richard tomlinson

hi all

i cannot comment on this piece as i know nothing of these pieces.

from what i can tell, only ONE person has attempted to date this piece (jim allen)

i would be interested to see other people propose a date. steve? i believe this is your area? how about a tentative dating, based on your experience?

regards
richard tomlinson

Steve Price

Hi Richard

I don't feel comfortable about date attributions in general, for reasons that are probably becoming tiresome reading. Certainly, it isn't 20th century. It strikes me as older than most of the comparables that I've seen, whatever that means. Mid-19th? Early 19th? Earlier than that? If so, how much earlier? I have no answers.

I like the blue gul center very much, am kind of puzzled by the dramatic difference in colors I see on my monitor in the images of the front compared to that of the back.

Regards

Steve Price

Martin Andersen

Hi Steve

I wouldn’t say there is much colour difference between front and back on the piece (the extreme close-up I posted earlier in this tread was taken in bad lightning, and the colours are way off)


As I am sure you all know photographing rugs is rather difficult. It’s hard to capture the colours, and of course even harder to capture the feel and texture.
Here are 2 shots taken in sRGB colour space, they are perhaps a bit too saturated (hope you can imaging something in-between these shots and the first I posted)





I have 5 other Tekke Torbas which would conventional be dated as pre-Russian, one was discussed here http://www.turkotek.com/misc_00091/torba.htm (and I have a few other Turkmen Torbas which are post-Russian). The texture and feel of the Torba in this tread is extremely fine, thin, soft and shinny. Nothing like my other pieces, or anything Turkmen I have been close to. Even if this Torba should turn out to be mid-19th (which I of course certainly don’t think it to be ) I sure think it’s an exceptional piece. Well enough bragging We all love our rugs, and I suppose we are all here to learn and share. I still hope more material regarding the border, or otherwise comparative material, will be posted.

Regards & Merry Christmas
Martin

(late edit : I may have been a bit to optimistic regarding my other Tekke bags. I suppose the correct conventional dating would be "last quarter 19th" or perhaps "last half 19th" all of course very uncertain)

Steve Price

Hi Martin

I know how difficult it is to photograph pile rugs, especially Turkmen rugs, and come out with colors that are reasonably close to reality (which, itself, changes with the light source). The images you just posted look pretty much the way I expect nice old Tekke colors to look in photos. The colors that left me very uncertain about what the torba looks like in the wool are these:







The colors in the third image are very different than those in the other two, on my monitor.

I find much to like about your torba: your description of its tactile qualities, the lovely blue in the major gul, the handsome border. If there was a way to date it reliably, it wouldn't surprise me if it turned out to be very old, and I feel comfortable about believing that it was woven no later than the fourth quarter of the 19th century. But it could be a lot older than that.

Regards

Steve Price

richard tomlinson

c'mon steve !

that is a cop-out. saying it is not 20C is too easy.

is this piece LIKELY to be 18C or older? could you at least poke your neck out from under that tortoise shell and answer this question?

you surely must have some inkling.....

richard tomlinson

Steve Price

Hi Richard

Tiresome as it is, here's my position again:
1. It isn't possible to create criteria for dating rugs within a range for which no documented example exist.
2. For Turkmen rugs, there isn't a single example of a rug that has a documented age within any narrow window of time between about 1700 an about 1875.
3. I'm confident that Martin's torba wasn't made much later than 1875, certainly not after 1900. I doubt that it predates 1700, although if C-14 dating gave a 16th or 17th century date I'd take it seriously.
4. I'm unable to be more specific, even with vigorous prodding. I don't think anyone else can do it, either.

That's my story, and I'm sticking with it.

Regards

Steve Price

PS - The rug was, in fact, woven at some specific time. We can't determine what that time was by a majority vote, as democratic as that process might be.

Martin Andersen

Hi Steve

I understand and respect your general sceptics regarding dating of rugs. And sure enough, dates are far too often presented as fact and not as estimates.

But personally I think you are too strict in your sceptics. Just to follow your argumentation to the extreme: one could say it is not possible to date any rug expect the ones which are c-14 dated, or those were there exists a specific recorded documentation of production time of the exact rug in question (the last would leave us up far after world war 1, and far after the production of any collectable Turkmen rug (with the exception of any possible earlier exact photo identifications)). This goes of course only for rugs without synthetic colours. A rug with synthetic colour could of course not have been produced before the introduction of the colour. But theoretically people could still somewhere be weaving like they have always done, with no development, with no deteriation from the 15th to today. With any rug without documentation or synthetic colour you would not be able say anything else then: it was not woven today (or perhaps even more strict: not woven tomorrow).

Personally I believe qualified estimates are possible (as estimates not as facts), and if you zoom out a bit in your timeline we do actually have a range where documented samples exist. It is not from about 1700 to 1875, but it is from the Hoffmeister Torba to the introduction of the synthetic colours. It’s of course a very fragile range, and of course estimates on this background can not be anything else then very rough estimates. But even if we did have a documented piece for every year we still would have to estimate within tons variations for every undocumented piece.
If we want to differentiate among the rugs, I cant see any way around qualified estimates in time and development. I may be naive but experience, knowledge, education, discussions and other human stuff like that is what I hope can be the basis of trustworthy qualified estimates.

No offence, and as I said I do respect your sceptics and the foundation they have in your scientific background.

Regards Martin

Steve Price

Hi Martin

In fact, it isn't possible to date any rug with certainty. Even with those having complete provenance, our confidence would be limited by the confidence that the provenance is accurate. That is, the attribution will always be a probability statement, although the probability could be very high.

Carbon dating gives a range within which there is a high probability that the correct date exists, so a C-14 date is just a probability statement, too. So is every statement that we might call scientific truth. I'm not at all bothered by that, but most people don't understand it and find it distressing.

Now, with that behind us, where does that put me in the dating game? My position is simple, actually. If there are no verifiable criteria for making an attribution, no attribution is possible. If criteria exist, then they can be applied and the conclusions will be probability statements based on them.

Since C-14 dating provides dates with 50 year windows that are highly likely to include the actual dates, such rugs can form a database from which we can try to extract age-related criteria on which to base attributions. Sadly, C-14 is useless for anything woven from wool sheared during the past 300 years or so. This means that if we're careful and work hard at it, and if there are enough C-14 dated rugs with which to do it, it's possible in principle to generate criteria for placing rugs at various approximate points between, say 1300 and 1700 AD.

There are lots of Turkmen rugs known to have been woven around 1900 and throughout the 20th century. Most collectors are familiar with them and can easily distinguish a probably-ca 1900 Turkmen piece from a mid-20th century Turkmen piece and from pieces older than ca 1900. Again, always with the understanding that these attributions have a small, but finite probability of being wrong. The troublesome area is from about 1875 on back to about 1700. Having no examples with documented ages, criteria pointing to specific 18th or 19th century intervals can't be generated. That might be unpleasant, but it's true.

Sadly, things are even worse than that. The number of C-14 dated Turkmen rugs is pretty small, which subjects using them to generate criteria for attributions to great uncertainties. To understand this more easily, consider the Pazyryk rug. It has a very high probability of being about 2500 years old, and its physical and technical characteristics are thoroughly documented. No sane person would think that comparing its properties with those of 16th or 17th century rugs would give us criteria for assigning ages during the period from 2500 years ago to 400 years ago. The Pazyryk rug is just too small a number of rugs from which to derive reliable generalities. To add an exclamation point to this, note that Jim Allen tells us that the 16th century (by C-14 dating) Salor pieces he's handled are virtually indistinguishable from Salors generally attributed to the 18th or 19th centuries.

The alternative to making guesses about the unknowable is to simply accept them as unknown and hope that something will come up to change that. I have no problem with accepting rough estimates that place rugs within ranges in which rough estimates are possible. I object very much to rough estimates being presented as facts (statements with high probabilities of being correct) that are pretty precise, which is common practice in Rugdom.

Regards

Steve Price

Jim Allen

In my experience with Turkoman rug sellers are vastly more “secure” in their provenance and dating than are buyers. Steve is a buyer and he is also trained in logical analysis. A book could be written about the duality of buying and selling rugs. Both Gurdjieff and Ouspensky were rug dealers on the side during their careers. I think these two “philosophers” were the prototypical rug merchants for generations to come.

There is a misconception amongst most Turkoman collectors that states in general that the better the color, space, and weaving virtuosity of the weaving the older it is. This was a workable hypothesis until people like me started testing Turkoman weavings with signs of great age and advanced aesthetics. We found out that 15th, 16th, and 17th century Turkoman weavings still existed and were dispersed among all the others across the globe. There are generalizations that can be made about 18th and 19th century weavings that are backed up by their relative values. The assurance of the accuracy of dates between 1700 and 1875 isn’t high but the generalizations follow along avenues of rarity, fineness of weave, the excellence of the dyed wool colors, and what is known about the relative fortunes of the different tribes decade to decade during this entire period.

Developing any confidence about such generalizations requires connoisseurship and this requires time and money to acquire. It takes money because it requires travel and time off to study the world’s museums holding Turkoman weavings. It almost always requires a reservation to do this at any given museum.

In conclusion it isn’t wise to lump all Tekke torbas woven between 1700 and 1875 with good natural dyes into one pot. One will find that all major factors vary decade to decade. The major factors are water quality, dye quality, wool quality, and a tribe’s relative security. If you have seen MONGOL recently you have some idea of what security really means in the nomadic context. Interestingly Jon Thompson has recently been saying that even the quality of spaciousness and its opposite crowded aesthetic runs in cycles in Tekke weaving during the period 1700 to 1875. He doesn’t actually spout those exact dates but that is the period he is talking about. This is an extremely complicated game and it takes many years to manage much confidence in it.

Martin PLEASE have some museum professional acquire enough weft threads to do an adequate C-14 study. Do not do it yourself or nobody will believe you. Raguth has said he lost some confidence in the European lab he was using. I know absolutely the big museums mostly use Jull at Arizona State U. The cost isn’t too much. I have an Anatolian rug I want to test. Perhaps you and your fellow collectors can come up with your best pre 1700 AD rug bets and we can get a full run. A full run contains 8 samples or it did years ago and costs less. I still have weft threads from some very famous classical carpets that could be used for a ‘known’ to test the accuracy of the technique. I have done this before and Jull’s dating was within a ten year window and that is fantastic!

Maybe Steve can keep this thread in limbo after he decides to archive it so we can revisit this topic when Martin’s C-14 testing comes back. If his Tekke torba is pre 1700 AD then the generalizations I make from it will be more interesting. Jim Allen

Steve Price

Hi Jim

I can live with generalizations if they're based on specifics. Otherwise, they're just affirmations of faith. The claim that age attributions can be made with moderate accuracy on the basis of rarity, fineness of weave, the excellence of the dyed wool colors, and what is known about the relative fortunes of the different tribes decade to decade during this entire period is a generalization of generalizations. Let me ask you to sharpen them up a bit.
1. Rarity is obvious enough. It means "not many like this one." Surely, you aren't saying that if a piece is nearly unique, it must be terribly old, are you? If you are, just how old would being nearly unique suggest? The Pazyryk rug is unique, at least until another one is discovered. It's 2500 years old. What does knowing that do for us when presented with something as unusual as, say, Martin's torba?
2. Translate "excellence of the colors" into a time continuum with reference points on it derived from documented examples. Likewise for fineness of weave.
3. What's the best track record on having his attributions confirmed by documentation, for any dealer? That, after all, is the only possible test of the hypothesis that some dealers have a special insight that makes them able to do accurate attributions at a subconscious level.

I agree that many dealers are more comfortable about their date attributions than most collectors are. They aren't disinterested third party observers, after all. But the fact that they're confident has no bearing on whether they're right. Every orthodox Jewish rabbi, every Roman Catholic priest, and every fundamentalist Muslim imam is 100% certain that his religious beliefs are correct. They can't all be right.

Regards

Steve Price

Jim Allen

Steve wrote, "I agree that many dealers are more comfortable about their date attributions than most collectors are. They aren't disinterested third party observers, after all. But the fact that they're confident has no bearing on whether they're right. Every orthodox Jewish rabbi, every Roman Catholic priest, and every fundamentalist Muslim imam is 100% certain that his religious beliefs are correct. They can't all be right."

That was my point. I used the word comfortable because it reeks of self interest and lacks authority. As for the question, "where is the beef"? This will take some time and effort. You started to dissect my statements by taking them one at a time using the example of rarity. Rarity is an important finding when coupled with fineness of weave and complexity and uniqueness of ground color. This is where a fine sense of tribal history comes in to modify what we expect from a tribe in terms of material qualities. A poor tribe marginalized by a stronger tribe or confederation does NOT HAVE access to the things that make an extraordinarily fine and beautiful weaving. Think Chodor! I have mentioned here many times that late 18th century Tekke 6 gull torbas are almost universally ugly due to their muddy colors, non traditional "Yomud like" borders, and generally lower knot counts. The Tekke were marginalized by the Yomud during the entire second half of the 18th century. This is but one example of how history and access to or interference with the best materials results in identifiable weaving characteristics decade to decade. We also need to talk about the fact that Pinner cataloged about 20 types of Tekke ensi leaving us with the impression these were clan based differences. If that was true for ensi it should be true for torbas and main carpets. I am able to discuss two main types of Tekke torbas based on knot ratios and their result on main gull appearances. There are two groups of Tekke torbas in terms of red versus purple red ground colors. That makes four main distinctions right off the bat for Tekke torbas. If we really gathered a big sample we might be able to make even finer distinctions. There were 40 Tekke clans in the middle of the 19th century. Let's see where this goes from here. Jim

Steve Price

Hi Jim

... late 18th century Tekke 6 gull torbas are almost universally ugly due to their muddy colors, non traditional "Yomud like" borders, and generally lower knot counts

Let's assume that this is actually a group in the usual ruggie sense - that they represent the output of a specific assembly of people during a particular period. My questions are, as usual, how do you know that this group is Tekke in toto rather than, for example, one of the 40 or so Tekke clans? How do you know that they are late 18th century pieces?

If, for example, I hypothesize that one of the Tekke clans wove ugly torbas until they were sent off to live among the lepers or the (shudder) Chodor, could you provide evidence that the hypothesis is incorrect? I haven't got a shred of evidence that it is correct, and can't think of a way to look for any. For that reason, I reject it. Do you have evidence for your hypothesis? If not, what makes it preferable to mine?

The notion that the economic fortunes of a tribe (or clan) affected the quality of their weaving is interesting, but not self-evident. Are there documented examples in other ethnic groups in which the quality of the crafts was correlated with the prosperity of the group at different times? If there are, that would support your hypothesis.

Regards

Steve Price

James Blanchard

Hi all,

This is a fascinating thread on many levels, particularly the mystery and beauty of the subject weaving.

Although my knowledge base is inadequate to offer anything more than impressions about the age, I think there a couple of more general issues that are perhaps relevant.

Steve has consistently pointed to the lack of a database of weavings with known dates. I would go one step further and say that I think there is a lack of a sufficient database, with or without dates. If we had a more extensive database we could at least begin the process of trying to sort them into groups based on various characteristics (materials, manufacture, palette, design, etc.). The first test would be the reliability of this approach. In other words, could a group of knowledgable rug experts actually agree on clusters? If not, the issue of the validity of dating is moot, and we might as well all admit it and enjoy the rugs on other grounds. However, if knowledgable experts can consistently assign weavings to groups, the next step would be to try to assign age and attribution. This would require a persuasive process of inductive reasoning, and I am not certain that the knowledge base will be sufficient for that. However, if the groups are arrayed in presumed chronology and there is some way of objectively dating the earlier ones (e.g. C-14), we might get somewhere. However, knowledge in this area is only going to proceed slowly and uncertainly as long as efforts to systematize the knowledge are delayed. If, instead, focus is placed on a single weavings here and there, it will take a long time before we even know whether accurate dating is possible, based on our usual cues (design, colour, materials, etc.).

James

Steve Price

Hi James

Within the Turkmen, there are decent numbers of pieces that are documented as to tribal origin, having been bought in situ. The Rickmers collection, the Russian collections, and so forth, all come with pretty good provenance records. Identifying Martin's torba as pre-20th century Tekke involves rather little uncertainty. The common criteria for separating Turkmen weavings into half a dozen or so major tribal groups are probably very reliable, despite the occasional ambiguous piece. Serious problems don't usually arise until someone tries to assign things more finely. There's a thread on this issue with Yomud subdividing in our Miscellaneous Topics section, by the way.

Regards

Steve Price

James Blanchard

Hi Steve,

I was thinking more about dating pre-1900 pieces, and especially earlier pieces, more than tribal attribution...

James

Martin Andersen

Hi All

I find Jim’s thoughts on the relation between weave quality and socio-economic fluctuations very plausible. We all know what dramatic consequences for both design and quality the Russian invasion had in just a few decades. But then again, figuring out theses complicated historical relations certainly seems like an overwhelming task.

Hopefully in time the digital revolution in which we are right now may help us. In a few years the internet might provide us with means to establish at least a complete visual database of what material has been photographed and scanned. Some kind of visual Google search on f.ex. the Hoffmeister border may not be far away.

Regarding the dating of my own Torba I personally design wice find both pro and contra for a very early dating. I am not sure if it gives any meaning to elaborate further into that now, its all rather speculative. I will look into the possibilities of a c-14 test here in DK. And I will be back here with the results. Whatever the results may be, they will be interesting, either placeing a lot of design elements ealier then 1700, or placing the Hoffmeister border as a possible variant between 1700-1875.

Jim‘s suggestion on an 8 sample full test run surely also sounds very interesting, hope someone else will join it. As far as I have understood a c-14 these days only requires 1 mg material for the tests, there certainly is loose weft-material around the edges on my Torba for that

Regards
Martin

Martin Andersen

There are of course a lot of variations in the layout of the Chemche Gul, but in general it looks something like this:



The Chemche Gul on my Torba doesn’t have the ram’s horn in the ends:



I have been through my books, and searched around the net looking for another Chemche Gul without the ram’s horn. I haven’t found a single one. I would be extremely interested if any of you could help finding one.

If my suggestion, that the Chemche Gul is derived from the internal geometry of the main gul, is correct, then it is possible that the rams horn design-wise is a later addition. And then the layout of the Chemche Gul on my Torba may not only be rare, but also close to the original prototype of this secondary gul.

Any thoughts on the development on the Chemche Gul are much appreciated.

Regards
Martin

Steve Price

Hi Martin

Scroll down to Number 35.

http://ne-rugsociety.org/gallery/bags/bags-navframe.htm

Regards

Steve Price

Martin Andersen

Hi Steve

Thanks, I post plate 35, from your link. It’s a Yomud, but certainly a beauty anyway


plate 35


plate 35 detail

A drawing like this Chemche Gul underlines that this Gul is a filling design of the negative space between the main Guls. And of course there is coherence between positive and negative space in this kind of geometry.


And searching the Turkotek achieve I once more draw on Jims finds.
In this tread http://www.turkotek.com/misc_00090/anatolian_turkmen.htm Jim posted this Karapinar.


Karapinar


Karapinar detail

The detail is interesting, and for me gives credence to a direct structural connection between the main Gul and the Chemsche Gul. And there’s no rams horns on this one either. Speculation of course, but I would think that the mother of the Chemche Gul was without horns

Martin

Edit dec. 31.: The Yomud seems to be in general more experimenting with the secondary Gul than the Tekke. Actually there are quite a few Yomud versions of the secondary Gul where the resemblance is close, but where I wouldn’t be sure if Chemche Gul would be the correct term. There certainly are samples like the one Steve linked to where it is the Chemche Gul without horns (or bows?).
But I still haven’t found any other Tekke Chemche Gul without the horns than the one on my Torba.

Steve Price

Hi Martin

Here's Jim's 17th century (by C-14 dating) Tekke juval.



It has chemche guls, and they have the ram's horn(?) extensions on them. It follows that the gul was invented by the 17th century. Perhaps the version without ram's horns was a subtribe, clan, or family variant, perhaps it was something used only briefly. Who knows? Not me. But the small amount of evidence that exists (Jim's juval) doesn't suggest that it's an ancient archetype.

Regards

Steve Price

Added Note: This isn't Jim's C-14 dated juval. It belongs to Yon Bard, conventional criteria would place it in the mid-18th century.