Home > Turkmen Rugs >RK reviews rageth's New Perspectives Part 7
Thu, May 12th, 2016 12:06:02 PM
Topic: RK reviews rageth's New Perspectives Part 7

Archetype soumak medallion khorjin front with heraldic beasts; RK collection

The third leg of rageth’s stance, no pun intended, is the art historical connection between various elements of Turkmen weaving iconography and ancient western pan-Asian design. As we have already made clear this is no new idea, and while rageth’s efforts publishing photos are more energetic than other researchers this does not make them better or more significant.

As for his arguments supporting those similarities? We have to conclude he provides no ‘new perspective’, as much of it can be seen as just verbose repetition of what others have already demonstrated, or ideas that have no chance of being correct.

For his efforts to collect and publish all those photo references rageth deserves credit. But just like the commendation we gave for the dye tests he organized and published it comes with qualification. Ours again lie with the interpretations he has made, his consistently overstating their importance, and his never ending presentation of opinion as fact.

In this installment of our ongoing extensive review we will examine whether rageth’s art historical analysis says anything new about the engsi, the most important and enigmatic of all Turkmen weavings.

Ancient Tekke engsi, gopaz type; published Weaving Art Museum “Turkmen Trappings” exhibition; RK collection

There are only ten engsi among the168 Turkmen weavings illustrated in color and black and white in this catalog. The two S-group examples are the best of the lot, the others not worth much mention as they are late and have no interesting ethno-cultural connection or unusual iconography. Therefore, and because rageth believes S-group weavings are the penultimate example of Turkmen woven art, the so-called Salor engsi, cat. no. 1, is a good place to begin.

RK needs once more make clear we do not necessarily believe Turkmen known as Salor made any of the weavings rageth and everyone else attributes to them. We prefer to refer them as either S-group, those with an asymmetric open left knot tied on a two level warp; pseudo-S-group, those with an asymmetric knot open to the right tied on a two level warp; and neo-S-group, those with different structural characteristics but designs and colors similar to the S-group examples.

And yes, we know, we just made up three new terms but they are highly descriptive, accurate, 100% provable definitions unlike the amorphous Salor moniker that is anything but.

S-group engsi, cat. no.1; type A; private collection

Because the engsi is the most iconographic weaving of all Turkmen woven products, the tentband a close second, it goes without saying it must have had a very special place in their weaving culture. The lack of any stylized and standardized gol, or what we call the grid-gol format that monopolizes almost all other types of Turkmen weavings, makes this more apparent.

Engsi are different, and little is known about them. Also their design elements, potent icons in the earliest examples and more ordinary patterns in the later ones, appear to have far greater connotation and connection to ancient, pre-existing, patterns than hardly any of the various gol-types suggest.

Engsi are also among the earliest known Turkmen weavings, and examples of great age are remarkably scarce. So scarce are the earliest almost none are published, causing them to go unrecognized by collectors. Often they are damaged, and this tends to make the subtle indications of great age and design significance appear as nothing special to the uninitiated. This could not be farther from the truth.

Speaking of the S-group engsi, RK does not accept the myth these weavings are very old, by this we mean earlier than mid-18th century. We have already mentioned their homogenous design pool, and their technical/structural differences. Also, their iconography is not nearly as complex and evocative as the early engsi of other groups, and their representational depiction of animals and birds completely different. Their flashy appearance compared to the reserved and quiet style of other very early Turkmen weavings is one more reason to doubt their antiquity or connection to historic Turkmen weaving culture.

No doubt S-group engsi are very rare, probably less than a dozen are known. However, even more rare are the archetypal engsi produced by other Turkmen groups. These in total number even less than that, and they are all different but do share some central iconographic elements. This is a comparatively important point considering all Salor engsi are very similar, and basically all the same age.

Although all S-group engsi are visually similar they are equally very different from all other Turkmen engsi. Their most exclusory features are the two representational panels, one of marching animals and the other a row of winged vulture-like birds.

Both of these designs are known from exceptionally early Anatolian weavings -- the animals from pile rugs and the vultures from kelim. However, they are totally unknown in other types of engsi, or any other Turkmen weavings. This is an important point along with their unique structural characteristic, the two level warp, convincingly implies they are somehow foreign and not really Turkmen products.

The particular stylized manner in which the S-group engsi animals are depicted, especially with one front leg raised, marks them as descendants of those beasts on the earliest of Anatolian pile rugs. The same association can be made with those on the soumak khorjin pictured at the beginning of this part. We have included it as a comparison reference to the S-group engsi animals, the soumak exhibiting an even earlier form much closer to the Anatolian originals.

In any event, the soumak is most likely no earlier than 1750-1800, a date RK feels it shares with all the S-group ensi. We do not believe rageth’s S-group c14 results, preferring to base our dating on the art historical implication the comparison with the soumak bag makes, as well as with other comparisons to Turkmen, specifically S-group, weavings we know. This dating does not play well with conventional thought, and while rageth wants to trace the origin of the Salor and all other types of engsi to western Asia, his compass is turned in the wrong direction -- Anatolia, not east, is far more probable at least for the S-group engsi animals and vultures.

Discounting the beasts and vultures, the rest of the iconography the S-group engsi displays is decidedly Turkmen, as we will now explain.

For more than a century Turkmen weavings have been passionately researched and collected. This effort has proven early engsi are the least encountered type of Turkmen weaving, fewer in number than even tentbands.

But as rare as early ones are a great proliferation, particularly those called Tekke, started in the middle the 19th century. Why did this happen? The answer, as we will show, provides some insight to help explain the S-group engsi.

Standard, common, Tekke engsi style, third quarter 19th century

This is an important question, one way past rageth’s comments.

S-group engsi and almost all those of the Tekke have the same vertical central panel with curl-leaf vine usually with a kotchak crown; white stylized ground curl-leaf vine with or without integral “s” icon in the border on both sides of the central horizontal panel; inner minor border of sainak icon; and rows of bifurcated “y” icon, or trellis-field as we have referred to it, in the four boxes that are the engsi field.

The S-group central horizontal panel is also found in Tekke engsi but only on an extremely rare group that are earlier than any of the standard, common types. None, however, as early as the oldest type of Tekke engsi, known as gopaz.

Rare Tekke engsi with central horizontal panel containing a row of curl-leaf “p” icon in small rectangles exactly like S-group engsi; 1750-1800; ex-RK collection

There is another slightly different S-group engsi, cat.no.2, type B.

S-group engsi cat. no. 2; type B; Marie and George Hecksher collection, San Francisco, Ca.

The star medallion and bracket iconography of the major side, upper and lower border of type B, its most defining feature from type A, can be found on a rare type of the earliest Tekke main carpets(MC). These MC also predate any S-group engsi. A coincidence? RK thinks not.

The archetypal Tekke gopaz engsi has the earliest, most detailed and best articulated versions of the curl-leaf “p” icon border we know, both in the somewhat different styled minor borders flanking the horizontal central panel and the standard white ground major side borders. This engsi is very early, well old enough to have provided the model for the S-group engsi stiff and rather lifeless curl-leaf that are sequestered in multiple small boxes in the central horizontal, and placed on top of each other in the vertical, panels that divide the field. Same goes for the diminutive white ground curl-leaf vertical inner minor border only present in type B examples.

Another part of the ancient curl-leaf iconography transmitted from the Tekke gopaz to the S-group engsi is “s” icon in the minor borders flanking the central horizontal panel, and the inner upper and both lower white ground minor borders of only type B. The “s” icon appears in a seemingly simple but actually very complex format the Tekke weaver achieved by carefully positioning pairs of curl-leaf to form the “s” icon in figure-ground between them. This dualistic balance of icon, curl-leaf “p” and “s”, was lost in the transmission process, the weaver of the S-group engsi only managing to weave a border of “s” icon, sans complicated figure-ground curl-leaf “p” and “s” icon.

The far more complex, animated, precise work, plus use of the complicated figure-ground drawing the weaver of the Tekke gopaz engsi was able to achieve hallmark the earliest Turkmen weavings, and the lack of such qualities definitely exemplifies later periods of production.

It is also not unreasonable to hypothesize the kotchak integrated within the intricate design of the gopaz engsi central horizontal panel also found its way, in a simplified form, into the S-group engsi configuration. This can be seen in the row of kotchak placed in a panel just above the upper elem of the S-group engsi. Interestingly, the rare Tekke engsi has the same kotchak panel in the same position. Another coincidence? Once more we think not.

Left: Detail showing kotchak panel from the S-group engsi; Right: Detail showing kotchak panel from the Tekke engsi illustrated above

So where are we going with this line of reasoning?

Basically at some indeterminate time in the past weavers of the first generation of S-group engsi copied some outstanding features from early Tekke weavings. Those we have mentioned from the gopaz engsi and the major border from a rare type of Tekke main carpet amalgamating them into the contrived S-group engsi format. Who were these weavers and why were they so familiar with Tekke iconography? This is impossible to answer, though we could guess they were trading partners of both the Tekke and the Salor, who saw an opportunity to create an item of extreme luxury (S-group weavings) to market to the wealthiest of their Turkmen contacts and had access to skilled weavers who could create it. The very fact there are S-group and pseudo S-group weavings that differ only in the type of asymmetric knot used, and in no other respects, supports contract weavers from different weaving backgrounds, hence the different knot they used, who were given all the materials necessary to create these luxurious but historically disconnected weavings.

That’s enough blatant supposition from RK, what isn’t at some particular time thereafter, as described below, elements of the S-group engsi format reappeared in the proliferation of later Tekke engsi produced from the middle to the end of the 19th century.

It is known about 100 years earlier, in 1822, some of the last remnants of the almost vanquished Salor Confederation joined with the Tekke of Akhal to attack Bokhara. It is logical this is the period when S-group engsi iconography became so influential in Tekke engsi design. And in fact the coloring of the rare Tekke engsi marks it as having been woven in the Akhal around this time fits well with it also displaying the S-group engsi static and stiff style of curl-leaf in its center panel and the standing row of kotchak in a lower one.

An important clue is the earliest engsi that can be attributed to the Tekke, the gopaz, has considerably different design elements than other Tekke engsi, implying it was woven long before S-group engsi elements became part of Tekke engsi iconography.

The Tekke gopaz engsi is so named for the unique stacked lyre icons in the major side borders. Some, like rageth, have erroneously called this icon a candelabra, see cat. no. 50, which we mentioned in a previous part of our review. However, the ancient example we illustrate clearly depict what look like strings, or a tuning mechanism similar to that on a guitar, at the top of both sides of the lyre. This feature is included in the few other known gopaz engsi like cat. no.50 but not quite as carefully articulated.

Detail Tekke ‘gopaz’ engsi showing lyre border and multi-arm candelabra elem; published Weaving Art Museum “Turkmen Trappings” exhibition, plate 3; RK collection

Tekke gopaz engsi cat. no.50 first half 19th century; private collection

Another question concerns the gopaz engsi elem panel of tripartite motif arranged in diagonal bands each with a different color, which are interpreted by other writers as flowers.

But are they really flowers or, as we believe the archetype example shows, multi-arm candelabra like the Hebrew Menorah?

We’d say the detail photograph above makes it obvious these are candelabra, not flowers. So all the other later ones must be as well, regardless their less detailed depiction.

A slightly similar, smaller elem icon with an actually quite different content appears on a group of S-group chuval, like cat. no.12, as well chuval of the Saryk, Kizil Ayak, Tekke and Ersari. The best and earliest of these appear to pre-date any S-group example.

Detail, early Saryk chuval with flower elem; ex-RK collection

These, however, are flowers and should not be confused with the gopaz candelabra elem. That said, their similar form, diagonal arrangement and color play is unique enough to be another design element S-group engsi have lifted from earlier Turkmen chuval.

There are basically two types of Tekke engsi, the very standard one with no notable or definable features but a number of differing minor ones, particularly elem panel decoration, and the animal-tree type with two small animals on either side of a flower- head tree repeated several times across one of the elem panels. But no example we have seen of either group is as early as the ancient gopaz engsi we illustrated. Plus, again, no S-group engsi appears to be as early as it either.

Astute readers might question what role the two outstandingly unique elements of the S-group engsi -- the large marching animals in the wide horizontal border within the field just below the four boxes, or the winged birds or vulture seen in the lower elem panel-- play in this scenario?

The marching animals most probably morphed into the far simpler animals in the animal-tree icon, but we have no clear idea what happened to the winged vulture. Its plausible the branched tree-like icon encased in their body is the remains of the skeletal tree found in the same elem position on a very small number of the earliest S-group chuval.

Left: Detail early S-group chuval with skeletal tree elem icon; Right: Detail S-group engsi vulture with body ‘tree’ ornament; ex-RK collection

We also know the brown ground of the S-group engsi vulture elem does appear on almost all Tekke engsi, however, it is decorated with rows of small multi-colored stars. This same brown ground elem with somewhat larger stars appears in the archetype gopaz engsi, linking all them again.

As for what role the S-group type A main border, which is usually repeated twice as it is in the cat. no.1, played in this transition? Undoubtedly the undulating vine and leaf and hash icon became much reduced, losing the curl-leaf to become the simple hash icon border seen with many standard Tekke engsi. The fact the hash icon alone appears nowhere else we know, except one rather late Tekke mafrash, but combined in the curl-leaf and meander border it can be found on the super rare bird and somewhat less rare animal-tree asmalyk. Examples of both types long predate any S-group engsi, and are decidedly Tekke weavings. These are more indications of the Tekke/S-group engsi connection.

Hash or tic-tak-toe border seen on some Tekke engsi

The above examples go good distance demonstrating the relationship between the S-group and standard Tekke engsi and that both contain earlier iconic elements from the gopaz engsi. The actual mechanics of any transition process are not known, but since no S-group engsi earlier than the 18th century has appeared -- RK believes only a scant few S-group weavings are earlier, cat. no. 15 is one -- it raises the question if indeed an earlier version of the S-group engsi ever existed. Or are they, as we believe, not part of the historic Turkmen weaving culture but rather contrived from it by foreign non-Turkmen weavers much after the fact.

Although we have no documentation, why do we believe this is the case?

It must be remembered no other Turkmen group utilized the difficult technique of knotting on a two level warp because the required weaving technology was basically limited to more sophisticated town and city workshop and atelier production. Normal weavers working with normal looms could not easily produce a weaving in this technique. Therefore, the noticeable lack of any really early, pre-middle 18th century, S-group engsi well implies there were none, nor any organized workshops or ateliers producing them before that time.

By the way this makes considerable sense considering the great internal societal changes and upheavals Turkmen tribes experienced until things relatively settled down, somewhat, by the middle 18th century. Another fact pointing to this possibility is the earliest S-group weavings are chuval and torba, no larger format weavings like engsi or MC are known. And probably will never considering the now 100 plus years of collecting interest.

Now then, were the makers of all the known S-group examples Salor or were they non-Turkmen contract workers producing these items for them, or other buyers?

From all the available evidence it does seem it was highly skilled professional contract workers, perhaps from Iran or India who were trained to weave on a loom setup with a mechanism to facilitate weaving with warps on two levels, working in a few small workshop or private atelier expressly producing the earliest S-group weavings.

This we admit is grand supposition but it does link together a number of factors to create an answer to a perplexing set of questions: Who wove the S-group weavings, why are there no very ancient ones, why are the earliest S-group examples bags and trappings, and why does the S-group engsi share many features of the earliest type of Tekke engsi, the gopaz?

The connection between the S-group engsi and Tekke was a two way street. An earlier inroad probably circa 1650-1700, when elements from ancient Tekke weavings were incorporated in the birthing of the S-group engsi. Then, circa the 1820’s a later transmission when features of the S-group engsi became part of the later Tekke engsi style.

What does jurg rageth have to say concerning the S-group engsi, rest assured nothing about connections with the Tekke engsi. Let’s now examine what he says about them and engsi in general.

As stated in the discussion on the origin of the engsi design in the chapter ‘The Turkmen ensi’ its concept might be very ancient, presumably ascribable to the high cultures of the ancient Near East..”

In that chapter rageth attempts to prove “…the ensi design is associated with sovereignty, with its roots in the cultures of the ancient Near East.” This far from demonstrated, poor reasoning and tautological paradigm render it absolutely unbelievable.

For instance “A first important clue to the antiquity of the ensi design is the use of the same complex-design plan among all Turkmens. This in turn suggests that the composition is older than at least the 9th or 10th century, the time of the formation of the Turkmen.” RK concurs the engsi format is ancient, but because all Turkmen use it does not prove this by any means. This universality carries no indication how long it might have been in use.

In addition, the ensi is not an ordinary door rug” RK definitely agrees with that but we don’t with what comes next “rather (the engsi) may have served the elite for representative purposes; as a symbol of status.” Were it an elite accessory, and there was a large elite class among the Turkmen from the nomadic clan khans to those in settled villages and towns, there would be more early examples than now known.

This might suggest ensi are rare, which surprisingly is not the case. There are many ensi in private and public collections, though most of them are from the 19th century, or even second half of the 19th century. Earlier pieces are rare.”

Here is another of the instances where rageth makes a sweeping highly questionable statement only to negate it immediately in the next breath. But more important than this penchant to discount his thoughts, his thoughts are wildly incorrect because he does not know enough about Turkmen weaving culture and their early weavings.

But even rageth’s limited understanding recognizes early engsi are very rare. This is an important central point, one which is part of our long publicized, but still far from proven, theory the origin of the engsi can be traced back to its original use as covering for the entrance of the shaman’s yurt. And as such it was laden with highly significant (and spiritual) iconography.

Sorry we cannot prove this any more than we could almost 30 years ago when we first proposed it. But the fact early engsi of any group are the rarest, least encountered Turkmen weaving makes a history of non-secular use very plausible.

RK knows how rare the archetype examples are, we have been collecting them since the middle 1970’s and so far know less than a dozen, some in our collection and others unpublished and unknown to most connoisseurs.

The dearth of evidence about the use of the pile woven ensi as a door rug is also significant” rageth states. But he does not say why, and although our answer lacks the same evidence our idea, it originally was used to cover the entrance to the shaman’s yurt, does suggest a far more plausible origin than rageth’s.

It is not surprising there is no mention of the shaman connection in any of the reports foreigners recorded about Turkmen weavings because these reports were made in the 19th century by which time the old animistic shamanic religion, its remote tenets and practices, had long been lost and supplanted by Islam and other forms of sreligious observance.

There is a corollary here, one that exists and can be shown viable throughout human history. Often important religious and spiritual ideas and icon remained part of consciousness and often in use, though their original meaning was no longer known or even associated with them.

Another point to note concerning our belief the engsi was a spiritual/religious object is the absence of any special engsi iconography on rugs and carpets meant to be walked or sat upon. No one walks on their most sacred icons, and the Turkmen were no different.

Therefore, the former importance of the engsi and its corresponding iconography to the ancient animistic, pre-Islamic, shaman religion of the Turkmen and their ancestors remained conceptually viable long after that tradition vanished.

It is likewise no mystery why the formerly sacred design iconography of the engsi eventually was completely forgotten and finally became a secular pattern associated with domesticity, as shown by the proliferation of engsi in the 19th century. This is a viable explanation for the engsi becoming a commonplace decorative article, when previously it had been the rarest type of Turkmen weaving.

Of course this negates rageth theory the engsi was a Near Eastern high culture article and design format. But let’s examine the proof rageth offers.

The ensi was reserved for the elite, Just as the luxury aq yup, the white tentbands for the wedding or the reception tent of the khan, had no practical purpose, but rather a purely representative function, such must have been the case for the ensi as well.”

Well, no proof or even any attempt at it there.

Dragging out a line drawing of a Turkmen yurt “Simpson’s drawing fig.1 shows not only a decorative tentband as a decoration of the yurt but also an ensi and a textile over it…” to support this claim rageth ignores this drawing was made in 1884 when any possible connections to what had been the historical use of the engsi was long forgotten.

Simpson’s drawing of the yurt from 1884

Besides the fact this image has been published many times before, the tentband is a design-less late example not worth mention in the context rageth is attempting to establish.

There is an interesting quote from Peter Andrews, a Turkmen tent specialist researcher rageth cities about the lack of information on the engsi as door rug he found in his contemporary field work “ I have never seen one used in this way, and only the old people remember how they should be used”. This is another proof of how meaning can eventually become lost though many generations.

Coming back to Simpson’s drawing rageth states “…the drawing shows further details which support the meaning and origin of the ensi and its design proposed here: a status symbol based on Ancient Near Eastern archetypes.” It does, really? But again no evidence the engsi is based on any ancient Near Eastern archetype is offered.

What rageth does do is proceed to dwell on the existence of another, this time a watercolor, picture of the same yurt with the same supposed Saryk khan standing in front of it. Far from showing anything ethnographic this supports ideas it was a completely staged scene for the artist’s benefit. How could it possibly have any ethno-historical value? And even if it were not staged the date, 1884, renders it so far after the time period of importance to this discussion it becomes valueless.

But rageth does not stop there. Going on to trace its geographic origin to the “…oasis territories of Central Asia… and because “These oasis districts correspond to the ancient cultural centers…” he infers this means the engsi was part of those early high cultures.

This is another tautology, ie a self-reinforcing pretense of supposed truth, not proof. In other words if you believe A, the engsi comes from the oasis territories of Central Asia, and B, these oasis were at one time in the past ancient cultural centers of high culture, that makes C, the engsi was a development of those high cultures, true.

This is a false argument that does not prove the engsi was a high culture invention, pattern, object, or had any association with it.

As for the historical origin of the engsi, rageth tries to suggest his belief in its high culture and not nomadic/shamanistic archetype is proven by the work of Ulrich Turck on the origins of the Anatolian kilim design. “In his work, Turck has shown representative architectural depictions of the Ancient Orient, such as the Mural crown or the City Gate, finding their way into the design repetitoire of traditional kilim weaving. Such designs can be found in kilims up to the early 20th century.

He then adds “It might be likewise with the Turkmen ensi design”.

This is on the surface laughable.

As someone who knows something about Anatolian Kelim RK can assure readers, and rageth, none of those designs, or anything else in Turck’s work appears in the iconography of the archaic group of Anatolian Kelim. And were Turck correct it would.

We also can answer rageth’s nagging question about the engsi “It is still not clear why this design did not find a wider diffusion and was used only in this narrow region, in spite of the fact that this region has been significantly involved in international trade since the 3rd millennium B.C..”

While no proof the ethno-historical role of the engsi was tied to shamanism exists, it is known many Turkmen became early converts to Islam, circa 800 AD onwards, and some others had different religious beliefs, like Christianity and Zoroastrianism.

By 1300-1500 AD this would have relegated shamanism, and any ties with the engsi, to only a very small percentage of Turkmen and might be the reason for the extreme scarcity of ancient engsi.

Being so hung up on trying to prove aristocratic and high culture western pan.Asian connections to explain the iconography on Turkmen weavings, rageth’s self-professed third perspective, it is no wonder he places the engsi in this framework.

To cap his evidence-less position he finally states the conclusion: The engsi was a “… representation the sovereignty in the form of a throne borne by the ruler’s subjects.

He then illustrates a group of line drawings showing rulers being carried on thrones from as far back as the late second millennium BC. But do any of these show an engsi pattern? Of course not, there are none rageth can dig up. End of story.

Then, as his last ditch effort, rageth traces at length the sainak icon appearances, and although there are many going back to the almost 5,000 BC, the sainak is just a part of the engsi iconography, tracing its history does not trace the history of the engsi. All the arguments and pictures rageth offers to prove the engsi was associated with high culture fail to supply any evidence this idea is factual. Starting on page 739 and continuing to page 779 he proves absolutely nothing but gullible readers will, no doubt, be led to believe different. We rest our case this book is not a success.

More to come, stay tuned

Author: john lewis
email: john_lewis@mac.com
Thu, May 12th, 2016 12:06:02 PM

RK Replies:

Glad you have enjoyed reading this, John.


I am still battling to fully comprehend this fascinating post.

My understanding was that although the central asian tribes were essentially shamanistic, the settled centres (cities) started absorbing muslim culture but then with the descent of the Mongolian hordes in the 13th century the shamanistic culture was "re-invigorated" under the Mongol religion - Tengrism - sky gods, eagles etc. This would certainly explain some of the iconography seen in Turkmen weavings.

I have also seen a theory that guls originated in the battle flags of the Mongols - which is why early weavings have a 3-dimensional "waving in the wind" - dimensionality, that is absent in later examples as the weaving skills degenerated. Older weavings seem to have been (deliberately?) woven with different gul heights. Whether or not these are best understood under the influence of narcotic substances I await the results of Professor Price's academic study. The idea that the winning tribe took over the gul of the losing tribe also seems to make sense e.g. Tekkie taking over the "Salor" gul.

Of course, we will probably never know but putting together possible scenarios as you have done is fascinating.

Home   Buy/Sell at the Kazbah   Terms Of Service