Home > Rug, Kelim, Soumak, Textile Post Archive >Milestones: Tabibnia and thompson Parts V&VI
Mon, Sep 12th, 2011 12:46:56 AM
Topic: Milestones: Tabibnia and thompson Parts V&VI

Another sticking point RK found in the “Milestones” text is the glossing over of what Turkmen weaving is all about.

The historical struggle between the Aqquyunlu(white sheep) and the Qaraquyunlu(black sheep), and the part they played various times in the struggles of warlords like Gengis Khan and his successors, Timur and his successors, and the later Ottoman and Safivid dynasties, is well documented and thompson’s brief treatment of it does little to explain what, if any, types of weavings these two Turkmen groups might have had, or might have made.

RK does not believe they were much different, in relation to carpet possession and production, than the warlords and dynastic large-scale societies they served .

By this we mean they were clients who ordered weavings and not actually in any way part of the economic and cultural process that made them.

Unanswered in “Milestones” went the question: What did the weavings, if they had any and surely they must have, of the white and black sheep Turkmen look like?

RK does not believe they looked like the Turkmen weavings collected today but were more classical in style – large medallion(s) rather than the ‘gol-centric’ style.

RK also believes ‘gol-centric’ weavings existed during this period (say 13th-15th century) but this type of weaving was made by the independent Turkmen tribes who lived outside the spheres, and power-zones of influence, of the Aqquyunlu and the Qaraquyunlu.

During the period thompson discusses the Aqquyunlu and the Qaraquyunlu were more sedentary/settled urbanites than nomadic, and it is very conceivable their carpets reflected this difference.

But: Who were these independent groups, and what was their influence on the development of what thompson refers to as the “Turkmen style”, is the great mystery the history of Turkmen pile weaving

By the way, thompson only mentions a Turkmen style but does not discuss how or what this might have looked like.

RK has no doubt there were Turkmen, read ‘gol-centric’, weavings from 13th ,14th, and 15th century but this is not the place to expand such a discussion.

Another contention we have with dr thompson, but far less important or as close to our interests, is his apparent belief in the “story” of the “found in Tibet” 14th century “animal rug” purchased by the Metropolitan Museum of Art in 1990.

Confronted animal rug;14th century; Turkey; Purchase Harris Brisbane Dick Fund, Joseph Pulitzer Bequest, Louis V. Bell Fund and Fletcher, Pfeiffer and Rogers Funds, 1990; Accession Number1990.61

The story, and there are several permutations, tells it was found floating in a river after it, and early other rugs, were discarded during the “cleaning” of an old temple.

The rug then appeared in Katmandu, passing thru several hands until it ended up in London.

RK saw the rug, and two others – the Faces rug and the Kirchheim animal rug --, with a now deceased dealer, Lisbet Holmes, in London.

Basically RK does not believe for one minute these rugs came from Tibet or were discovered under the circumstances thus described.

Nor do we believe the Met’s rug is 14th century or for the matter even made in Turkey.

RK believes, and has said so in private for almost 20 years, it is 16th century and an Afshar made in Iran.

Not to put this out of proportion we will end here, as, after all, thompson only briefly mentions the rug once on pg.27. But in doing so he does state

…the acquisition of a carpet by the Metropolitan Museum of Art in New York that is believed to have been preserved in a monastery in Tibet”(10)

The footnote refers to daniel walker’s, the former curator at the Met, publication about the acquisition in 1990.

We also do not accept thompson’s rational for the scarcity of rugs from the 15th century

The survival of carpets from the fifteenth century has been a matter of chance, determined by the historic patterns of trade and the accidents of political history, and in part by dry conditions, such as those in Egypt, which favor the preservation of of organic archaeological material.”(pg.28)

As RK sees it the paucity of 15th century, and earlier, carpets is due primarily to the fact very few were produced. This might be hard to imagine from the voluminous amount of later, ie 16th -18th century, carpets but this was, RK is sure, the case.

During those times, pre-16th century, these early carpets were far more non-secular than they eventually became, and while in some parts of the Near East, like western Turkey, there was “commercial” workshop carpet production, ie secular, it was not in near the volume seen in later times.


Dating carpets, be they classical, village or nomad, is at best educated guess work, regardless of the validity, or not, of a C14 analysis.

There are, however, certain parameters and clues one can rely on to provide evidence and substantiation for back-up.

However, it appears to us in the “Milestones” text dr thompson has ignored some pertinent factors and gone ahead and over-dated more than a few of Tabibnia’s carpets.

Let’s look, for example, at some of those in the Ottoman section.

Plate Two, a small-pattern Holbein carpet is no doubt an early example, but not mid-fifteenth century as C14 dating and dr thompson both claim.

Detail, Plate Two, “Milestones”

RK would date it a century later, well into the mid-sixteenth.


Well, for starters, when it is compared to a carpet RK believes truly is mid-fifteenth century there are significant differences documenting our position.

Fig.46, “Milestones”; detail of a small-pattern Holbein carpet fragment in the TIEM, Istanbul, 303, RK believes mid-fifteenth century

First, and most important, is the visibly compressed articulation the field pattern has undergone in Plate Two’s rendering of the iconic small-pattern Holbein design.

Any honest comparison of the smaller of the two main field patterns shows Plate Two’s are not nearly as spherical and well-proportioned as Fig. 46’s. This is due to vertical compression of the design (compare below the upper right medallions to see this compression).

Left: detail Plate Two showing the vertical compression as compared to Right: detail Fig. 46

The smaller medallions are not the only compressed element -- notice the octagon/stars from the interior of these medallions are equally compressed in comparison to what the weaver of Fig. 46 was able to achieve.

Compression like this is a tell-tale sign of later workmanship, but there are others.

The far more delicate, intricate and fluid interlace surrounding the larger of the small-pattern Holbein medallions, as well as the more detailed drawing of the central cruciform in those medallion, also demonstrate these tell-tale post-period qualities.

And, as the side-by-side comparison of the “Kufic” borders on Plate Two and Fig 46 below shows, it is quite apparent Plate Two is the later version and, again, Fig. 46 the earlier.

Left: Plate Two; Middle: Fig 46; Right: Fig 31

We do not think the TIEM’s small-pattern Holbein is 14th century, and concur with thompson’s calling it 15th, but how could he then believe Plate Two is its contemporary?

Here is a detail of Fig. 31 (“Milestones”), which is in the Museum for Islamic Art in Berlin, I 5526.

Fig. 31, “Milestones”; Museum for Islamic Art, Berlin

It is a large-pattern Holbein with a “similar” Kufic style main and a secondary border with the rosette and star double kotchaks ‘cartouche’.

This is why we included it along with the Kufic borders from Plate Two and Fig. 46 to illustrate what we feel is the archetypal, and earliest, version.

We are not surprised thompson misjudged the small-pattern Holbein (Plate Two) as being contemporary with a genuinely earlier one (Fig.46), as he and many other classical-carpet snobs frequently view nuance like those we cited as insignificant, something they surely aren’t.

When all else fails it is often illuminating to carefully examine borders, as their drawing and iconography, and not necessarily what appears in the field, can often provide important clues to a weaving’s age.

By the same token Plate Three, dated early sixteenth century in the “Milestones” text, is in our estimation likewise over-dated by 100 years.

Plate Three, “Milestones”

It also is a small-pattern Holbein but anyone who believes there is only 50 or so years between it and an archetypal example like Fig. 46 is not only dreaming, they are sleep-walking as well.

Here compression has changed the proportions of the interior of the “holbein” medallion, but the far more serious error is the lack of delicacy in the articulation of the interlace surround, which has now become somewhat crude and obvious – the mystery of the original so well expressed in Fig. 46 lost or forgotten.

This problem is also present in the Kufic border, which shows what RK calls an early version of this border’s final incarnation.

Courtesy of dr. thompson Plate Four, a charmingly degenerate version of the small-pattern Holbein design, is also the recipient of a century over-dating in the “Milestones” text.

Plate Four, “Milestones”

And although thompson hedges his bet, as the quotation below shows, we still do not agree with his rational this rug is anything but late 17th.

A carpet such as this with no known previous history is difficult to date. It has an unusual and intriguing design with archaic features. It is easy to ‘play safe’ and offer a seventeenth-century date and it takes a little more courage to make it earlier: but the form of its interlaced roundels is essentially fifteenth-century and they barely appear in later carpets. An exceptional feature is the structure, which has something in common with the Konya and Beyshehir carpets that are probably fourteenth-century. A sixteenth-century date thus seems easily possible.

Well, anything is possible but what is probable, or more to the point what is obvious?

This rug, as charming or intriguing as it might be, is not sixteenth century. Just look at the main border in comparison Fig 46 and Plate Two – it screams: I am not 16th.

And by the way, in his short description for Plate Four dr thompson does not devote one paragraph, sentence, or even a few words, to discuss the “exceptional” structural feature he cites.

Again, here’s thompson talking talk but stumbling badly walking the walk.

However, RK does see eye-to-eye with thompson when he calls the Kufic main border of Plate Five, a “Lotto” carpet, “…the earliest form”; however, this particular rendition displays marked vertical compression and significant loss of internal proportion.

Therefore, we cannot agree with his dating it “…beginning of the sixteenth-century, though a date in the late fifteenth-century is also possible”

Plate Five, “Milestones”

RK would date it early 17th for many reasons, not only the border shortcomings.

Plate Nine, a “Tintoretto"-type double niche rug, is one of probably hundreds made from the 15th to the late 18th century.

Plate Nine, “Milestones”

It is an interesting variant, both for its candy-cane coloration and large filigree escutcheon suspended between the central medallion and upper niche.

These features are for more interesting than archetypal, but it is the drawing in the niche that signify for RK thompson has over-dated this rug.

The two-dimensional and somewhat fuzzy character belies any possibility this rug could be “…late fifteenth-century or early sixteenth…

The same could be said for the main border, which though competently drawn lacks the salubrious mystique archetypal 15th century Tintorettos exude and a mid-17th century date is generous enough.

Plate Ten, “Milestones”

Plate Ten sets up an interesting comparison, as it is, in fact, actually a Tintorotto variant down to the similar escutcheon placed under the niche.

The Milestones text offers an “…end of the sixteenth century…” date, which RK thinks is over-dating to an extreme.

We’d prefer, instead, a more conservative and probable circa 1700 guesstimate.

Again the two dimensional quality of the field pattern, the visible vertical compression the ancillary octagon/stars in the field suffer, and the lack of any abrash in the red background of the field, all point to the later dating we suggest.

Over the years RK has spent some time analyzing how the pattern seen in the wide main border developed, as we formerly owned a late 18th century Anatolian village rug with a similar one.

We came to the conclusion this border is a very abbreviated version of the far more involved cloud-band on Plate 9.

Notice the minor element placed between each group of four “S” has been lifted from Plate Nine’s border, and, in fact, the curvilinear “S” motif and their perfect placement are actually a simplistic attempt to recreate the cloudbands from Plate Nine.

Left: border element Plate 9, “Milestones”; Right: border element, Plate 10, “Milestones”

This might, on first take, appear far-out but when you let your eyes relax on each group of those “S” motif, which are exactly and uniformly repeated in each group of four, they simulate the movement those cloudbands impart to the border’s overall design.

The lappets in both end borders” thompson says “ are probably the relic of a style fashionable in medieval times and preserved in rural communities.

This is laughable hogwash when thompson offers not one iota of evidence to support such a contention.

And besides didn’t thompson place himself on record, as we quoted this earlier, believing village weavers were too ignorant of the history of their design vocabularies to know their histories? If so, then, how could such a design remain potent for such a long period?

Let RK set dr thompson straight as the probable history of this symbol, and unless he knows what we know and for some reason does not want to mention it we must again say his statement cannot be anything but hogwash.

Archaic long-stitch embroidery, RK collection

We have published this embroidery several times, both on RugKazbah.com and on the Weaving Art Museum website, where we have stated our opinion the large medallion represents some type of astrological map or perhaps a calendar.

But no interpretation is needed to see the similarity the detail below has with the Plate Ten’s “lappets”

Left: detail long-stitch embroidery, RK Collection; Right: detail Plate Ten “Milestones”

This icon then, we surmise, shows up in later Turkmen weavings as a primary main border element.

Many long years ago RK coined the phrase “bat border” to describe this infrequently seen Turkmen main border.

Here are some examples

“Bat border” found on certain Yomud chuval

None of these are from an archetypal chuval and perhaps one day when one does appear the question as to whether or not this icon originated in Anatolia, Turkmeistan or elsewhere will have to remain unanswered.

We say elsewhere because at this point we are not so sure, as we formerly were, the long stitch embroidery is trans-Caucasisan or “Armenian”.

Plate Eleven, “Milestones”

Plate Eleven is an interesting rug that reminds RK of several other variants of the type. We have taken the liberty of illustrating it upside-down and not the way it appears in the “Milestones” catalog because we believe it is a prayer rug.

This becomes clear when it is viewed as we have presented it.

The idea of the implied mirhab is one RK has championed for a number of years and while we agree it is a toss-up with a weaving like Plate 11 we do believe our perspective is correct.

But regardless of whether or not it is a prayer rug, dating it “…earlier sixteenth century…” as thompson did is no toss-up – it is incorrect.

This rug’s use of ornaments that never appear in 16th century rugs and its borders, which are also unknown from that period, discount any possibility it could be that early.

RK would propose a mid-seventeenth century date.

More to come, stay tuned…

Home   Buy/Sell at the Kazbah   Terms Of Service