Edwards also notes: “The Persian weavers of the Sirjan valley far outnumber the 40, 000 nomadic Afshari and their rug production is greater.” (ed.: Edwards retired from the rug business in 1947 and was writing in the later 40s.)

One additional sign of the diverse background of Kerman Afshars is that the weavings they produced are the most varied of any of the Persian tribes.

As indicated above, the Afshars were (in the past) pastoral nomads. They were among the “black tent” variety.


Austin said that those in the Kerman area migrated between the cool Jabel Barez mountains (which sometimes reach 4,500 meters) in the summer, and warm winter encampments in the lowlands (which extend west to the Persian Gulf).

He said that the Kerman Afshars are now mostly sedentary, with only a few thousand still living as nomads. There are a dozen Afshar villages where a traditional Turkic dialect is still spoken. Most Afshar descendants have mixed extensively with the Persian people in the villages south and west of Kerman.

Although Afshars are generally thought of as a Kerman area tribe, in fact, Afshar populations exist in a number of other areas as well, notably in Khurasan in northeast Iran, and in the Bijar area of Iran’s northwest.

Next Austin sketched some of the deeper historical background of the Afshars.  He said that the Afshars have roots in the Turkmen Oghuz who left Turkmenistan, east of the Caspian Sea, in the 11th century. They traditionally spoke a Turkic dialect, and some settled in eastern Turkey. but the majority of the tribe settled in Khuzestan at the head of the Persian Gulf.

At the beginning of the 16th century, the Afshars were instrumental in assisting six other tribes, all members of the Kizl Bash Confederation, in placing Shah Isma’il on the throne of Persia (1499-1524). By this time important segments of the Afshar tribe had migrated again to Azerbaijan in the region of Lake Urmia.

Also in the 16th century, the Afshars were forced to migrate from Azerbaijan and were resettled in several parts of Iran. Still in the 16th century, Afshar khans were given control of important parts of Persia and gained considerable power.

Their constant rebelliousness led Safavid shahs and rulers of later dynasties to command the tribe’s dispersal and resettlement.

Nadir Shah, who ruled Persia for about 10 years in the mid-18th century, was an Afshar.

The few remaining Afshars in Azebaijan, who had not migrated, have lost their tribal identity. Afshars of the Khamseh area around Hamadan were powerful until the 19th century.  They have now mingled with other Turkish speaking groups.   Some claim that the best Bijar rugs were woven by Afshars.

P.R.J. Ford indicates that Afshar groups in Khorasan and Mazandaran are loosely associated with the Kurds and their weavings are usually classified as Kurdish.  Afshars in Yaz, Fars and Khuzestan imitate local styles and techniques and are indistinguishable from local production by others.

Austin next listed some of the indicators, most of them structural, that are used to attribute particular weavings to the Afshars.


Afshar Attribution Indicators

Afshar pile rugs tend to be square-ish: 4 feet by 5.5 feet is a frequent approximate size.

Tribal Afshars are all wool (city woven workshop rugs are woven on a cotton foundation with depressed warps). Woolen structures tend to predominate generally, until the 1930s, when cotton was adopted.  There are numerous exceptions to these rules, in which cotton foundations can be seen in very old Afshar rugs, which have other characteristics of rustic origin.

The warps of Afshar rugs are usually ivory wool.  Warps are invariably depressed, usually about 45 degrees, but town rugs tend to be more deeply depressed than tribal rugs.

Afshars usually have two orange-red weft between the rows of knots, which help to distinguish them from Khamseh pieces, although single-wefted weavings are encountered.

Tribal rugs are usually symmetrically knotted, with the presence of asymmetric knots indicating either a village rug or a strong village influence.

All old Afshar rugs have long, flatwoven end finishes between 10-15 cm deep.  Generally. these are done in plain-weave stripes using a few colors. Occasionally extra-weft wrapping is used.  The presence of a row of diagonal bars in the end panels done in extra-weft wrapping strongly suggests an Afshar attribution.

Afshar rugs woven in towns tend to have a firmer handle than do tribal Afshars. In general, Afshars have a somewhat less flexible handle than do Khamsehs. They are somewhat more flexible than are Qashqa’i pieces.

Nearly all Sirjan rugs, whether Afshar or not, tend to have two picks of blue cotton weft between each row of pile knots. Almost all Sirjans have some degree of warp depression.  There is little evidence that either pattern or structure distinguish Sirjan Afshar weavings from those woven by non-Afshar weavers.

Tribal Afshars almost always have reds based from madder.  This is true despite that fact that reds from towns the nearby Kerman area are often derived from cochineal dye.

Some colors are seen to be Afshar indicators. The reds used tend to be distinctive, as is a particular shade of electric blue. A distinctive rosy-brown is used in Shahr-i-Babek rugs.  Color in Afshar rugs also benefits from the fact that the sheep in the Kerman area produce a nice white wool that takes dye very well.

Silk Afshars are rare (See Hali 29, page 77).

Afshar rugs tend to have geometric designs. There is sometimes a visible design influence from Kerman, especially in the form of the boteh.  Kerman, like India’s Kashmir, was a major producer of shawls in which boteh designs were heavily used.  Some of the earliest Persian pile rugs with boteh designs were from the Kerman area, including apparent Afshar products.

Afshar rugs frequently have seven borders.  Border designs include some with diagonal stripes and double boteh borders separated by a serrated column.

There is no real demarcation between designs in rugs by indigenous Afshar weavings and more commercial types.  Nevertheless, some frequent Afshar design usages can be listed. They include:

Afshar rugs often have over-sized hexagonal medallions, hanging lamps and 2-1-2 designs with substantial corner brackets. (See Pacific Collections, p.70; Opie’s Tribal Rugs of South Persia, p.183; Hali 126, p.24).

Afshar rug designs include infinite repeats of oversize vases. The likely source of this usage is a Kerman city design described as the gol-e-bolbol pattern which dates back to the Safavid period. (See village rug examples in Pacific Collections, p. 87 and in MacDonald, Tribal Rugs, p. 111. See tribal rug examples in Opie, Tribal Rugs, p. 224; Sovereign Carpets, p. 82; Hali 139, p. 87. For classical Kerman vase carpets see Hali 112, p. 83; for Lady Baillie Kerman vase carpet, Hali 128, p.111. For early Afshar rug with city influence see Hali 136, p. 46 and Hali 114, p. 85.)

Boteh repeat designs are also frequent in Afshar rugs. Afshars were among the first tribal weavers to use the boteh. As mentioned above, one Afshar boteh design usage featured double botehs with serrated columns (See Middleton, 118). Afshar boteh usages mimicked those of the fine shawls both imported from India and woven in Kerman. These shawls were seen to be the finest garments for tribal chieftains.

Large Afshar botehs often have an interior device that appears to be resting on a butterfly. (See Hali 34, pp. 17 and 65; Pacific Collections, p. 84; Tribal Rugs of South Persia, p. 195; Sovereign Carpets, 81. There is also pre-1800 Eastern Fars carpet fragment from the Burns Collection, Hali 120, p. 82, with a later derivative in an Afshar rug in Hali 140, p. 129.)

Afshars also weave a large latticed “tulip” design, usually on a dark blue ground. Donald Wilburg and David Milberg divided tulip Afshars into two groups. Type 1 with six narrow borders, and Type 2 which features a wide main border flanked on either side by several narrow guards (See Pacific Collections, p. 85.)

Shield designs are also notable in Afshar rugs. These shields are reminiscent of those in rugs from East Anatolia which borders the traditional Afshar homeland in Azerbaijan. They are probably derived from palmette devices. (See Hali 34, p. 18; Pacific Collections, p. 83; Opie’s Tribal Rugs, p. 221; Hali 120, p. 50; Hali 127, p.57.) There are also shield-shaped cartouches that often contain a palmette; and shield separated by a spikey shrub (Antike Orient-Teppiche, p. 97).

We have referred above to the fact that lattice designs are included in “tulip” Afshars.

Afshars also use compartmented designs, which divide field into rectangles or lozenge-shaped compartments (See Hali 57, p. 98; Antike Orient-Teppiche, p. 97).

Murgh (chicken) designs are also encountered in Afshar rugs, but less commonly than in Khamsehs. The usual version seen is that of chickens opposing one another around a vertical pole (See Hali, 29, p. 39).

Afshar rugs also sometimes exhibit Phoenix and Dragon designs (See Hali 116, p. 45; Opie’s Tribal Rugs of South Persia, p. 181; Antike Orient-Teppiche, p. 96).

The influence of Kerman city usages on Afshari weaving is visible, but not overwhelming.  As mentioned elsewhere the Afshars did adopt boteh designs that originate in Kashmir Indian and in Kerman shawls.  Cotton was grown in Kerman province and its presence in the foundation of some older, but more frequently in younger Afshar rugs is likely another sign of Kerman influence.

Kerman city production was marked by large rugs with realistic floral designs, cotton foundations, asymmetric knots open left, used in a distinctive fully depressed structure that included multiple wefts, and reds frequently derived from cochineal.  In contrast, the archetypal production of the Afshars features smaller rugs, closely clipped pile, geometic designs, which is symmetrically knotted on a wool foundation, with reds derived from madder.  While it is not unusual to have apparent tribal Afshar pieces with asymmetric knots, this usually indicates the presence of a Persian villager weaver.

Some nineteenth century Afshar rugs are exceptionally large and have European-style floral designs. They also have asymmetrical knots, deeply depressed warps and cotton foundations, suggesting that they are products of town workshops, although their coloring is distinctly Afshar.

Proximity allowed the Khamseh tribes of Fars and Neyriz to the west, to exert considerable influence on Sirjan weaving, so many of the latter, especially the flatweaves, are hard to distinguish from those of the Khamseh. Striped rugs and those with tree designs are made both in Sirjan and Neyriz.

Afshar designs are frequently variations on hexagonal schemes and stylized flower and foliage motifs. There are also frequently stripes in the spandrels of Afshar rugs.

Next Austin talked about the regional groups who are implicated in Afshar weaving.


First is the Sirjan area, mentioned frequently above. This area is a large valley west of Kerman city in Kerman province. It has the greatest production of tribal and village carpets in Kerman province. Here are the largest concentrations of Afshar weavers, who are outnumbered by Sirjan Persians who also weave.

Jabel Barez and Afshar-I-Kuhi are a mountainous area stretching southeast of Kerman all the way to Baluchistan.  Its rug collection center is Bam. The rugs are termed Afshar Jebel Barazi or Kuhi.  Afshars co-existed with Lak and Baluch tribespeople.  Afshar-I-Kuhi rugs are easily distinguished from other Afshar rugs, being deep-piled with soft, shiny wool in dark colors reminiscent of Baluch rugs (See Hali 58, p. 105). Old Kuhi rugs are symmetrically knotted with two shoots of weft between rows of knots, with depressed warps. Their foundation is either all wool or a mixture of wool and cotton. The Kuhi have a unique khorjin in which both the front and back are piled. Kuhi means “from the mountain.”  Flatweaves resemble those of Sistan Baluch (MacDonald, Tribal Rugs, p. 122.)

Shahr-I-Babak and Dehaj are areas located in the extreme Northwest of Kerman province. Dehaj is a village north of Shahr-I-Babak with a long history of excellent weaving, generally by Arab residents. Rugs have a high-quality weave with blue wefts and symmetric knots (See MacDonald, Tribal Rugs, p. 120; also the boteh rug in Ford, p. 69).

Pockets of Afshar (also Khamseh) are found in the neighborhood of Neriz, a town in Fars province to the west of Kerman. Designs peculiar to Neriz include triple-trunk trees supporting angular flowering, bows with numerous birds.  Saffron or white fields are noted, as well as deep indigo botehs and sophisticated flower borders (Hali 34, cover; Hali 20, p. 30; MacDonald, Tribal Rugs, p. 115; Hali 113, p. 34; Middleton, p. 119).

“Outback Afshars” (See Hali 117, cover) is a term popularized by Tom Cole to refer to primitive and archaic Afshar rugs, often with lazy lines, which have turned up in the bazaars of the Pakistani province of Baluchistan. These rugs have a structure similar to that of Azerbaijani weaving with cotton or mixed cotton and wool foundations, coarse weave, with uneven backs and slightly exposed wefts. Some of the oldest Afshar rugs known have a similar structure (Tanavoli).  Some rugs in this group have an asymmetric knot open to the right. Warps, in some cases, are cotton twisted with animal hair. Colors include very saturated reds and greens and an electric blue, plus peach.

Afshar weaving includes the following formats/weave techniques:

Khorjin: many pile saddle bags were woven.

Namakadans : salt bags used to carry salt or grain.

Jol-i-ash: horse covers

Qur’an bag: for carrying a Koran volume

Dozar: a rug two meters by one and a half meters or less

Zaronim: a rug one and a half meters by one meter, an older format.

Sofrehs: and similar flatweaves; mostly in concentric or zigzag pattersns. Done in a mixture of plain tapestry weave and double-interlocking brocade with delicate patterns in weft wrapping and weft substitution techniques. Sofrehs were used in a variety of ways, among them, for wrapping (there are “bread” sofres), and as eating cloths.

Sumak: less common than in the Caucasus but some technical similarities between Afshar gelims and Caucasian sumaks suggest a common origin.

Shiraki peech: another square-ish format, a flatwoven cover about five feet by eight feet, with a complicated structure in which plain weave is combined with weft wrapping, brocading and weft substitution to produce images and motifs that are often diamond shapes.

This is the end of Austin’s introductory lecture.

He and Michael now moved to examine some “in the fabric” pieces they had brought in to illustrate particular aspects of Afshar weaving.  To go to this Part 2 double click on the link immediately following or copy and past it into your browser.

Please note that there is also a Part 3, in which pieces brought in by members of the audience were examined.  This third part is at the following link:

Afshar Rugs and Textiles, Part 2

Posted in Seidman, Michael, Uncategorized on September 9, 2009 by rjohn

This is the second part of a three-part virtual presentation of a Textile Museum program on Afsar rugs and textiles conducted by Austin Doyle and Michael Seidman.

It is likely advisable to read through Austin’s lecture in Part 1 since it provides context for the illustrative pieces in this part.  Here is the link to that lecture:

A third part is devoted to piece participants brought in.  This link takes you to Part 3:

Austin and Michael had brought a number of Afsar pieces


arrayed on the front-of-the-room board.

Michael began, preliminarily, with some older pieces Harold Keshishian had brought that were possibly relatable to Afshar weaving.

The first of these preliminary pieces was the mounted shawl fragment below.


Here are some closer details of this piece.


The botehs are on a silk ground.


Harold estimated this fragment to the 18th century.


It is not clear whether this fragment is from an Indian or a Persian shawl, but Kerman shawls (and this is the possible link to our Afshar topic) were noteworthy and are thought by some to compete favorably with the more famous Indian shawls of Kashmir.

A second piece that Harold brought WAS a shawl from Kerman.


Here are some closer details of this colorful piece.



Its stripes with their finely detailed ornamentation are reminiscent of the similarly colorful and embroidered pantaloons of Zoroastrian women.

A third piece that Harold had brought was the Kerman brocade below.


Harold dated this richly textured textile to 1750 and said that it had been reconstituted from several pieces.

Here are some closer details.




You may recall that we saw some other images of this fine piece in our virtual treatment of the textiles at a party that Harold and Melissa held during the holiday season in 2008

Harold had also brought in an image of Nader Shah, the great Afshar military leader and somewhat less distinguished ruler of Persia in the mid-18th century.


Michael treated the material on the board, beginning with a series of Afshar bags with botehs used prominently in their designs.

The first such piece was a complete khorjin set.


This piece was attributed to Afshars in southwest Iran.

Notice that the botehs in its respective field areas are reflected so as to be seen upright on both sides when the khorjin is in use.

Here are some closer details.

First of the bridge, with chevron designs common to many SW Persian tribal groups.


And then of a corner of its lower half.


The back of this piece is a plain, brownish shade.

The next piece was a single khorjin face with a rural version of the boteh device.


The effect is subtle because of the close colors, but notice the diagonal use of color in its botehs leaning to the right.

Here is a closer look at one corner.


I mentioned from the audience that Afshars often seem to have a distinctive blue in their palette and Austin and Michael agreed that there seems an identifiable Afshar color palette.

The next piece was the smaller bag face below.


The scale of the botehs in this piece are somewhat larger and add to its appeal, as does the framing effect of its white ground main border.  The spiky floral meander of the white border is very characteristic of Afshar weaving.

The next piece was a khorjin face of the more usual size.


Notice again the use of the distinctive blue mentioned above. The intricacy of the designs around the closure system draws attention.


Here is a closer, more comprehensive look at this upper right corner.


The next piece was the interesting bag face below.


Here, an effective striped border frames a field with large-scale, instrument botehs, alternating with forcefully colored armatures.



The next “boteh” piece was a sizable rug.


As with the previous piece, colorful, instrumented botehs are placed in colums between a meandering lattice of substantial armatures.


This piece was described as featuring “serrated leaf forms.” The heavy armatures alternate between sections that do seem to be clear plant forms to others that may well be also, but that seem nearly mechanical.


The field is framed by two major borders. The outside one with its white ground is especially effective and quite characteristic of an Afshar border design.

The next piece was the very small bag below.


It was described as “finely woven” and its virtues are captured in this single image.

The following piece was a salt bag.


Skillful use of red and a yellow-orange enrich this piece, especially in areas where is is combined with a dark ground.


Note that the lattice of the field emphasizes the rectilinear while both the forms inside that lattice and the main border move toward the curviliner.

The design combinations used in the top opening-flap are unusual.


Again, there is a rectilinear-curvilinear field-border contrast, but this time is it reversed.

The bag of this salt bag is also unusual


Attractive striped flatweave is combined with a pile treatment of the opening-flap similar to be distinctive from that of the front.

The next piece was a classic, published Afshar rug, now locally owned, but once in the Ralph Yohe collection.  Rugs with this “tulip” design are thought to include some of the oldest known Afshar rugs.


Multiple sets of four richly drawn tulips are opposed on a dark blue field and bracketed by an intricate lattice.


Three smaller scale borders frame the dramatic field without competing with it.


This design is seen to be drawn from the shawl tradition and a dealer in the room said it looked Kerman to him.

There were some other examples of this “tulip” design in the room. The piece below


was this khorjin face (closure slits at the bottom in this image).

Here is a sightly closer look.


A third piece with this “tulip” design field was a small rug.


A distinctive white-ground main border with polychrome medallions frames its field.

Here are some details of it.




The next piece was the large rug below with a field of diagonals.


The colorful diagonals a composed of abstracted plant forms. The main border is an Afshar striped usage.


There was some question about whether this rug was an Afshar. Some thought that this rug, which had somewhat darker warp threads, might be from Fars province.

The next piece has a distinctive zigzag field design.


It was seen to be a Sirjan valley town rug, with depressed warps and a stiff handle, woven in the early 20th century.

Here is a closer corner detail.


And one that shows its “stars and blossums” field devices.


The next piece was the small bag below.


Its field features a large star and a number of smaller stars in background.

This time the zigzag designs are on the back.


And on the small panels between the slits in the closure system on the pile side.


The next piece was another khorjin with star devices arranged diagonally.


Here is a closer top center detail, showing the decoration of the closure system area.


There was some question about whether this piece is Afshar or Khamseh.

A further piece was the rug below.


This piece has seven or eight borders.

Here is a closer detail of one lower corner.


And his is a closer look at its field devices.


This rug was seen to be a city product.

The next was also a rug, this time a three-medallion design.


It has brown wool warps and a “eye-dazzler” field design surrounding its medallions,


and a subtle. but well instrumented, system of borders that frame the field effectively.


The careful composition and controlled execution of this piece suggests that it was woven following a cartoon.

We next turned to another khorjin face.


It has larger-scale floral-like devices in its field.


The stark white of its border contrasts dramatically with the strong colors of the field.


The colors of this piece are strong and beautiful.


Despite its careful composition, the spacious drawing of the main border design projects, for me, an unusual vitality.

In my view this khorjin face is one of the best of an aesthetically strong group of pieces presented in this session.

The next piece was a pile panel AM22

with Memling guls.


It has the shape and size of a Turkmen torba, but its border systems seem Persianate.


It could conceivably be a side panel of a small, pile cargo-bag type mafrash but that, too, could be questioned.

The next piece had a single Memling gul


but this time it occurred on a salt bag.


The colorful bag face below was attributed to the Fars province.


It “chickens” might suggest a Khamseh weaver, but it has white warps and a distinctive border that might license an Afshar attribution.


The next two piece were khorjin faces with similar designs.  The first khorjin face has a field of tiny boteh and a very fine weave.


These field designs are very similar to those sometimes seen on pieces attributed to the Qashqua’i.

Here is a closer look at details of the first one.


And here is a detail of an upper corner of the second one.


It was thought that both of these pieces are probably Afshar, with their white warps and characteristic borders.

The next piece was the rug below.


This rug was described as a rustic version of a “vase” design.

It has a camel hair field, which is unusual for Afshar production.


and an asymmetric knot.


The vase designs have the appearance of faces.  A local rug dealer of Persian extraction claimed that the faces were intentionally drawn and representated “div” or demons.  Here is a closer detail of its side border systems.


The following piece was the rug below.


It features a large central stepped medallion.



It was described as having a classic Afshar design.

Here is a closer look at a lower corner of it, with a classic Afshar spandrel design.


The next piece was, to my mind, one of the prettiest rugs of the day.



This is a Kerman area city rug with a lovely boteh white border.

Its indigo field effectively recedes to give the impression that its center medallion


and corner devices



“float” on it.

It has an asymmetric knot open to the right and “vase” motifs. It was estimated to have been woven in the 3rd quarter of the 19th century.

The next piece was yet another rug.


It was described as having a “medallion and vase” field design.



Its reds are from madder.

And it has “niche” spandrels.



Its knot is asymmetric open to the right. It was estimated to have been woven about 1900, but had classic Afshar colors of peach, electric blue, and a strong green.

The last piece among those that Austin and Michael had brought in was the one below.


Here are closer looks at its field.



Among its colors are an apricot and a peach shade.

It has star medallion corners.  The 2-1-2 design of the medallions in the field may be a more archaic version of the medallion and spandrel design seen in the other rugs of this design, shown above.


The knot is asymmetric open to the right. It was estimated to have been woven in the 19th century.

Perhaps its most interesting feature is that it exhibits “lazy lines.”



This rug appears to have the weave, colors, and lazy lines described as characteristic of “outback” Afshars, in the Hali article by Tom Cole, and certainly does have a primitive and archaic appearance to its drawing.

Folks in the audience had brought in a number of related pieces.  They can be seen in Part 3.  Here, again, is the concluding link to that material:

Afshar Rugs and Textiles, Part 3

Posted in Seidman, Michael, Uncategorized on September 9, 2009 by rjohn

This is Part 3 of a Rug and Textile Appreciation Morning presented at The Textile Museum here in Washington, D. C. by Austin Doyle and Michael Seidman on Afshar Rugs and Textiles.

Part 1 is Austin’s lecture in which he summarized the literature usefully.  The link to his lecture is here:

Part 2 is devoted to the pieces Austin and Michael brought to illustrate various aspects of their topic.  Here is the link to Part 2:

In Part 3 we moved to look at pieces others in the audience had brought in. The fist piece was a small bag the field of which was dominated by a flower form.


The next brought in piece was the rug below.


Again flower forms dominate the field. Here are two closer looks.



The floral forms in both of these pieces are seen to be instances of western influence in oriental rug and textile design.

The next piece was the bag face below.


Here is a closer look at the abstracted floral forms that populate both its field and borders.  The meandering floral main border is a characteristic Afshar design, and this particular floral border was seen in several Afshar rugs in Part 2.


The next piece was a “mystery rug.”


It was described as having been woven in Khorasan. It is full of Turkmen usages, mostly Yomut, but as drawn by a member of another weaving group.

Here is a closer look at an upper corner.


And here is a lower one.


Notice that there are pile elems at both ends, a sometime Yomut usage.

The guls in the field are a conventionalized version of the “tauk naska” gul


in which the “animal” forms in the quartered major guls have become “H’s,” (this happens with some Turkmen pieces too).

The gesture at a minor gul is a “beach ball” device seenon some Middle Amu Dyra Turkmen weavings, but more frequently on Caucasian rugs and textiles. This same device is employed as minor borders flanking a meander main border that lacks any recognizable Turkmen roots.

It was estimated that this odd rug was likely to have been woven by Afshars about 1930. This is plausible since the Khurasan Afshars live close to both Kurds and to Turkmen groups. A few years ago Michael Craycraft drew my attention to another piece with Turkmen designs that he attributed to the Afshars.

The next brought in piece was the one below.


This bag face was described as 20th century with Kurdish designs. Here it is reversed top to bottom.


Its feature of most interest, of course, is the unusual elem-like panel on a seeming khorjin face.

The next piece was also a small bag.


It was attributed to Kurdish weavers.


Here is a closer detail.


The next brought in piece was the large sumak below.


Here are some closer details of it.



Notice that the central part of the “bird-on-a-pole” devices contain “Greek keys” often seen to signal an Armenian presence.


The attribution of this sumak piece was uncertain, but the border designs and colors were thought to be possibly consistent with Afshar work.

The next piece was a very small, vanity-type bag.


European style flower forms are heavily abstracted.


There was conjecture about whether this piece is better attributed to the Afshars or the Bakhtiaris.

The next piece was another large sumak.


It had a field composed of left-leaning “stripes” of small poly-chrome medallions.


Here is a closer look at the internal intrumentation of these medallions.


A number of sofrehs had been brought in and the piece below was the first of them.


Sofrehs are distinghuished by a variety of uses. This one is seen to be for carrying bread or bread dough.

Here are some closer looks at parts of this piece.


My notes draw attention to the side edges of this piece.


It was attributed either to the Afshar or the Khamseh.

The next piece was another “bread” sofreh with lovely colors.


Again, a closer detail image.


The attribution conjectures about this piece paralleled those of the previous one.

The following piece was also a sofreh, but of a different type.


It is an “eating cloth” type and was put down on the ground for meals. “Eating” sofrehs are also sometimes called “bridal paths” because they were apparently also on occasion used in wedding ceremonies.

This is a substantial textile, 12 or 14 feet long, with a gray-abrashed camel ground field, inward pointing zagged black borders and


heavily decorated ends.


Once with it in his hands, Tanavoli initially opined that this piece was likely Kurdish. When I indicated that some others had thought it Afshar he immediately agreed that it could be that as well.

In his lecture, Austin mentioned that one frequent Afshar color usage is a distinctive blue and this piece (although it is not readily visible in these images) has enough of it to suggest to me that it is mostly likely Afshar. You will see this distinctive blue more readily in some of the other sofrehs that follow here.

The next piece, another eating sofre, with a design very like mine immediately above, but shorter, WAS attributed to the Afshars.

Note: From this point forward the owner of these pieces has, at my request, supplemented the descriptions, made in the text from my notes, with captions of his own.  Since his knowledge of these pieces, and access to them for purposes of description, is far superior to the indications in my notes, his captioned indications should be taken to be the accurate ones.

Afshari dining sofreh from Khorasan, NE Persia

Here are some closer details of this piece.

Beautifully decorated end panels of Afshari dining sofreh from Khorasan

The distinctive Afshar blue usage is more visible in these end panels.

Here is a detail of a device in its field.

Characteristic dendritic zig zags on borders of Afshari dining sofrehs

The next piece was also a sofreh attributed to the Afshars, but with a very different palette.

Afshari sofreh or cover from Jiroft, south central Persia

The caption above provides the indicated attribution.

Here are some closer detail images.

Detail of border design in Afshari sofreh or cover with Luri/Bakhtiari influence
Detail of Afshar sofreh or cover showing Lori/Bakhtiari design influences

The next piece was a Sirjan-valley sofreh.

Afshari bread sofreh with simple yet powerful graphics

Again some closer looks at parts of it.

Detail of Zig zag border of Afshar bread sofreh from Sirjan
Detail of playful central field of Afshar bread sofreh fron Sirjan

A next piece was yet another bread sofreh below.

Afshari bread sofreh from Sirjan in typical design format

Again, the caption provides the attribution.

Here is one closer detail.

Typical highly decorated end finish to Afshari bread or flour sofreh

The next piece was this complete khorjin set.

Complete Afshari khorjin in soumak technique

Its field is a tesselated version of the “bird-on-a pole” design with internal “Greek key” instrumentation.

Detail of Afshar khorjin in soumak technique, bird design with a short bridge.

The last of the brought-in pieces was the one below.

Piled large bagface or sofreh woven by Afshars or Veramin

It owner attributed it to Veramin…Afshar Veramin.

Here are some closer details.

Heavy brown wool warps star design border typical of Afshari weaving from Veramin
Unusual Afshar weaving of a large bag or a piled sofreh from Veramin

Austin and Michael answered questions,

the program was adjourned and folks moved to the front.


others compared notes on pieces in hand.



My thanks for Austin and Michael for permitting this virtual version of their program and their considerable editorial assistance in producing it.

Pat Reilly provided a good set of notes and Tom Xenakis helped editorially as well.

I hope you have enjoyed what seems to me an ambitious program, well executed.


R. John Howe