kilim fragment

Horst Heinzlreiter · November 25th, 2009, 03:21 PM

Dear forum,

let me show you a kilim frag, which is not mounted professionally yet. It comes from the late Tom Weisbuch (California).
Would be great, to get some ideas about this stuff, origin, age etc.

Thanks in advance
Horst

Doug Klingensmith · November 27th, 2009, 02:01 PM

Wonderful old fragment Horst. Thanks for the look.
My uninformed amateur first impression is Central Anatolia (Konya?)
d.k.

Pierre Galafassi · November 28th, 2009, 02:45 AM

Hi Horst,
I haven't got the foggiest clue about kilims, but met the two fellows below, which might be kin to your beautiful kilim, in Orient Rugs. Volume 4. Turkish. Pages 110 & 122. K.Zipper and C. Fritsche.

Both are said to be from central Anatolia, Konya area. XIX century.The first one is called Karaman by the authors. A once strong turkish tribe beaten by the Ottomans.
Hope that helps.
regards
Pierre

Horst Heinzlreiter · December 2nd, 2009, 09:50 AM

Thank you for the pictures and your guess about the origin of this kelim.
The striped border on the pics,which surrounds the main motive,
is really similar to our kelim,
but I have not seen this big hooked ends on any other kelim,
did you?

Once again,thanks to all!
Horst

Joel Greifinger · December 3rd, 2009, 10:41 AM

Quote:

but I have not seen this big hooked ends on any other kelim

Hi Horst,

I'm not sure that this is what you are referring to by "big hooked ends", but here are two kilims from the Konya region published in Petsopoulos's Kilims.

Joel Greifinger

Jim Allen · December 6th, 2009, 05:54 PM

I spent some time viewing the kilim exhibition set up in conjunction with the Hajji Baba club's 75th anniversary. This was the kilim collection of Mr. Wolf and I discovered one kilim very much like yours with a similar archaic drawing that he had C-14 dated to the 15th century. There is every reason to believe that your fragment is that old. One thing was made clear by Mr. Wolf, if you got him to talk about the C-14 dates like I did, was that his kilim collection spans 500 years and that kilims that old still circulate in the world. He still collects! This white ground type of Konya area kilim was one genre that seems to have surviving examples from every century over the last half a millennium. Let me also say that the very best restorers in Turkey can bring that kilim all the way back to look like it was never damaged. I don't know if I would do that but then I don't own the kilim. I would be surprised if that fragment doesn't date to the 15th or 16th century after comparing it with what I saw of Mr. Wolf's white ground Konya collection.

Joel Greifinger · December 6th, 2009, 08:03 PM

Quote:

This white ground type of Konya area kilim was one genre that seems to have surviving examples from every century over the last half a millennium.

Quote:

he had C-14 dated to the 15th century. There is every reason to believe that your fragment is that old.

Hi Jim,

Given the fact that iterations of white ground Konya kilims survive from every century in the last half a millennium, could you explain why, based on the evidence at hand on your computer monitor, you can say that there is "every reason to believe" that Horst's fragment is from the 15th or 16th century?

Joel Greifinger

Jim Allen · December 6th, 2009, 11:07 PM

Their white ground Konya was hung near the window in front and the date was not attached to the label. Most people just thought it was a really nice old Konya kilim. I was there late and heard Mr. Wolf expound on the piece and say how surprised he was that it dated to the 15th century. His kilim had the delicate and beautiful colors Horst's piece has. The drawing was very well thought out as per the negative space patterning. Look at the negative space patterning between the big medallions in Horst's example. The negative field pattern (kochank) is actually about as strong and big as the medallions we 'automatically see as the main design ornament' are. This tendency was as strong in Mr. Wolf's kilim as it is in Horst's. So it is the pallate, the depth of design patterning, and the great simplicity of design that gives me the impression that Horst's kilim is so old. I would love to see Martin's torba dated with this kilim. I have a yellow field Konya or Karapinar yastik I would like to add to the run. That leaves 5 empty cells as a run contains 8 samples. Anybody else out there think they have an unidentified classical period weaving to test? You get a discount on an entire run, or at least one did years ago.

Louis Dubreuil · December 7th, 2009, 06:03 PM

Bonsoir à tous

I think this kilim fragment a "very good one". The composition is simple and majestic. The field is clear and not crowded with too many motifs. The design is very well done with a singular clarity. Colors are subtle. The blue is perfect.
This piece displays all the qualities that make a very great old weaving.
I see in the medalion design a very realistic and bold representation of a tortoise, that could have been either the "totemic" animal of the tribe who wove this piece or also a great symbolic picture . I agree with Jim about the perfect negatifve/positive game we can see in the white kotchak design made by the space between the "legs" of the tortoise .

All those qualities are more frequent in older pieces. The fact this piece could be XV° is likely. It is certainly not a XIX° production.

Louis

Horst Heinzlreiter · December 8th, 2009, 05:09 PM

Good day,

thanks you for your stunning comments, never thought this kilim as old as
your given age, Louis.
Would be interesting to take a C 14 age determination. Will see.

We believed it an 18th century piece, because of it's powerful gaphics and
colors.

Your replies led me to my kilim books - there in Vok's "Anatolia", plate 70,
I discovered a kilim, with a similar designed motif in the center (please have a look at the pic below). Vok's kilim is dated to the 17th c.

I found a few others, which seem to belong to the same group, for example
B. Frauenknecht, Anatolische Kelims, plate 17

B.Frauenknecht, Early Turkish Tapestries, plate 27, dated 1846

Konzett, Gewebte Poesie, plate 54, about 1800

Konzett, Gewebte Poesie, plate 55, before 1800

Y. Petsopoulos, Einhundert Kelims, plate 76

Best regards Horst

December 8th, 2009, 05:27 PM

Dear Mr. Heinzlreiter,

I was just about to add some of your most recent references when you posted. I think the most one can say is Central Anatolia perhaps closer to Sivas. There are lots of ivory ground kilims from greater Central Anatolia so one should be careful, not all are necessarily related in design. The main point I wish to offer is a comparison to the piece from the collection of David Lantz of New York. This is plate 33 in Rageth's Anatolian Kilims and Radiocarbon dating. Also there are close comparisons with three other kilims including plate 34 from a "private collection". In these pieces the C wrenches extend out like yours. Lantz's piece is dated 1427-1692 with 86.4 confidence limit. Characterized by large C-shaped pendants at the hexagons Rageth writes the secondary motifs (small hooked hexagons) always form a 4 + 1 composition together with the primary motif while the large hexagons form a center on the vertical axis - as does your piece. The border design of the Lantz piece corresponds to the secondary field design - also as yours.

I think your piece could be 17th century though it is hard to see the colors clearly on monitors. In any event Tom Weisbuch was a strong collector with a keen eye and you must be very pleased to have such a evocative kilim from his collection.

Best wishes, Michael Wendorf

December 8th, 2009, 06:30 PM

Dear Horst,

It's a very nice kilim you have. This is a type that was produced for quite some time in Central Anatolia. Here is the earliest example that I'm acquainted with. This is published as Plate 33 in Jurg Rageth's book, ANATOLIAN KILILMS & RADIOCARBON DATING. And here is the calibrated age of this piece with 95% confidence limit:
1427-1692 AD (86.4%)
1728-1815 AD (9.8%)

In later examples, the central figures tended to become a little wider and more squat, and the outlining bands bolder, so that these figures visually filled the space more completely. As the aesthetic gradually changed, the figure/ground/border proportions altered a little. Everything became more compact, with less space around the secondary hooked medallions. The color palettes generally became a little warmer. If you would like, I can post some later examples, though radiocarbon dates on these are not reliable.

Best,
Marla

December 8th, 2009, 07:19 PM

Horst,

Here is a later example of a kilim with the same motif from Rageth's book--an example carbon dated :
AD 1690-1730 (19.2%)
AD 1814-1924 (80.8%)
Of course we have to realize that most authorities put little faith in carbon dates on pieces as late as the 19th century.

And a third example from Rageth's book, with no carbon dating, but listed as "presumably from the 2nd half of the 19th century."

Horst Heinzlreiter · December 9th, 2009, 05:03 PM

Thank you all so much for your inputs.

@ Marla you were the guide - you made my day!

@ Michael: Yes, it would have been a great thing to see all the pieces
of Mr. Weissbuch's collection. We are proud of the kilims we were able to catch.

To legitimate my reply, please let me show an other kilim fragment:
It is mounted, dimension - 222x157cm and probably Yoruk?
Best regards Horst@Helga

December 9th, 2009, 05:18 PM

Dear Readers,

Thanks to Marla Mallett for supplying images of plate 33 and related kilims in Rageth's Anatolian Kilims & Radiocarbon Dating. It is an interesting group that spans many generations.

The image that Horst found from Vok's collection has also been radiocarbon dated. That piece came back with 1469 - 1673 with 95% confidence limit and is also plate 26 in Rageth's book and appears also at page 79 in Hali 92. The special feature of the ivory ground Vok kilim, according to Rageth, is the design of two large medallions with another drawing between them, this is a varient of type with 3 large medallions of the same size and design. The 3 large medallion type is much more common. The Vok varient is probably related to another well known piece also published by Vok on an orange ground (Plate 28 in Rageth's book).

Horst's kilim and the Lantz (plate 33 in Rageth) and other pieces published by Rageth belongs is probably not a varient but is it's own group. The organization is different than the 3 medallion type although both designs often share an ivory gound.

A word about color... it's dangerous to draw too much from computer monitors. And it is dangerous to make too many conclusions even when you are confident of the colors. One generality that I remember Rageth making after handling many kilims for his dating project was that the few kilims that date from the 15th to the 17th centuries show a homogeneous, restrained palette. The first changes can be seen in the first half of the 17th century and become more pronounced into the 18th as the spectrum of colors changed. You have to look closely at many kilims to understand what he is talking about. Then you must add to that deep knowledge of the design pool.

Again, congratulations to Horst for his wonderful kilim fragment.

Michael Wendorf

Joel Greifinger · December 10th, 2009, 08:20 PM

Hi all,

Here’s a piece published in Bohmer’s Nomads in Anatolia that holds a notable resemblance in design to Horst’s initial fragment. Bohmer does not attempt to assign a date of production.

This clearly looks to be one of what Michael Wendorf referred to as not a variant of the “three medallion type” but rather its own group.

I am rather struck by the confidence with which age attributions have been made at times in this thread. Jim declared that there is “every reason to believe” that Horst’s fragment is from the 15th or 16th century. Louis writes that it is “likely” from the 15th century based on a “simple and majestic” composition, a field not crowded with too many motifs and a” singular clarity” to the design.

Michael Wendorf cautions to not draw too many conclusions as to age from color, especially on a computer monitor and that judgments need to be bolstered by a “deep knowledge of the design pool.”

Since my knowledge of the design pool is unfortunately rather shallow, I am hoping that other contributors can help me out. Is this kilim from Bohmer in the same design group as Horst’s fragment, as it appears to be? Do you see this as also likely to be from the 15th or 16th century and, if not, why not? If color is a crucial factor in the differentiation, can you say more about how you draw the distinction?

Thanks.
Joel Greifinger

Steve Price · December 11th, 2009, 08:40 AM

Hi Joel

I agree. In the absence of a substantial database of pieces with documented age, generating reliable criteria for age attributions is impossible.

Rageth's book (which I haven't seen yet) apparently includes a number of kilims with C-14 dating, and this could be a pretty good starting point. On the other hand, unless things have changed a lot since the last time I did any serious reading about C-14 dating, the method is useless for anything less than 250-350 years old. For that reason, Rageth's inclusion of pieces C-14 dated to the 19th century disturbs me. Perhaps methods have been developed for overcoming the problems introduced by the industrial revolution; if so, I'll be more comfortable once I know what they are.

Maybe the criteria that have become conventional wisdom are correct. Maybe not. Until we really know, I prefer phrases like "this piece fits the criteria ruggies generally attribute to the 15th century" to phrases like "this is a 15th century piece."

Regards

Steve Price

Louis Dubreuil · December 12th, 2009, 04:34 PM

Hello Joel

In my opinion the Bohmer's example looks like a copy of an old kilim. It could be likely XX°.
The design is stiff and mechanical, the colors are without any vibe. I think this is a good example of how uninteresting can be a soulless reproduction.

I have'nt said that the kilim was XV°, but that it was likely XV° comparatively with a XIX° hypothesis.
In my opinion, among the other old examples of this group shown by Marla, this fragment displays the best design.

Cordialement

Louis

Joel Greifinger · December 12th, 2009, 05:38 PM

Hi Louis,

While he doesn't give a specific age estimate for the kilim I posted, Dr. Bohmer includes it in a section of Nomads in Anatolia devoted to pieces woven "in earlier centuries." Although the book was originally published in 2004, I don't think he meant to include the 20th century in this category.

I think we would have to agree that Dr. Bohmer has what Michael Wendorf referred to as a "deep knowledge of the design pool." His assessment of the design and colors seems to have led him to the conclusion that, rather than a 20th century "copy", this was to be included in a group of kilims that were, "more beautiful than their successors" based upon "the vitality of their colors, their boldness, their rhythm, and the variety of their forms."

It isn't surprising that knowledgeable folk will disagree about the aesthetic merits of a given artifact. It's also evident that in rugdom the attribution of age, like beauty, is to a significant degree, in the eyes of the beholder.

Regards,
Joel Greifinger

Steve Price · December 12th, 2009, 06:55 PM

Hi Joel

... in rugdom the attribution of age, like beauty, is ... in the eyes of the beholder

Absolutely correct, and a serious problem. Beauty is, by its nature, subjective. The age of a rug is an objective reality, and the nearly universal position ("I like antique rugs. I like this rug more than that one. Therefore, this one is older than that one.") is as pervasive as it is absurd.

Regards

Steve Price

Louis Dubreuil · December 13th, 2009, 07:01 AM

Bonjour à tous

Yes there is always a subjective factor in weaving appreciation, but the exercice consisting among a group of weavings of the same design in classifying them from the best to the poorest can be "standardized". The appreciation is based on composition, colors, balance...In a weaving group that can have made a type of design during a period there are good weavers and medium and bad. The woven products are generaly homogenous from the point of wiew of waving technic (same finess, same wool material) but can be different from the point of view of colors (the skillfulness of dyers can differ from one to another) and the point of view of design. This latter depends of the aesthetic capacity of the weaver. Traditional weaving art is a "collective" creation under the influence of the group that is the keeper of the "model" through the tradition and through the mother to daughter transmission, but it is also an "individual" creation because the result depends of the capacities of one person : design execution, choice of the colors and of the balance between design, composition, colors...
This is exactly the same thing in traditional music : few persons are able to reproduce correctly the traditional model and more few are able to enrich the music with their personnal contribution or to play the piece with the "perfect spirit".

In kilim and rug making this is the same thing. But there is also an other point to take : a specific design has a proper "life". It appears in various circumstances : as a volonteer "political" decision (the decision of an authority, king, khan...), as a creation of a master weaver (from the traditional found or as a personnal "fantaisie"). It can be also inspired by foreign designs and locally adaptated. The design is then adopted by the group folowing the "fashion" process. There is a period in which the design is at its optimum. After comes the progressive shifting of the design, and it can desapear because beeing not yet fashionable or because the chief who decided it has been beaten.
With an expert exerced eye I think it is possible to find among a lot of similar design pieces which are the "good" ones (possibly the "prototypes" or the first models, or the good copies and interpretations of the optimum period), and to find the bad copies or the shifted design corresponding to the decline period of it. It is also possible to detect copies that have been made out of the period the design was "living" : too perfects but without any "feeling". I think this is the case of the Bohmer's example : too exact, too stiff. In this case the perfect aspect can come either of an entier copy of an old fragmented piece or of a very complete restauration of a fragment. from the result we can say that it is better to have a fragmented piece than to try to restaure it with the danger to kill any of its qualities.

Steve Price · December 13th, 2009, 07:38 AM

Bonjour, Louis

I'm pretty sure you and I and most ruggies would agree on aesthetics of groups of rugs, but that doesn't prove that our aesthetics are those of the weavers or of the population at large. The overwhelming majority of people in the developed world neither collect nor even see the beauty that we see in antique rugs. We learn the aesthetic from other ruggies, and the belief that its roots lie in some exotic and now-dead cultures is likely to be fantasy.

There was a most interesting example of this on a website run by Rugdom's Village Idiot. He posted two pairs of soumak khorjin faces that he asserted were an older and a younger in each pair. I suspect that most ruggies would agree with RVI's relative age dating, by the way. He believes that the age differences are self-evident. A novice wrote in, listing what he saw as the differences, in his own words. It would be hard to defend the position that his observations were inaccurate - that the two "older" pieces were more worn, had less saturated palettes, had more primitive drawing (I think his word was "charming"), and had fewer filler motifs. RVI took offense, and refused the novice's request for the criteria by which he determined the ages of the four khorjins.

The novice probably represents the general public's aesthetic.

Regards

Steve Price

December 15th, 2009, 08:15 AM

Dear Readers,

Joel Greifinger has poised several legitimate questions in this thread that have been largely ignored save a few general comments about subjectivity and the dating method used in Rageth's book Anatolian Kilims and Radiocarbon Dating that apparently only Marla Mallett and I have access to.

The testing that was conducted for Rageth's study was the dating of milligram samples with the Accelerator Mass Spectrometry or AMS method or technique. AMS directly counts the number of C isotopes in a sample. The basic idea, according to Georges Bonani of ETH Zurich, is to first accelerate the C ions produced in a negative sputter ion source to high energies and then to eliminate the isobaric and molecular interferences with filters. You have to read the bood if you want a full discussion. The sample of pieces that were tested by Rageth were all chosen as possible examples from before 1800. Dye testing was also done. It was true that results prior to AMS were reliable only for samples from the 16th century or before. AMS arguable allows this window to expand to approximately 1800. The issue for objects from the 16th - 18th centuries appears to be less the reliability of the age of the object or the amount of C present but more the calibration, i.e., the the conversion of the radiocarbon age BP into a calendar age BC or AD. An interval of 100 - 200 years for an object that is between 250 and 300 years old is sub-optimal. That said, Rageth study seems to suggest that there are three age groups for old kilims: 1450 - 1650, 1650 - 1800 and 1800 - 1950. These periods relate to the shape of the calibration curve. See Rageth pp. 23 - 28.

So you see, not everything must be as subjective as some writers would suggest.

As for Joel's specific questions - I think that of the few pieces dated reliably to the 1450 - 1650 group you can begin to see the colors and color theories (refer to Goethe's color theory, for example) that begin to guide your eye. These pieces serve as a kind of anchor. You can also see the designs of these pieces, look for commonalities. Marla Mallett and I have already identified a few concepts, conventional wisdoms that might help you on this journey. Perhaps you will find some continuum. There could be other evidence based on weave and materials that would inform only by hands on examination.

Regarding the piece published by Bohmer - if you compare it to the Lantz piece and and Horst's piece as well as the other examples from Rageth's study posted by Marla I think you will begin to see the answer. Bohmer's piece would seem to be a relatively later piece based on both loss of design and the intensity/saturation of color and color palette. Of course its relative age does not necessarily define its beauty as a work of art either objectively or subjectively. However, I am not a kilim expert so you should form your own opinion.

Wishing you a pleasant journey, Michael

Steve Price · December 15th, 2009, 09:26 AM

Hi Michael

Thanks for the update. I wasn't aware of the accelerator mass spectrometry advance in C-14 dating, and the fact that it makes reasonably reliable dating possible for pieces as recent as the 19th century is a great leap forward.

I'll have to get a copy of Rageth's book. But even without having one,I can see that it holds the seeds of a database from which evidence-based dating criteria can be extracted.

Regards

Steve Price

Marla Mallett · December 15th, 2009, 02:29 PM

In almost any art field, curators with substantial experience can, almost at a glance, with gut reactions, place pieces within particular time frames. Curators of Western European paintings, for example, can almost automatically place pieces within the Baroque, Rococo, or early 19th century periods. It's a matter of style... difficult to summarize aesthetics that changed over time. To me, it seems no different with Anatolian kilims: 17th/early 18th century pieces from Central/West Central Anatolia are different from late 18th century/early 19th century pieces; these have a different character from mid-19th/late 19th century pieces. Most pieces from after about 1910 seem to have a character of their own. Dr. Bohmer has always been vague about dating pieces, but this piece published by him seems to me to almost surely to belong to the early 20th century.

It's of course only an opinion, but upon seeing Horst's wonderful fragment for the first time, I assumed that it dated to the end of the 18th century or very early 19th century. As I've looked at it more, that conclusion has strengthened. I agree with Louis that Horst's fragment represents the "best design" of the group...This is actually characteristic of many of the kilims from that period, when a synthesis seemed to occur in central Anatolia and some of the most powerful pieces were created. After that, an over-refinement seemed to occur and as the weave balances gradually became more and more refined, the designs weakened. There were of course talented kilim artisans in every time period, as well as people who weren't very adept at designing, so that pieces of varying success were produced in every period. But the overall aesthetic is a different matter, and THAT did evolve gradually over time, while tending to be pervasive and consistent within a community or group.

Best wishes,
Marla

Steve Price · December 15th, 2009, 02:44 PM

Hi Marla

Paintings, sculpture, music - almost any western art form - can be given pretty accurate age and geographic attributions based on style if the person making the attribution is reasonably knowledgeable. The reason is that there is an extensive database of documented examples from different times and places, and they (we) extract criteria from those. Even a Phillistine like me can usually place pieces of music that we've never heard before in time and place, often even name the composer.

That happy situation doesn't exist in the world of rugs. It sounds like Rageth's book will lay a foundation for creating such a database for kilims, and it's probably reasonable to extrapolate certain things from kilim to soumak and pile. Until those things get a little further than they are now, age attributions will continue to be reliable for late 19th century and later, will continue to be guesswork for older pieces.

Regards

Steve Price

Marla Mallett · December 15th, 2009, 04:13 PM

Steve,

There actually is information on early kilims stashed away in the Turkish Vakiflar collections and archives--sometimes info gleaned from mosque inventories. We are fortunate that because of the long tradition of Vakiflar donations to mosques, many more early pieces have survived in Anatolia than in other places. When many of those were on public display in the past (before the kilim museum closed because the pieces were rotting), kilim enthusiasts (including me), studied them thoroughly. Pieces done for the early Ottoman courts are easily dated by style, as their parallels are seen in other art forms associated with court workshops. Then too, we have the records of exported Turkish pile rugs in European paintings. Since there were very few soumak pieces produced in Anatolia, it would be difficult to extrapolate information relating to those (Persian and Caucasian?) from Anatolian kilim databases.

Best,
Marla

Steve Price · December 15th, 2009, 05:24 PM

Hi Marla

If there is a database of kilims with reasonably well documented dates and a significant number of specimens of various ages, somebody ought to work up the criteria extracted from them (and their foundations) and publish them. The criteria can then be refined through critical analysis of the supporting data. I'm not doubting your word, but, "I've looked at lots of documented pieces, so I know how to make reliable attributions" isn't compelling evidence all by itself.

Regards

Steve Price

December 16th, 2009, 09:13 AM

Dear Readers,

Of course, Marla Mallett (as Steve Price acknowledges) is correct that in most any art field, curators can place a particular piece within an equally particular time frame based principally on style. And Steve Price is also correct that much of this has to do with a fairly well documented database of anchor pieces - pieces with known provenance.

The field of Anatolian kilims is, however, not most any art field. For one thing, Anatolian kilims is an anonymous art form and anchor pieces are still very few. For another, interest in Anatolian kilims is relatively recent and, with few notable exceptions, confined mostly to enthusiasts rather than trained curators. Similarly, many of the conventional dating methods that help curators, enthusiasts and even Phillistines like Steve and me place pieces such as piled carpets are not available to assist in the dating of Anatolian kilims.

For example, Anatolian piled carpets have been collected or been part of the stream of commerce for centuries. Some have a relatively long collection or inventory provenance and some are depicted in paintings of Italian and Dutch manufacture over these centuries, Anatolian kilims have not been collected outside of mosque donation until recently and first make their appearance in Orientalist paintings of the late 19th century such Les Rideaux by Cezanne. Inscribed dates is equally frustrating. While there are generally frustratingly few for pile carpets, there are none that I am aware of among Anatolian kilims prior to about 1815. While much work has been done on dyestuffs, the most significant dyes that can be used to pinpoint palette changes are largely the appearance of synthetic dyes such as Fucsine and Mauvine in the 1860s, as well as uses of cochineal and indigo sulphonic. Too late to really help us firmly date early Anatolian kilims. The pieces in the Vakliflar colletions are better known to Marla than most and certainly I do not know them. However, Rageth makes little mention of them nor do the papers of most presenters at the symposium. This may be because the best documented "Ottoman" pieces are different stylistically from the Anatolian kilims we are discussing.

To demonstrate my point, consider that as recently as 1997 when Rageth organized his group of Anatolian kilims for testing (with the help of kilim enthusiasts he sought only kilims that were felt to be pre-1800 for testing) there were still some significant surprises in dating even though the vetting group for testing had generally concurred on probable age. This occurred both in pieces that ended up testing significantly younger and older.

That said, Marla's opinion about Anatolian kilims from the period of very roughly 1650 - 1800 and (if I read it accurately) and particularly Anatolian kilims from toward the end of that period when a possible synthesis occurred in Central Anatolia that gave rise to many beautiful and powerful kilims seems to have some consensus among curators/enthusiasts. The restrained palettes and spare drawing of the earliest Anatolian kilims start to be replaced with bolder design and color while remaining harmonious. As Marla and others have stated, it is difficult to summarize these changes. Of course, there are others, such as David Lantz who would argue passionately that the something iconic and cultural is lost in this shift or synthesis and therein lies part of the dilemna and controversial nature of Anatolian kilims.

I hope the discussion in this thread has helped Horst appreciate his wonderful kilim even more and that others have become interested in appreciating this artifacts.

Best wishes, Michael

Steve Price · December 16th, 2009, 10:18 AM

Hi Michael

Thanks for your reasoned and thoughtful post. Just a few points.

You wrote, ... roughly 1650 - 1800 ... and particularly Anatolian kilims from toward the end of that period ... seems to have some consensus among curators/enthusiasts. The restrained palettes and spare drawing of the earliest Anatolian kilims start to be replaced with bolder design and color while remaining harmonious.

Vis-a-vis the consensus among curators/enthusiasts, I quote another piece of your post, ... when Rageth organized his group of Anatolian kilims for testing (... only kilims that were felt to be pre-1800 ...) there were ... surprises in dating even though the vetting group for testing had generally concurred on probable age. This occurred both in pieces that ended up testing significantly younger and older.

This looks to me like one of those dramatic moments when a beautiful hypothesis comes face to face with some ugly facts. It demonstrates the unreliability of consensus, even when the consensees have credentials as experts.

You and Marla both cite the presence of certain carpet designs in paintings of known age. This, of course, proves that those designs were being woven and that carpets with those designs were being exported to Europe at least as far back as when those paintings were created. But it's important to note the limitations of those facts as attribution criteria. The paintings do not (and cannot) include information about structure or accurate depictions of colors, both of which are usually important elements of age attributions. So it's certain that Holbein pattern carpets date back to Holbein's time (as opposed, for example, to Pak-Bokhara, for which such a date is implausible), and that's interesting. But it doesn't provide a basis for attributing a carpet to Holbein's time. You and Marla both know this, but I thought it worth mentioning here.

Regards

Steve Price

Marla Mallett · December 17th, 2009, 12:32 PM

Steve,

This an old topic, discussed several times before, in which your views are now well known: you find little credibility in the judgments of people who have spent lifetimes immersed in the world of aesthetics, and who have studied, and handled thousands of kilims or carpets over periods of 30 or 40 years. Of course everyone makes mistakes. But in the kilim field, I personally have found it invaluable to have the counsel of three or four very wise people in Turkey who, in our discussions, have made excellent cases for their views. I have been able, then, to add to these discussions what I've learned from my studies of design/style evolution and disintegration, as well as technical evolution. While it is normally impossible to put exact numerical dates on pieces, to me it is quite reasonable to see the evolution of styles and weaves in terms of generations of weavers. The evidence is right there before us. Sorry to bring this up again, but it is a sore point that I can't help protesting.

Unfortunately, marketing within our field has complicated dating issues. Especially at the top end of the market. It is natural that any collector who has bought the seller's story along with the kilim is predisposed to believe and defend it, possibly even going to the expense of having his piece carbon 14 tested. Unfortunately, we need to be cautious about placing complete faith in dealers' catalog dating. As a kilim dealer myself, I learned many years ago to make my own kilim buying decisions in Turkey...then only after finalizing a sale, to ask the seller's opinion as to the age of the piece. A much more candid opinion was likely to be forthcoming. We could then discuss other pieces more freely as well. Separating knowledge and substative information from marketing is difficult, but most of the expertise in the kilim field does indeed reside in the experience of dealers--whether or not it is candidly conveyed. Museum people handle many fewer examples, and usually have to rely on people in the market for their information, except in the case of those scattered pieces with documentation. We can even cite an example of a 1990 museum catalog with descriptions written by the seller of the kilims--an indisputable conflict of interest that intensified many people's doubts about what seemed to be very questionable dating.

Another particularly flagrant market abuse occurred a few years ago when H. Kircheim was sold a group of very coarse, simple "17th century" yellow-ground Konya pile carpets. He admitted to me, when we talked at his ORIENT STARS exhibition opening in Hamburg, that carbon-14 tests had just come back with probable age estimates of 40 years old! He couldn't wait to dump them. This dating may not have been accurate, but at least the carbon 14 tests could not come even close to the supposed early dates. Well... the Turkish experts that I know were not at all surprised...except for the sellers.

Best,
Marla

Steve Price · December 17th, 2009, 01:25 PM

Hi Marla

You note, correctly, that I find little credibility in the judgments of people who have spent lifetimes immersed in the world of aesthetics, and who have studied, and handled thousands of kilims or carpets over periods of 30 or 40 years. It's clear that you disagree with this point of view, and that you're somewhat offended and perhaps angry about it.

Let me take this to a realm not directly rug related. There are at least hundreds of thousands of people, probably many millions, who have spent their lives immersed in their religion's teachings. To most of them, the evidence is right there before us. It seems self-evident to me that they can't all be right; their positions are mutually contradictory. This raises the fundamental (I think) problem with finding credibility in the judgements of people because they devoted themselves to diligent study and belong to a large coterie of adherents to their beliefs.

The fact is that age attribution for any genre of weavings involves, consciously or not, a graph in which some characteristic (or combination of characteristics) appears on one axis and time is on the other. Having large error bars doesn't bother me at all; I know how to deal with that. But such a graph cannot be constructed without some known (documented) standard specimens of known age and characteristics. Any other approach is based on faith, not on evidence. Belief based on faith is perfectly OK for anyone (even me!). But my fervent belief isn't compelling evidence to someone else.

Regards

Steve Price

Marla Mallett · December 17th, 2009, 03:17 PM

Steve,

No I'm not angry. In my advanced years, having spent my entire life involved with aesthetic issues, I'm quite used to people who have little respect for aesthetic insights gained from long study, little regard for knowledge based on elusive visual factors that can't always be quantified or neatly pigeonholed....who can't accept that there are logical progressions or logical evolution in human artistic development just as in other fields. For anyone who has spent time studying the evolution of woven design, one can make very good guesses about the probable appearance of "missing links." And certainly put the pieces in the proper order. Of course there are some folks who doubt the veracity of biological Evolution too.

Horst,

In all this time, I think that nobody has commented on your second fragmented kilim--the Karakecili/Aydin piece. I find it extremely interesting...though now I'm reluctant to advance any theories about its possible history.

Best wishes,
Marla

Steve Price · December 17th, 2009, 04:10 PM

Hi Marla

I have great respect for the aesthetic insights that serious students of the subject have. But, transforming aesthetic insights into criteria for reliable date attribution requires having examples of known dates with which to do it. A corpus of pieces with documented dates like the kilims in Rageth's book, for example, could allow that transformation to be made in the case of Turkish kilims.

Just as an aside, there is a lot of evidence that Darwin's two major hypotheses (evolution happened and natural selection was the driving force) were fundamentally correct. Additional data might change the conclusions some day, of course. Most important, though, is that the evidence (the data) is public information and that it has been and continues to be subject to critical analysis.

Regards

Steve Price

Louis Dubreuil · December 17th, 2009, 05:56 PM

Could the central medalion motifs be symbolic drawings of turtles ?
I have not found in the classical typology of kilim design (Erbek's book) any design of turtles. In the other examples of related kilims shown there is a complicated medalion with many hooks all over the outer line of the motif. Erbek says it is "bereket" = fertility motif (fig 195, page 59). Bereket seems to be a kind of "hold-all" concept that is applied to many different designs.

I have found in a paper from Filiz Nurham OLMEZ about symbols in anatolian weavings a paragraph about turtle design :

"The Motif of Turtle
Turtle was used at the lower parts of the columns on the Tonyukuk monuments, and it is a motif that comes from Shaman traditions and used against the evil eye in our ethnography (Diyarbekirli 1972, 55). The turtle indicates long life and death (Coruhlu 1993, 15).
The motif of turtle which is also known as “tortoise” is presented in a geometrical form which looks like the motif of elibelinde. The motif of turtle is used in the kilims of Kula, Manisa-Selendi, and Kutahya region kilims and Karakecili kilims (Onuk&Akpinarli 1995, 20; Deniz 2000, 196)." from “DEATH” SYMBOLISM IN TURKISH WEAVINGS
Filiz Nurhan OLMEZ
Dr. Suleyman Demirel University,
Faculty of Fine Arts Department of Traditional Turkish Handicrafts,
http://filiznolmez.sitemynet.com/, fnozan@hotmail.com

Unfortunatly there are no illustrations for this turtle motif in this paper. If anybody can find something among the cited issues...

Cordialement

louis

Steve Price · December 18th, 2009, 10:48 AM

Hi Louis

The fact that turtles were significant cultural icons in the region makes it possible that this could be what the central motifs represent. It would be great to be able to narrow the list of candidates to a small number, though. I read a relevant comment somewhere else yesterday: every rug is a Rorschach test. I don't remember where I saw it, and I apologize to the author for not being able to give him proper credit. If I have a lucid moment later, I'll add it.

Regards

Steve Price

James Blanchard · December 18th, 2009, 01:44 PM

Hi all,

It is interesting to me that various folks who have a fair bit of experience have suggested date estimates ranging over 4 centuries for Horst's kilim. Perhaps a group of Turkish experts would tighten those estimates. I would be interested in a study wherein a group of experts anonymously estimated the age of a relatively large population of weavings. It might at least give some confidence that there was agreement within an expert group, though we might still have some doubts as to whether they agreed because they were all accurate, or because their social and professional interactions over the years had created a "common wisdom" about how to date pieces.

I am generally sympathetic to Marla's viewpoints on the generational changes in design, but less certain about how much we know about how quickly this might have occurred, or how evenly across weaving groups.

For the large group of non-experts, like me, the main thing to avoid is paying way too much for a piece based on an erroneous age estimate. I try to follow the following rules:

1) Buy only what you really like, and resist the urge to buy something purely because you are persuaded that it is "collectible".

2) Try to "start small" with purchases, because your taste will change over time as you learn more and become socially conditioned by the advice and opinions of other collectors.

3) Beware of the allure of the "great deal", unless you really think you want to become a dealer. Just like you, everyone else is looking for a great deal and usually can afford to wait and look around.

4) Books are usually cheaper than rugs, so invest in some. They not only make collecting more interesting, but also accelerate your ability to avoid major mistakes.

5) Wait a day or two before larger purchases and do some interim research. There are examples and references for just about every sort of weaving that one might buy, so one can usually become at least minimally educated about a potential purchase before opening one's wallet.

James

Steve Price · December 18th, 2009, 02:48 PM

Hi James

Michael Wendorf's post #29 describes something similar to your suggestion about experts making independent age attributions of what they agreed were pre-1800 Turkish kilims.

Regards

Steve Price

Louis Dubreuil · December 18th, 2009, 06:08 PM

Hi Steve

I think that the main parallel between rug or kilim motifs and Rorsach test is the use of the symetry. Symetry is the main property shared by 99,99 % motifs. All kinds of symetry : axial, central, diagonal...And symetry cannot be just a technic consequence.
I think it could make a good salon starter. And there is a lot of litterature about this subject.
In the "rug test" we can play at "what kind of animal you see here". It is a very simple game. But we can play also at "what do you see in this very bizare multi toothed or multi hooked symetric medalion" and we are already very near of the sofa of the good doctor Sigmund on which there was also a rug with odd serrated and symetric medalions....mind the abyme !

Cordialement

louis

Marla Mallett · December 19th, 2009, 03:37 PM

Steve,

Just a couple more notes: Before you accept carbon-14 dating and Ragath's book of 1999 as a panacea--the answer to all questions of dating kilims and a reliable method of establishing "benchmarks"--there are a couple of factors to consider. First, you will find that age estimates for the vast majority of pieces shown encompass as much as three centuries. Sometimes an even broader range. The percentages given just suggest the most likely part of that range. Of what practical use is a suggestion that a weaving originated sometime between the early years of the 16th Century and the early part of the 19th Century?

There are, of course, still strong doubts about the veracity of even this vague dating for recent objects. There are strong concerns about the effects of contamination, even when test materials are repeated washed. In some cases the results are laughable: Consider the example of a Central Anatolian piece shown in Ragath's book as Plate 39. Just about anyone familiar with old kilims would assume this piece to be from the early or middle years of the 19th century. Indeed, it was published in the Vok catalog as "beginning of the 19th century" (Udo Hirsch's date). The radiocarbon dating came back, however: AD 1899-1901 (2.7%), and AD 1954-1956 (97.3%)....with 95% confidence limit. Ragath's text states that the piece was tested three times. If you would like, I can scan this photo for you. So how does one select useful "benchmarks"? Does one choose which test results to accept and which to reject based on other, more subjective criteria?

Best,
Marla

Horst Heinzlreiter · December 19th, 2009, 04:42 PM

Dear members,
first thank you again for all your informative comments, which we followed
with great interest.

concerníng age:
I contacted Mr BONANI (ETH Zuerich) and he told me that C-14 dating
for textiles younger than 300 years is unsatisfactory - it has to be combined with design studies.
So for me the exact age of the kilim will remain in the dark.
But anyway through it´s powerful drawing age speaks to us.

It seems to be a difficult job to date a kilim and needs a lot of experience.
Let me show you an other one, which illustrates that.
It comes from Bertram Frauenknecht and he dated it as about 1700 refering to the design of the mural crown motif. As Marla pointed out: It needs people,
who studied and handled thousands of kilims.

However your attempts to date the textile gave us a scale to come closer to the "real" age of the piece.

Thank you Horst

Steve Price · December 19th, 2009, 05:02 PM

Hi Marla

The reliability of the results can be no better than the method. Until not so long ago, any C-14 date more recent than, perhaps, 1700 was routinely ignored because of technical problems that were well understood but not fixable. That is, all anyone could conclude from a C-14 result giving a more recent date was that the date was more recent than 1700 or so, but otherwise indefinite. Even under the best of circumstances, C-14 dating of rugs doesn't do an awful lot more than assign them to rather broad date ranges. One fine day, a much better method will emerge. The late Leslie Orgel and I chatted about amino acid racemization, for instance, but never got down to the nitty gritty with it.

I gather from Michael Wendorf's account in this thread that C-14 is now useful up to the early 19th century. I don't know how the new methodology works, but my guess (and that's all it is) is that a C-14 date more recent than 1800 can't be trusted to mean anything more specific than that the piece was woven after 1800. That is, even though the 1955 date was reported to be highly reproducible, it probably only means "post 1800". As a side issue, I'd be interested in knowing whether the three assays were truly independent. For example, I can weigh myself repeatedly on my bathroom scale and get a very narrow range of variation; standard deviation probably less than 1 pound if I do it five times. But I wouldn't be surprised if my true weight is 5 pounds higher or (sadly, less likely) 5 pounds less.

The rest of your post asks, basically, what do we do if there's no objective way to make a date attribution? My answer is, we just have to live with that. My own habit for years has been to say, for example, that such-and-such a piece has the characteristics that would lead most ruggies to attribute it to the mid-19th century (or whatever was appropriate). This acknowledges the conventional wisdom without actually endorsing it or any specific alternative.

Regards

Steve Price

Steve Price · December 19th, 2009, 05:15 PM

Hi Horst

Let me put you into a hypothetical situation. You have been locked in a library with 10,000 books, all written entirely in Mandarin Chinese, for 25 years. You've had no access to even a single translation of any character, no mentor, and none of the books includes any illustrations. During that 10 years, you've pored over every page many times. So, it would be fair to say that you've got extensive experience with books written in Mandarin. How well do you think you would be able to read the language?

Regards

Steve Price

Marla Mallett · December 19th, 2009, 05:57 PM

Steve,

My point is: What good does a 300-year date range on any specific kilim do us, as far as establishing the "benchmarks" you are asking for, for pieces prior to 1800....A range from early 16th century to 1800? EVEN if the tests are absolutely accurate?

Your "parallel" with Chinese text hardly makes sense. We ARE very familiar with 19th and 20th century kilim production, so those provide a starting point.

Marla

Marla Mallett · December 19th, 2009, 06:40 PM

One more thing: Georges Bonani, who has written the technical section of the Ragath book, emphasizes that "Whenever calibrated data is reported, ALL true time intervals have to be quoted. It is NOT permissible to only report the interval with the highest probability and to omit the intervals with the smaller probabilities."

Marla

Marla Mallett · December 19th, 2009, 07:00 PM

As for tests from labs that match or don't match, Rageth cites one example:

"The kilim on Plate 25 exemplified this problem. A dating performed in Arizona in 1993 placed this piece into the 14/15th century. A second test undertaken at the ETH Zurich, in spite of several replicate measurements could not confirm this result but dated the kilim to the period 1650-1800." Ragath then goes on to say that in such a case a weighted mean value should be calculated, but then says that the client rarely does this....(possibly just preferring to accept the earlier date???).

I wonder...If one of the Labs' set of tests was obviously in error, or the materials were contaminated, why should those results be averaged with the other? Don't we just need to admit that the methology has serious problems?

Marla

Steve Price · December 19th, 2009, 08:32 PM

Hi Marla

Dating rugs, bags, kilims, etc. made during the last 100 to 150 years is not a problem. Dealers knew what they were buying from the sources during this period, and much of that has been passed down. That's another way of saying that we have a database of textiles from western and central Asia that makes it possible to extract criteria for making attributions within it. We can even be pretty confident that what we think predates 1850 really does predate 1850. Trying to assign more specific dates within this group is where the troubles arise. That is, attributing this piece to 1750-1800 and that piece to 1550-1600 has much more uncertainty than most people are willing to admit. Carbon-14 can be helpful in this range, but as you note, it's pretty imprecise. I haven't seen Rageth's book, but had the impression that the date ranges reported for individual pieces were much narrower. If that's not the case, the much needed database of pieces with documented ages isn't in there.

What to do when a piece tests as 15th century in one lab and 16th century in another? The folks running one or both labs need to look carefully at the methods, look for subtle differences that might cause errors, eliminate those differences, and test again. Lab to lab discrepencies are common in lab based science, and finding their basis is much of what keeps the process moving forward. Simply averaging the disparate results is not general practice unless it is certain that the methods were identical in every significant way. On the other hand, one of the facts of life in labs is that if you do the same procedure five times you will get a range of results, and reporting an average and standard deviation is considered the most constructive (and honest) way to do it.

Carbon dating differs from most garden variety lab procedures in that the method gives multiple date ranges with probabilities for each rather than a single mean plus error estimates. In most scientific work, a 95% probability is taken as sufficient - that is, a 5% error rate is considered acceptable. For life and death issues, greater probabilities are needed. In the example you quoted, where the result came in as AD 1899-1901 (2.7%), and AD 1954-1956 (97.3%)....with 95% confidence limit, I think most scientifically trained folks would accept the whole thing as meaning "post 1800" and recognize that nothing more specific can be concluded.

The Mandarin text thing was included only to emphasize the fact (at least, I think it's a fact) that having handled and studied thousands of Turkish kilims doesn't give a person the ability to accurately make date attributions any more than handling thousands of books in Mandarin will give him the ability to read it. Of course, this doesn't apply to the period during which reliably documented specimens exist. I'm sorry if I've been unclear on that point.

Regards

Steve Price

Joel Greifinger · December 19th, 2009, 08:51 PM

Þa cydde man me, þæt us mara hearm to fundode, þonne us wel licode: and þa for ic me sylf mid þam mannum þe me mid foron into Denmearcon, þe eow mæst hearm of com: and þæt hæbbe mid godes fultume forene forfangen, þæt eow næfre heonon forð þanon nan unfrið to ne cymð, þa hwile þe ge me rihtlice healdað and min lif byð.

Hi Steve,

I think your thought experiment with the Mandarin library gets to an important point but we can best see the larger analogy if we instead have a group of modern English speakers/readers trapped for the duration in a library stocked with books and documents in Old English (such as the excerpt above). This provides them with knowledge of recent practice in materials that descend from this tradition. Over time the group may begin to form theories about the content of the volumes and begin to generate standards for judging the beauty of certain arrangements of characters on the page. Norms may develop for estimating the age of individual books not only from their state of wear but other perceived similarities and contrasts with contemporary English phrases and the contents of the books shelved nearby. By the end of this confinement, the most experienced members of the group would be extremely knowledgeable about many aspects of the library and its contents. They might even have generated a set of more or less coherent narratives about the history of the library’s holdings, citing consistencies within its pages as evidence. But, barring some independent means of verifying their translations, we can’t say with any certainty that they really read Old English.

Horst wrote:

Quote:

However your attempts to date the textile gave us a scale to come closer to the "real" age of the piece

There is no question that Horst’s (and every) piece has a real age. What seems questionable, given the paucity of evidence independent of the circularity you’ve pointed out, is whether we can justifiably believe that our estimates of the age of any particular piece is more than very broadly accurate. Without such evidence, a consensus of experienced opinion primarily tells us a great deal about how knowledgeable commentators view these objects. Pointing out the epistemological limits of that expert consensus doesn’t in any way denigrate its value in helping us appreciate the objects of its study.

Joel Greifinger

James Blanchard · December 19th, 2009, 11:45 PM

Hi all,

Perhaps this is a naive question, but I if there are enough experts who could agree with each other on the dating of kilims, why would Rageth or anyone else embark on a project to use C-14 dating? Doesn't the mere fact that an imperfect estimation method such as C-14 dating is being used suggest at least a general unease among experts about dating based on gestalt? (I don't use the term "gestalt" in a pejorative way, since I generally do give substantial weight to the wisdom of experts).

James

Steve Price · December 20th, 2009, 06:01 AM

Hi James

I'd expect the rationale to be in the book's preface; I haven't seen the book, so I don't know. But even with minimal discomfort about expert opinion, getting independent evidence about the dates would be useful. If expert opinions were mostly confirmed, it would give future expert opinion more credibility by establishing an objective track record for the success of the methods experts use.

Also, C-14 has a history of generating surprises in Rugdom and you can't catch fish unless you go fishing. Jim Allen's Turkmen collection, for example, includes pieces for which there is now pretty good evidence that there are extant Turkmen weavings from the 16th century. Many (but not all) Turkmen collectors/scholars/obsessive-compulsives believed that the oldest Turkmen things still around were from the 18th century until it was proven otherwise.

Finally, like you, I give great weight to expert opinion. But I don't think experts on kilimology have expertise on the date attribution of kilims woven before about 1850. An expert's expertise has limits. In most fields, experts know the evidence that forms the foundations of their expertise and, for that reason, they also know its limits.

Regards

Steve Price

Marla Mallett · December 20th, 2009, 09:10 AM

To get back to the original subject: How about putting Horst's kilim alongside the Plate 33 kilim from Rageth's book that I posted earlier.

What does the scientific data tell us? That the Plate 33 kilim was made (with 95% confidence) between the years 1427 and 1815. How helpful is that? A spread of 388 years! It just suggests that it MIGHT be very early production. Yet if we study these pieces from an aesthetic perspective, a design evolution perspective, we should be able to reach much more satisfying conclusions than that. Especially if we then add examples that were made, almost certainly in the 19th century. The conclusions become still more convincing if we study massive numbers of other kilims from the same geographic area.

We do indeed need to heed George Bonani's words: that it is improper to cite only the date intervals with highest probability. It is certainly dishonest (or careless) to cite just the highest probability date interval and then claim "95% confidence." Bonani is the guy who did most of the tests in Ragath's book. There are very few carbon-14 tests which can pin down the origins of a piece to a single century, and any such claims should be regarded with suspicion.

Marla

Marla Mallett · December 20th, 2009, 09:22 AM

James,

Why would anyone embark on a project to carbon-date pieces? To SELL them.

Marla

Joel Greifinger · December 20th, 2009, 09:45 AM

Hi Marla and Steve,

I'm trying to get a handle on the current state of the discussion.

Marla, you wrote:

Quote:

if we study these pieces from an aesthetic perspective, a design evolution perspective, we should be able to reach much more satisfying conclusions than that. Especially if we then add examples that were made, almost certainly in the 19th century.

Earlier, Steve's recommendation for stating age attributions was to say:

Quote:

for example, that such-and-such a piece has the characteristics that would lead most ruggies to attribute it to the mid-19th century (or whatever was appropriate).

Since (I think) you both agree that, at least at present, C-14 dating doesn't provide the specificity that would be of most use in dating these kilims, and that the study of design variation in large samples of pieces about which we have more certain information provides the most satisfying conclusions we are able to draw, given the nature of the evidence, what is the substance of the disagreement? Is it about the degree of tentativeness or certainty with which age attributions are expressed?

Joel Greifinger

Steve Price · December 20th, 2009, 10:10 AM

Hi Joel

I think the fundamental difference of opinion is whether we can attach a time line that has reasonably well defined points of known states in design evolution, assuming that design evolution is unidirectional. Marla (and many, many others) believe that we can. I don't, nor do some others.

Regards

Steve Price

Marla Mallett · December 20th, 2009, 12:32 PM

Of course design evolution is NOT unidirectional-- not in kilim weaving or any other of the visual arts.

Marla

Steve Price · December 20th, 2009, 01:22 PM

Hi Marla

Obviously, not unidirectional over really long periods and different for different genres. But unless it's unidirectional for significant periods, I don't see how anyone can even assign relative ages from considerations of design evolution. It seems to me that this approach to dating is paralyzed unless it assumes unidirectional evolution within the particular genre and time period. Not so?

There's also the problem of not knowing the rates of design evolutions and (usually) not knowing the forces driving it until we get to the commercial period - roughly, 1850 to now. Ironically, the commercial period is the least problematic in terms of date attribution anyway.

Regards

Steve Price

Marla Mallett · December 20th, 2009, 02:22 PM

The difference between Steve's view and mine seems to be this: While he would like the convenience of a design "time line," I find that concept absurd when dealing with artistic expressions. From my perspective, we need to study until we recognize different aesthetics--artistic attitudes that changed significantly over time in each geographic area. It's not a matter of taking one motif and tweaking it, then tweaking some more, and then a little more, then plugging each version into a graph. There are far too many interconnected and subtle variables that make up that elusive thing we call "style."

Ruggies normally seem far too focused on motifs rather than style in both dating woven objects and in assigning provenance. Caragaggio and Manet may sometimes have focused on similar subjects (motifs) but it's style that allows knowledgeable individuals to distinguish their works. The same is true of kilims separated by 250 years.

Marla

Steve Price · December 20th, 2009, 02:47 PM

Hi Marla

Is it correct that the term "design evolution", used throughout this thread and elsewhere, actually means "style evolution"?

Regards

Steve Price

PS: I think the focus collectors often place on motifs is an outgrowth of the common notion that the "magic" in a tribal textile has to do with the meanings of motifs.

Marla Mallett · December 20th, 2009, 05:08 PM

Steve,

I suppose that I've used the terms "design evolution" and "stylistic evolution" somewhat interchangeably. They are difficult to separate. It is apparent that some people tend to ignore the elusive but critical elements of style which are part of "design evolution." I have NOT meant the term "design evolution" in this discussion to mean only the evolution of individual motifs. This is a much more narrow concern; while it is logically part of the larger picture, it's only one part.

Marla

Steve Price · December 20th, 2009, 05:51 PM

Hi Marla

Thanks for the clarification.

Regards

Steve Price

James Blanchard · December 20th, 2009, 05:57 PM

Hi all,

I would be very interested to see a cogent explanatory exposition of the design evolution for various types of Turkish kilims, and several other types of weavings. In fact, I think that there is too little attribution without explanation. So, just as Rageth is reportedly working on assembling a collection of kilims dated by C-14, wouldn't it be of some interest for someone to put together chronologically ordered sets of kilims based on an appreciation of design evolution? One might still discuss the actual timeline, but at least it would give a better appreciation for how experts date kilims relative to each other. For example, there is a lovely set of Yuncu kilims in Paul Davies book on Anatolian kilims, but it is not clear what criteria he has used to assign dates (i.e. 18th, early 19th, mid-19th, etc.).

Does anyone know whether such an analysis has been done? If not, wouldn't it be a useful way of communicating the experience of experts?

James

P.S. My daughters just called me to see a Tekke engsi displayed in "Ben Hur" (with Charlton Heston). I now have to revise my assessment of the possible timeline of this sort of weaving, ranging from Biblical times to Charlton Heston's early years.

Rob van Wieringen · December 21st, 2009, 06:36 AM

Hi All,

Quote of Michael Wendorf in #29:
"....when Rageth organized his group of Anatolian kilims for testing (with the help of kilim enthusiasts he sought only kilims that were felt to be pre-1800 for testing) there were still some significant surprises in dating even though the vetting group for testing had generally concurred on probable age. This occurred both in pieces that ended up testing significantly younger and older".

It seems to me very logical in this discussion, between learning mandarin versus learning style, to focus also on the score of the vetting committee.
How many of the total vetted "felt pre-1800 kilims" were in fact correctly predicted? Is there any review of these sub-results in Rageth's book?
And to extend this principle a bit further, it would be interesting to know how the scores generally are in all C-14 testing labs worldwide for rugs/textiles over time. When these data would be available it would show nicely two curves: one for the expectation dates of the consigners and one for the actual C-14 dates.
The availability of those figures could help in forming an opinion about the expertise of experts/consigners in dating items.
If they were doing significantly better as a random pick, which is actually what I expect, that would be a point in favour of Marla's view.

Regards,

Rob van Wieringen

Marla Mallett · December 21st, 2009, 12:29 PM

Rob,

Rageth's sample includes 64 kilims. I've not been able to find any summary of the data in the book. But I can give you my quick review.

Of these, the results for only one kilim were returned with probabilities listed completely within the 19th/20th centuries--that is, with no part of the estimated dating listed as before 1800.

The results for seven kilims showed probabilities that completely excluded dates after 1800.

So, according to my quick count of the test results for the remaining 56 kilims, all of those results showed intervals that fell within the 19th/20th century along with date intervals that lie within earlier centuries. Within these, some of the latest date estimates overlapped 1800 just slightly, others by a great many years. When considering the dates given for 95% confidence, as G. Bonani insists is proper, most of the age probabilities range from between two to three centuries; a few of the ranges are a bit narrower, a few a bit wider. A very "typical" example might range from the mid-17th century to the mid/late 19th century. Many seem to fall into this category.

There is no uniformity between the breakdown of the date intervals listed. For some pieces the highest probability intervals form a large percentage, but cover a long span. For some pieces, two or three intervals are given that show more equal percentages in each.

Best,
Marla

Steve Price · December 21st, 2009, 01:09 PM

Hi Marla

OK, there were 64 kilims that a panel of enthusiasts agreed were all pre-1800. Of those, 7 (about 11%) were unambiguously that old by C-14 dating. Many, perhaps all of the other 89% were probably woven after that date.

This forces me to rethink my optimism about Rageth's book being able to provide a database of kilims with documented dates that could form the basis for extracting criteria for dating Turkish kilims. Obviously, it can't.

What to conclude from the fact that the panel of kilimologists who agreed that all the kilims in the sample were woven not later than 1800 were often wrong about that, perhaps with as many as 89% of them? To me, it strongly reinforces skepticism. It seems clear that people who have handled huge numbers of kilims and spent years carefully studying their aesthetics and styles cannot reliably make age attributions as coarse as pre-1800 vs post-1800. And, if their estimates were that unreliable with pieces on which they reached consensus, why should anyone place faith in the reliability of one or two or three expert opinions even if they're unanimous?

Just one side point. Today, Cevat Kanig posted a nice capsule summary of the history of Ottoman rugs. It illustrates that if reasonable documentations of ages exist (as in court carpets; sadly, rare in rustic or tribal weavings), a time line with fairly well defined points in the evolution of style and aesthetics can be constructed.

Regards

Steve Price

Marla Mallett · December 21st, 2009, 02:01 PM

Steve,

Apparent I wasn't clear, or you didn't read what I wrote. The data does NOT suggest that great numbers of those kilims were later than 1800!!! We still just don't know. Nearly all of the pieces are dated WITHIN A TWO TO THREE CENTURY AGE SPAN, and since the dates overlap two or three centuries, covering spans within the 17th century, 18th century and 19th century, for example, that does NOT prove that the vetting was in error....It just says that the piece COULD BE 17th century, 18th century OR 19th century. Some of the ranges are broader than that.

Best,
Marla

Steve Price · December 21st, 2009, 03:34 PM

Hi Marla

My recollection from earlier posts is that the AMS method for C-14 dating cannot give meaningful results for things later than 1800. It gives numbers (dates), but those dates are meaningless beyond placing the item somewhere between 1800 and the present. That is, even though the method gave 1954-1956 in one instance, with a very high calculated probability, the only conclusion that can be drawn is that it dates to some time after 1800.

If the italicized statement above is true 100% of the time (hardly any laboratory finding is, and I doubt that this is one of them), then none of those 57 kilims was pre-1800. But even if the rule about how to deal with ambiguous or unambiguous dates that are more recent than 1800 is correct only half the time (a very poor level of reliability), then the largest number of pre-1800 kilims in the group of 64 would be 36; 7 that were unambiguously pre-1800 plus 50% of the remaining 57. Being correct 36 times out of 64 is still only about 56% correct. And that's in a coarse category sorting, pre- or post-1800. The percentage of correct attributions would become lower if narrower age ranges were used.

I agree with you that C-14 often isn't much help in date attribution beyond separating pre- from post-1800 and, in some cases, putting pieces into even earlier (but still very broad) ranges. And if objective evidence that some people can actually do it by some other method should come up, I'm prepared to become a convert. But until that happens, I don't see the question of date attribution for rustic or tribal weavings as any more answerable than most of the questions about significances of motifs that crop up so often.

Regards

Steve Price

Marla Mallett · December 21st, 2009, 04:44 PM

Steve,

OK...Perhaps it would clarify things if I gave you some random datings, typical of the major group--all with calibrated ages with 95% confidence limit. Looking at these, I wonder if you could still be so sure to put most--or even 50%--in an " almost surely woven after 1800" category. I don't want to list 57 pieces, but I've tried to include a variety of kinds of kilims, to give a rough idea of how these listings play out--excluding any for which the probabilities listed were exclusively prior to 1800.

Plate 10. AD 1644-1694 (27.4%)
AD 1726-1816 (53.0%)

Plate 11. AD 1487-1610 (27.2%)
AD 1611-1689 (37.6%)
AD 1733-1813 (25.0%)

Plate 12. AD 1514-1593 (11.1%)
AD 1620-1696 (38.2%)
AD 1724-1817 (36.4%)

Plate 13. AD 1655-1706 (20.2%)
AD 1714-1820 (55.7%)
AD-1838-1873 (4.5%)

Plate 14. AD 1482-1702 (59.8%)
AD 1718-1819 (29.0%)

Plate 15. AD-1636-1672 (51.4%)
AD 1748-1805 (33.4%)

Plate 16. AD 1635-1709 (29.1%)
AD 1711-1822 (48.9%)
AD 1834-1881 (3.9%)

Plate 17. AD 1623-1779 (43.0%)
AD-1797-1945 (56.5%)

Plate 21. AD 1639-1824 (74.0%)
AD-1828-1886 (7.9%)

Plate 23. AD 1680-1761 (31.5%)
AD 1803-1938 (68.5%)

Plate 24. AD 1659-1823 (70%)
AD 1832-1883 (10.9%)

Plate 27. AD 1644-1694 (27.4%)
AD 1726-1816 (52.0%)

Plate 28. AD 1642-1708 (25.6%)
AD 1712-1821 (50.6%
AD-1835-1880 (5.2%)

Plate 30. AD-1652-1707 (21.5%)
AD 17123-1821 (53.4%)
AD 1836-1878 (6.0%)

Plate 32. AD 1691-1729 (17.8%)
AD 1814-1923 (82.2%)

Plate 33. AD 1427-1692 (86.4%)
AD 1829-1815AD (9.8%)

Plate 36. AD 1649-1706 (22.7)
AD 1714-1821 (53.4%)
AD 1837-1875 (4.7%)

Plate 37. AD 1679-1764 (30.4%)
AD 1803-1938 (69.3%)

Plate 40. AD 1514-1594 (10.75%)
AD 1620-1705 (34.0%)
AD 1715-1820 (38.6%)

Plate 43. AD 1511-1599 (33.3%)
AD 1617-1677 (53.2%)
AD 1773-1801 (8.6%)

Plate 45. AD 1642-1682 (41.3%)
AD 1747-1806 (41.1%)

Well...I'm tired of typing. This represents a pretty typical sampling...I've tried hard NOT to skew it in one direction or another, but just to show how difficult these figures can be.

Best,
Marla

Steve Price · December 21st, 2009, 05:52 PM

Hi Marla

Thanks. I didn't realize how little some of those ranges extend into the post-1800 period. It will simplify things to presume that there is no significant difference between 1800 and 1825. I don't know whether that's actually the case in terms of the limitations of AMS C-14 dating, but it's probably not a travesty.

Your sample includes 21 kilims. Of these, 3 (numbers 23, 32 and 37) are clearly post-1825 weavings; 16 (numbers 10-16, 21, 27, 28, 30, 33, 36, 40, 43, 45) are pre-1825; 2 (numbers 17 and 24) are completely unclassifiable. Let's chalk those two up to methodological screwups and just ignore them.

That leaves 19 that can be sorted into pre- or post-1825 with reasonable confidence. Three of them (about 16%) postdate 1825, contradicting the consensus opinions that every kilim in the book was woven before 1800. The other 84% or so are consistent with the opinions of the experts. That's not a horrendous error rate; I'd call it "not bad". But it doesn't knock my socks off, either, especially because it only tests the panel's ability to sort things into gross categories. Clearly, the finer the categories into which we ask them to place the kilims, the less consensus there is likely to be among them, and the greater the probability that any individual attribution of a particular kilim will be in error. That is, if the kilimologists were mistaken 15% of the time about whether a piece was pre- or post-1800 (lumping together everything from the 15th to the 19th century), assigning each piece to a 50 or 100 year window would make the error rate much greater. I observe that it isn't unusual for certain people to project great confidence when assigning tribal and rustic weavings to windows as narrow as 25 years, and 50 year windows are very common.

Obviously, too, not every expert can be right when there are differences of opinion among them. The fact that the 64 kilim sample excluded every kilim about which there was no consensus skews the statistics. That is, the sample consists of what the experts considered to be slam dunks. That, too, makes the 85% success rate somewhat less spectacular.

To make a long story short, I think the best conclusion one can draw from the data is that if we start with a population of kilims that a panel of experts agree were all woven before 1800, their collective wisdom is correct about 85% of the time, if we cut them a little slack and raise the error trigger from 1800 to 1825.

I keep getting back to where I started: for rustic and tribal weavings, age attributions for pieces made during the past 150 years or so can be very reliable. For things woven before 1850 or so, attribution to a narrow window (I'd call 100 years a narrow window for this discussion) in the absence of documentation (for instance, a provenance or good C-14 data) is virtually impossible.

Thanks for your patience.

Regards

Steve Price

Marla Mallett · December 21st, 2009, 07:50 PM

Steve,

I have to point out that your figures would be more accurate if I was not a lazy typist, and had presented a complete listing for all of Rageth's tested pieces. So we have just made a stab at assessing the results.

Out of the entire group in the book, I would say off-hand that there were maybe three or four kilims that provided true "surprises" for the vetting committee. One example is a Fethiye kilim that I have always wholeheartedly disliked, so I was delighted to see that it was among those that proved to be later than anticipated. It was Plate 59, and the dates: AD 1677-1773 (36.1%) and AD 1800-1941 (63.9%). Harald Bohmer later did some dye tests on this piece and indeed found indigosulfonic acid used for the blue component of greens in the piece. As this dye was used primarily in the second half of the 19th century in Anatolia, the carbon date probabilities seem pretty respectable.

Best,
Marla

Marla Mallett · December 21st, 2009, 07:59 PM

Steve,

Also...keep in mind that in choosing my random sample I EXCLUDED kilims in which all of the date intervals given were PRE-1900.

Best,
Marla

Steve Price · December 21st, 2009, 09:56 PM

Hi Marla

Since the sample of 21 was selected to exclude those with a specific kind of result, it isn't really random in the statistical sense of the word. So, I retract the conclusions I drew from the data. Beyond the obvious - that the panel of experts performed better than what my analysis of that data indicated - there's not much that can be said. After the first of the year my life should be a little less hectic. I'll get a copy of Rageth's book and see what I can extract from it.

Again, thanks for your patience and effort.

Steve Price

Marla Mallett · December 21st, 2009, 09:57 PM

Steve,

I've taken more careful notes now and need to correct my earlier statement about the number of kilims in Rageth's book that might qualify as "surprises"--7 pieces out of the total of 64 kilims which carbon-14 tests gave later dates than those folks apparently expected. Here they are. (Three of these were also on my "sample" list above).

Plate 23. AD 1680-1761 (31.5%)
AD 1803-1938 (68.5%)

Plate 32. AD 1691-1729 (17.80%)
AD 1814-1923 (82.2%)

Plate 37. AD 1679-1764 (30.4%)
AD 1803-1938 (69.3%)

Plate 39. AD 1899-1901 (2.7%)
AD 1954-1956 (97.3%)

Plate 51. AD 1688-1733 (18.8%)
AD 1812-1926 (79.2%)

Plate 59. AD 1677-1773 (36.1%)
AD 1800-1941 (63.9%)

Plate 63. AD 1677-1773 (36.1%)
AD 1800-1941 (63.9%)

So what should we conclude? George Bonani, who did most of the tests, explains: "Especially for objects younger than 300 years, the temporal variations of the C-14 production almost always lead to ambiguities (two to five possible true age intervals). C-14 analyses alone of samples from this historical era are therefore not too meaningful. Only if additional information or other data is available (e.g. historical or stylistic evidence, etc.), the exclusion of certain time intervals can be considered."

Marla

Patrick Weiler · December 22nd, 2009, 01:27 AM

This has been a most enlightening discussion, with two seemingly polar-opposite, mutually exclusive stands:

1 We cannot accurately determine the age of very old weavings, even with carbon dating.

2 We have a pretty good idea how old these weavings are based upon significant art-historical research, handling and experience.

Marla has just stated a link between the two:

"Only if additional (to the carbon dating) information or other data is available (e.g. historical or stylistic evidence, etc.), the exclusion of certain time intervals can be considered"

Without this factor, we would need to consign kilim dating to the dustbin. With it, more likely age ranges deduced by carbon-dating can fairly confidently be assigned. If we depend upon only one or the other, the accuracy of dating is less accurate and dependable.
Considering that significant value, importance and appreciation of these textiles is often dependent upon their assessed age, a combination of these two approaches is desirable.
If we all agree that both the art-historic and carbon-dating methods of assigning approximate ages is bunk and useless, we will see a flood of old kilims on the market for pennies on the dollar.
Therefore, I believe that carbon dating and experienced research is totally useless for assigning age and value to old kilims.
Maybe then I could buy a few of these now worthless, old rags and use them for carpet padding.

Hopefully yours,

Patrick Weiler

Steve Price · December 22nd, 2009, 06:13 AM

Hi Patrick

I hate being the Grinch in this, but I offer some comments about Bonani's statements. Here's the relevant statement (emphasis added):

Only if additional information or other data is available (e.g. historical or stylistic evidence, etc.), the exclusion of certain time intervals can be considered.

He's right, any evidence that would more or less eliminate one of the date ranges would sharpen the results. There's plenty of that for court weavings. Little, if any, except anecdotal evidence for most rustic and tribal weavings. Anecdotal evidence makes a good starting point for framing hypotheses, but cannot be the basis of sound conclusions by itself.

I don't think anyone agrees that both the art-historic and carbon-dating methods of assigning approximate ages is bunk and useless, even for Turkish kilims or other rustic and tribal weavings. It appears that everyone involved understands the limitations of C-14. It can determine whether something was woven before or after about 1750 rather reliably. For those things that it gives pre-1750 dates, the attributions are usually to distressingly wide ranges of dates. And I don't think there's much disagreement about whether kilimologists can put fairly reliable dates on things woven between about 1850 and today, and can pretty reliably sort things into pre- and post-1850 piles. The disagreement is about whether the kilimologists can then put fairly reliable date tags onto the stuff in the pre-1850 pile. Marla (and many others) believes that they can. Until I see some specifics about how they do it and what the foundations for those specifics are, I'll remain skeptical.

Regards

Steve Price

Rob van Wieringen · December 22nd, 2009, 07:47 AM

Maybe I have missed something, but I think it is important to conclude that the possibility of age attributions based on only stylistic evolutioniary markers is actually a fact and not just based on a myth.
This judging mechanism for age attribution has apparently a value on its own, and can develop, apparently, without the use of benchmarks.

The evidence for this is in the figures.
The vetting group made a sample of 64 kilims which were thought to be pre-1800. Roughly I guess their score after C-14 testing was something like 75% correct.
When one would pick a random test sample of anatolian kilims ( not included the ones with obvious synth. dyes ), I guess only some 5% of these would be pre-1800.
So it seems there is something going on here.
On what grounds the vetting group has a significant better outcome as random guesswork would give?
I can only think of the reason Marla Mallett mentioned: age attribution through stylistic evolutionary markers.

Steve's remark (#26) ".....age attributions will continue to be reliable for late 19th century and later, will continue to be guesswork for older pieces." looks a bit overstated, seen in this light.

Regards,

Rob van Wieringen

Steve Price · December 22nd, 2009, 08:12 AM

Hi Rob

Nobody disputes that experts can do a very good job of making attributions of kilims woven after 1850 or so, or that they can do pretty well at identifying a kilim as being older than that. The reliability of the experts at doing this with kilims about which they were unanimous appears to be in the ballpark of 75% to 95%.

The disagreement is about whether they can reliably assign pre-1850 Turkish kilims (or any other rustic or tribal weaving from before 1850) to relatively narrow windows of time. The data in Rageth's book can't address that issue.

I'm more than a little baffled by, This judging mechanism for age attribution has apparently a value on its own, and can develop, apparently, without the use of benchmarks. Are you suggesting that some people are able to reliably identify the times at which old kilims were woven even though there isn't a single documented specimen from which to extract criteria for doing so? I should add that this is the view of the vast majority of ruggies. To me, it sounds like a statement of faith.

Regarding my statement in post 26, ... age attributions will continue to be reliable for late 19th century and later, will continue to be guesswork for older pieces. is carelessly written, and I apologize for that. For "late 19th century", read "1850" or "mid-19th century". The "guesswork for older pieces" doesn't mean that older pieces can't be recognized as being pre-1850, but that assigning them to specific intervals before 1850 is guesswork.

Regards

Steve Price

PS - The elephant in the room mentioned only once, by Marla, is that there are often reasons for doubting expert age attributions even if we believe that the experts can make them. One reason is commerce - sellers of low integrity will misrepresent in order to justify their price. Another is self-aggrandizement.

Rob van Wieringen · December 22nd, 2009, 09:19 AM

You got me wrong there.
I meant solely the judging mechanism , which apparently can develop without the use of benchmarks.
In other words: Apparently it is possible for experts/enthusiastics to use certain skills, by means of stylistic evolutionary markers, and these skills have developed without the use of age benchmarks, because they just are not there in the tribal anatolian kilim history.

In my previous post I just wanted to emphazise the merrits of Marla Mallett's contribution: The possibility of age attribution based on stylistic evolution. Within limts, but still a value on its own.

Regards,

Rob van Wieringen

James Blanchard · December 22nd, 2009, 10:06 AM

Hi all,

Based on this discussion so far, I think we are still missing some very important information about the process for the C-14 dating analysis.

First, it seems that these kilims were selected based on a consensus of experts' assessments that they were pre-1800. But to know the reliability of the experts, we would need to know the total population of kilims that might have been considered and the number of kilims upon which the experts did not agree. For example, if 1 or 2 of the experts believed a kilim was pre-1800, but others did not (and it was not included in the C-14), that kilim has to go into the "unreliable" bin. So, for example, if 75-80% of the kilims upon which the experts agreed were actually pre-1800, but there were another 40-50 kilims that some thought to be pre-1800, but there was no consensus, then the accuracy drops to 50%.

So, to reiterate a point that I made earlier, the issue is not just accuracy (i.e. whether kilims that experts agree upon are of a certain age), but also reliability (i.e. how often experts agree). So, for example, in the present analysis if experts agreed on which kilims were pre-1800 weavings about 60-70% of the time, and they are about 80-85% accurate as a group (based on C-14 dating), then the combined accuracy would be about 50%. In other words, if any one of the experts indicated that a kilim was pre-1800, on average he/she would be correct about 50% of the time (assuming that the experts are equally accurate).

I give this only as an illustration, without data, to show that to develop confidence in the age estimates based on expert opinion, it is critical to assess how reproducible those assessments are. As yet, I have not seen any evidence about how often experts agree (or disagree). I would note that in order to be persuaded about the level of agreement among experts, assessments would need to be done anonymously since achieving agreement through discussion eliminates the concept that experts can use inherent features of a weaving to make independent assessments of age.

James

Marla Mallett · December 22nd, 2009, 11:16 AM

Hi Folks,

OH, you KNOW I'd never suggest that merchants would misrepresent their goods to extract a bigger price! Kilim promotional history actually suggests a slightly different mindset at work.

In the years between 1984 and 1990, prices on early kilims hit fantastic heights. $50,000 kilims were not unusual, some pieces sold for much more than that. This was the time when archaeologist James Mellaart's fantasies supposedly based findings at the Turkish site of Catal Huyuk were being used to promote the notion of a 7000-year-old kilim history, and to promote Anatolian kilims as cult objects. When Mellaart's material was exposed as fraudulent, the kilim market collapsed, and there was much embarrassment all around. Many European dealers were left with inventories of unsold pieces in which they had large investments. I see the carbon-14 dating exercise, begun in the mid 90s, largely as a face-saving venture--whether or not it was completely practical. Ragath's book was published in 1999.

Here's a time-line: Mellaart's first very sketchy introduction of the notion of "Neolithic kilim production" first appeared in a 1983 ICOC lecture in London, then was expanded upon in a 1984 article, published in a kilim book of Bertram Fraeunknecht's. Mellaart's own publication, along with Belkis Balpinar and Udo Hirsch, in late 1989, The Goddess from Anatolia, gave us a full exposé of his theories, along with fraudulent drawings and claims. If you would like a more complete discussion of this matter, see my article, The Goddess from Anatolia: An Updated View of the Catal Huyuk Controversy for a summary of events surrounding this whole debacle. My lengthy original article, published in 1990, A Weaver's View of the Catal Huyuk Controversy, is a detailed examination of the fraudulent evidence with all of the documentation. Both of these pieces were published originally in ORIENTAL RUG REVIEW. Both are included, in their entirety, along with illustrations, on my website, http://www.marlamallett.com/. The link is near the bottom of the HOME page.

Best wishes,
Marla

Steve Price · December 22nd, 2009, 12:41 PM

Quote:

Originally Posted by Rob van Wieringen

... it is possible for experts/enthusiastics to use certain skills, by means of stylistic evolutionary markers ... without ... age benchmarks

What skills are needed to be able to assign times based on style evolution in the absence of age benchmarks? I haven't seen anything more specific than assurances from Believers that it can be done.

Regards

Steve Price

Marla Mallett · December 22nd, 2009, 02:30 PM

Steve,

Well...One could begin with aesthetics courses and art history studies to hone one's abilities to recognize the basic ingredients of artistic evolution and understand works conceptually. That's required in most curatorial fields. Then one must apply these gradually achieved abilities to specialty areas and spend time in more intensive study--both aesthetic and technical. It is difficult (though not impossible) to claim expertise having skipped this basic education.

Marla

Rob van Wieringen · December 22nd, 2009, 03:02 PM

Steve,

We seem to have a communication clash here.
It will be my broddle English, sorry.
I will try again in shortcuts.

Apparently it is possible for experts/enthusiastics to use a certain knowledge in assessing age, concluding from the Rageth figures.
This knowledge is based on stylistic evolution.
As an aside remark, and only that: This knowledge has developed itself without the use of age benchmarks, because they just are not there in the tribal anatolian kilim history.

But the real issue is this:

"The possibility of age attribution based on stylistic evolution. Within limts, but still a value on its own."

Can you agree with this statement?

Regards,

Rob van Wieringen

James Blanchard · December 22nd, 2009, 03:21 PM

Hi all,

I highly value expertise and education, having spent a large part of my adult life in formal education (receiving or dispensing). But I believe that it is incumbent upon experts to both demonstrate and explain the basis for their expert viewpoints. So if experts have developed criteria for dating kilims, then they should be able to relate this with illustrative chronologies and articulate the criteria they use. For me, it is simply not enough to say that the expertise is based on "handling hundreds of old kilims" and applying unexplained aesthetic and technical analyses.

In my field of medicine, clinicians and researchers can spend decades honing their clinical and diagnostic skills, agree substantially with each other, and still be wrong. Consider peptic ulcer disease. There was a huge body of clinical and research literature about the causes (e.g. stress, diet) and treatments, and then we discovered an unexpected cause (a micro-organism, Helicobacter pylori) that completely changed our views about the pathophysiology and appropriate therapy.

This does not mean that I directly equate art studies with medical science, but rather that expertise needs to be openly explained and tested and revised when necessary to gain and retain credibility.

James

Steve Price · December 22nd, 2009, 03:33 PM

Hi Rob

What we know is that when experts have a unanimous opinion that 64 kilims were woven before the commercial period, they are right about 90% of the time. There is no data showing how well they would have sorted those kilims into, say, 100 year intervals.

It should not be astonishing that they are pretty good at identifying commercial period textiles and, therefor, ending up with a population of kilims only 10% contaminated by them.

The ONLY EVIDENCE that they can do what you believe they can do is their insistence that they can do it.

Hi Marla

Can you walk me through a few of the art historical principles that can be applied to date attribution of pre-commercial folk arts and crafts from western Asia?

Thanks, and regards.

Steve Price

Rob van Wieringen · December 22nd, 2009, 04:38 PM

Hi Steve,

Quote: "The ONLY EVIDENCE that they can do what you believe they can do is their insistence that they can do it."

I am not quite sure what you are meaning by this (and by 'they'), but I can asure you I believe nothing, until there is proof of it. Only then I am able and willing to change my position. I will try not to accuse you, but I thought you had the same standards.

To go back to the topic, you are introducing a new parameter: commercial production. For the sake of definition, when did commercial anatolian kilim production started in your view?

Even commercial produced anatolian kilims are part of stylistic evolution, or not?

You didn't react to my question, I will take it as a no.

Regards,

Rob van Wieringen

Marla Mallett · December 22nd, 2009, 04:43 PM

James,

Sure, "Illustrative chronologies" would be lovely, but organizing such a publication would require rounding up pieces, getting permissions for the use of photographs from both individuals and museums, to say nothing of the problems of getting to each of these pieces, which are scattered throughout several countries, for first-hand study. So we normally deal with the pieces in any particular genre on a one-at-a-time basis. There would be the problem of finding financing for such a publication, and then finding someone willing to devote the necessary time and effort to such a project. On sites such as Turkotek, we confront a policy which normally prohibits the discussion of dealers' pieces on the market. That eliminates an immense amount of material--including many pieces that appear in some of the most prominent publications--listed euphemistically as from a "private collection"...often meaning a dealer's inventory. Many of the pieces in museum or conference exhibition catalogs have indeed been pieces owned by dealers. A review of the list of lenders to the Rageth book, reveals a large number of kilim dealers' names.

Keep in mind that a majority of kilim publications have been promotional materials produced by dealers. These authors' concerns lie primarily with promoting their merchandise, and there is little to be gained by comparing their pieces with those belonging to others. Yet because clients always want dates attached to individual pieces, dealers are obliged to furnish their best estimates. A number of kilim books have been vanity publications, in which collectors have proudly displayed their possessions. Who among these individuals is going to want to get into the "mine is older than yours" game? If the rug world is a small one, the kilim world is much smaller still, and there are endless problems in discussing and comparing pieces that belong to friends, colleagues or customers.

Best,
Marla

James Blanchard · December 22nd, 2009, 04:59 PM

Hi Marla,

I suspected that the situation was as you have described, which is perhaps understandable though disappointing.

I suppose that one could still develop a "review" article that organized sub-sets of published kilims along some rational chronology, along with an explanation of the criteria that were used to order them. This could perhaps be a more structured sort of exposition than Tom Cole's interview with Jerry Anderson, wherein the focus was on tribal attribution more than age. I, for one, would be very interested to read this.

Might I also remind folks that Steve Price actually wrote an article along these lines illustrating design evolution for Turkmen trappings (http://www.rugreview.com/1juvals.htm). What I liked about that article was that he explained the various criteria he used to place the weavings into different age categories. In my view, we need more of this sort of specific articulation of criteria for age categorization. At the very least, we could better understand the basis for experts' assessments.

James

Steve Price · December 22nd, 2009, 06:06 PM

Hi Rob

"This" refers to the ability to reliably sort the dates of kilims that are earlier than 1850 or so. "They" are the people who claim that experts can do it.

The "commercial period" doesn't have a precise starting date. But it appears to have begun around 1825-1850 in central and western Asia and was firmly in place by 1900.

You seem offended by my saying that the only evidence that "they" can attribute relative ages to pre-commercial tribal and rustic arts and crafts from western Asia is "their" claim that "they" can. Do you know of any other evidence? I don't.

Regards.

Steve Price

PS - I think the art historical approach basing date attribution on style evolution works pretty well with court weavings. Why? Because of the documentation of the dates for many court weavings

Rob van Wieringen · December 22nd, 2009, 07:03 PM

Hi Steve,

I am not offended, just disappointed in your ability to reason logical and to stay on that track.
It is obvious to me you are emotionally to much involved in this matter, and you are unable in the end to tackle rational arguments with other rational arguments.
I am not a Believer, as you like to put it, not at all.
My only intention in this discussion was to come to some kind of conclusion, formed by logical reasoning, whatever the outcome might be.
This seems to be impossible, alas.
In the end I am sorry to have to conclude that you are as much a Believer in this matter as the Believers you so much dispise.

Best regards,

Rob van Wieringen

Steve Price · December 22nd, 2009, 07:40 PM

Hi Rob

I wrote that I know of no evidence that anyone is able to reliably assign moderate ranges of weaving dates to pre-commercial rustic or tribal weavings. You seem certain that it can be done. I asked you to cite evidence for your position. You ignored the question and replied with ad hominem remarks.

I direct your attention to the rules of engagement in the paragraph at the top of this page. Please stay within them.

Thanks.

Steve Price

Marla Mallett · December 22nd, 2009, 07:41 PM

Steve,

I can't imagine where you got the notion that kilim weaving in Anatolia was "commercialized" in the 1825 to 1850 period and was "firmly in place by 1900." "Commercializati on" of production normally means that pieces were made for sale outside the community producing them. That didn't begin in Turkey until perhaps 35 or 40 years ago. Prior to that, kilim weaving was strictly limited to the production of pieces made for home use...first by nomadic groups. As these nomads gradually settled, they continued to make a few pieces for use in their own houses--mainly for sedir covers and wall hangings. it was pile-rug weaving that was commercialized and products made strictly for the marketplace, even for export to Europe in very early times.

Best,
Marla

Steve Price · December 22nd, 2009, 08:02 PM

Hi Marla

I stand corrected on that. My belief until now was that commercialization of Turkish rustic weaving was more or less simultaneous with commercialization of Turkmen, and tribal/rustic NW Persian and Caucasian weavings.

My apologies for the misinformation and thank you for correcting it. In the posts in which I refer to "precommercial period", I should have been saying pre-1825 or pre-1850.

Regards

Steve Price

PS - The fact that the kilimologists didn't have a corpus of commercial post-1825 kilims on which to rely makes their success in assigning 90% of 64 pieces to the correct side of that dividing line less ho-hum than I thought it was.

James Blanchard · December 22nd, 2009, 08:57 PM

Hi all,

Perhaps it would be useful to get back to Horst's kilim, and the basis for date estimations. From my reading of this thread, date estimations range as far back as the 15th century, with several kilims of a similar type being shown along with various dates ranging up to the 19th century.

Since I am almost certainly the least well-educated in this field, here is my attempt to put a few kilims in chronological order, based solely on bits and pieces of conventional wisdom that I have acquired over the past few years.

Okay, here is my list, starting from presumed oldest to presumed most recent. I have tried to make this selection without referencing the dates listed by publishers, though I am sure I have been somewhat biased by that information. I don't think the latest example (#5 below) fits exactly with this group, but my naive assessment would place it latest among this group anyway. I would be happy to hear a critique from more experienced folks, along with their rationale...

James

1.

2.

3.

4.

5.

Marla Mallett · December 23rd, 2009, 02:57 AM

James,

OK. I'll make a stab at comparing these pieces. First of all, I should say that every different kilim genre requires a slightly different kind of analysis, so that comments regarding these pieces may not apply to any another group.

I would definitely rearrange your sequence, James. I would switch your numbers 2 and 3. I would eliminate your number 5, as unrelated. Although it will be a bit confusing, I'll use your numbers for my commentary.

With this group, the most obvious factor to consider is the relationship between the primary and secondary motifs in each piece, and the ways they occupied the space. The most carefully considered placements are present in #1 and your #3 (Horst's kilim). Most of these small hooked medallions were carefully placed in pairs to fit alongside the indentations of the large medallions. They were important auxiliary design elements. By the time the #2 kilim weaver came along, that relationship had been lost and small secondary hooked medallions were just squeezed in--as many as would fit, making a busy background. By the time this major theme was adopted by the #4 weaver, any meaningful relationship between primary and secondary motifs was completely lost.

When we consider spatial relationships and proportions, we can see that in the earliest pieces there was a tendency for the artisan to give each separate motif plenty of room, because when weaving without a cartoon or model, it's hard to calculate their placement. When building a kilim design from bottom to top, it is oftentimes quite difficult to locate the exact proper starting point for each downward projecting design part. The weaver of #1 was very cautious and methodical; she was extraordinarily concerned with clarity and with the dominance of her primary motifs. In order to retain a sense of stability and continuity in her rather wide-open layout, she injected very small hexagonal tertiary motifs. By the time of #3, Horst's piece, everything had become more compact. Closely packed parts are much easier to handle if the weaver has either experience with the design or a model on hand to copy. Very small tertiary motifs present again in this kilim, but they are no longer so necessary for continuity or evenness of tone in the patterning. The weaver of this bold piece seems to have been far more concerned with connectivity, and the presentation of ONE large coherent statement, instead of a series of repetitive parts. There was a quite different attitude toward the design. The small scattered bits and pieces seem to have been inserted because such "fillers" were an "expected" part of this traditional layout, but here they seem like extraneous clutter. By the time the weaver of #2 came along, the carefully planned relationship between primary and secondary motifs was lost, size relationships changed, and the placement of small hooked medallions became arbitrary. Even half medallions were injected, simply to fill background spaces. The hierarchy of the various parts in the earlier pieces was substantially altered. Instead of the spare refinement of the first piece, we now find squat hooked hexagons with a busy background of fussy and relatively insignificant secondary features.

Both #1 and #3 kilims have consistent and well articulated motifs throughout. #1, which I would guess to be probably a 17th century piece, conveys a sense of restraint and elegance-- with careful, rather pinched and isolated motifs that are combined with logic and clarity. I would expect a significant time gap between this first piece and #3 (Horst'spiece), with its more generous forms and large, bold statement. It seems quite characteristic of the synthesis that occurred in many 18th and very early 19th century Konya area pieces. #2, most likely a mid- to late 19th century piece, displays a less logical, much more intuitive rendering of the basic idea. The same parts are present, but they are used in a more careless manner. By the time we get to #4, which I would suggest is a 20th century piece, all consistency and any clear logical relationships between the elements are gone. It's just a collection of separate parts and pieces plopped on a large white panel, with little thought given to any integration of the parts. it's a "traditional" piece in name only. The transition has thus run its course from a strong concept through gradually more mannered, hackneyed renditions. Though the same motifs are present in these four kilims, they are four radically different artistic expressions. The first is regal, the second (#3) is bold and dynamic, the third (#2) is dowdy, and the most recent is disjointed and bland. The differences are in style and attitude--not in the way individual motifs are formed.

The relationships between positive and negative space are always important in slit-tapestry design...That's more complicated to discuss with these pieces than with some other types. This goes beyond the simple consideration of reciprocal motifs.

Relationships between border systems and field are normally important in making stylistic comparisons. Here, since #2 is missing its borders, unfortunately, we have to ignore this aspect.

Beyond these observations, when speculating on dates and even proper sequencing, it is immensely helpful to compare these pieces to the general aesthetic displayed by groups of other kilims from the same general geographic area. In this case, #5 does not provide a good parallel.

Well, one could go on and on, and speak in esoteric terms of the "spirit" of each piece or the conceptual notion behind each composition; I've tried to refrain from that this time around. It's impossible to talk intelligently about color with these photos, as the on-line versions distort even the two Ragath book plate renditions. Likewise, without either handling the kilims or at least seeing excellent close-up photos, it is impossible to consider the critical relationships between weave balance and patterning that is such an important part of the aesthetic development in kilims. After all, these are WOVEN STRUCTURES, not just flat patterning. The scale of each weave element is critical in fashioning the pattern. The way the whole series of scale relationships are handled within a piece should be an integral part of one's judgment.

Within each generation of weavers there were artists, there were competent craftsmen, and there were women who simply produced kilims because that was expected of them. Considerations of artistry and excellence of design must be separate from judgments concerning age and design evolution. We cannot assume that designs always deteriorated...that quality was always spiraling downward. When copywork is involved, artistic degeneration can be expected. But whenever a creative individual produced her own fresh take on an old idea, or took off on her own, there was a good chance of superior results.

Well, this is a quick evaluation. With the pieces at hand, and a little more time, one could surely produce a better summary.

Best,
Marla

Steve Price · December 23rd, 2009, 06:57 AM

Hi Marla

Many thanks. Your structure-related perspective helps a lot.

Regards

Steve Price

Marla Mallett · December 23rd, 2009, 11:42 AM

There's one more thing I could say about this topic in general, and the difference in the ways we approach pieces from different collecting areas in order to assign even general dates.

In analyzing Turkmen pieces, for example, we deal almost entirely with production that fits within a 100-year period. When we date those pieces, or even just try to put them into reasonable sequences, we see that a lot of simple copywork is involved and a broad set of general principles legitimately apply. Analysis becomes almost solely a matter of pointing out quality degeneration over the century. It is practical to speak of layouts becoming more compact (less "spacious" to use a favorite ruggy term). From almost every tribal group the pieces become increasingly crowded, as borders and excess ornamentation proliferate. Individual motifs may also become degraded.

With available Anatolian kilims, which span five centuries and present an extremely wide variety, the problems of analysis become far more complex. There are vast gulfs between pieces from the Northwest, Southwest, Central area , Southeast/Syrian areas, and the primarily Kurdish areas of the East. Brocaded cicms form almost a separate category. Thus each genre must be treated separately, and there are few general principles that can be applied across the board. it's easy to confuse quality issues with design evolution.

Best,
Marla

James Blanchard · December 23rd, 2009, 12:02 PM

Thanks Marla,

This is very enlightening.

What approach to you take to estimate the time lag (in years or generations) in the evolution of designs?

Are there any specific reference points that can be used to know how to peg a particular stylistic expression to a particular era?

James

Marla Mallett · December 23rd, 2009, 01:19 PM

James,

You've asked: What approach do you take to estimate the time lag (in years or generations) in the evolution of designs?

I'd say fancy guesswork. Especially without the pieces on hand to examine. With the four pieces in your series, I see the largest time gap between the oldest and the next oldest piece. All of the same design elements are there, and the same theme, but the way in which they are used is radically different. The attitudes toward the pieces differ. We have no choice except to work backwards on a time line. The general character of Horst's dynamic piece is so typical of work from the late 18th/early 19th century, that it seems the easiest to place. There seems to have been a significant period between it and the earlier piece. I can't say how much, but three or four generations of weavers would not seem unreasonable. In other words, a granddaughter or great-granddaughter of the first weaver may have woven Horst's piece. Any closer connection is a little hard to believe. They were living in different times, and styles changed, even though they worked with the same elements. This is necessarily an intuitive approach, but when we see that the carbon-14 dating actually places the first piece most likely into the early part of the 17th century, I feel pretty comfortable with my assessment.

Working forward in this sequence, the piece that I've placed in the mid-19th century (your #2) has become so disconnected from the "tradition", that its general character has been lost. The "theme" has been distorted--which is a different issue entirely from the fact that the main motifs are awkwardly articulated.

Best ,
Marla

Horst Heinzlreiter · December 24th, 2009, 01:07 AM

Wow, what an input all around here!
Thank's so much MARLA for your object-lesson, your "kilim critique" is an eye-opener, I have never got such interesting detail informations.
Since I have posted my fragment I have learned quite a lot in doing so.
Great forum, bravo TURKOTEK

Best regards Horst

Steve Price · December 31st, 2009, 08:52 AM

Quote:

Originally Posted by Marla Mallett

"Commercializati on" of production normally means that pieces were made for sale outside the community producing them. That didn't begin in Turkey until perhaps 35 or 40 years ago. Prior to that, kilim weaving was strictly limited to the production of pieces made for home use...

Hi Marla

I ran into this sentence about Turkish kilims, written by Jack Cassin ("RK" is his nom de keyboard) just two days ago. By 'seriously old' RK means kelim appearing to have been made well before the orientalist ‘craze’ created foreign demand and the ensuing beginning of commercial production that began in the late 1880’s to satisfy that demand.

The late 1880's is more than 80 years earlier than 35 or 40 years ago. Do you know why he might think that, or is it likely to be pure fabrication? He does have the IPAA (International Pants Afire Association) Lifetime Achievement Award, after all.

Regards

Steve Price

Marla Mallett · December 31st, 2009, 01:17 PM

Hi Steve,

The commercialization and export of knotted-pile carpets was certainly given a major boost by the Orientalist Craze in the 1880s--both in the US and in Europe. But kilims were not a part of that commercialization, and remained largely unknown in the West until quite recently. Senneh kilims, 16th and 17th century silk Kashan kilims, and large Caucasian soumaks were among the few types well known in the West.

When Anthony Landreau and Ralph Pickering mounted a Textile Museum (Washington) exhibition in 1969 and published a catalog FROM THE BOSPORUS TO SAMARKAND: FLAT-WOVEN RUGS, nomadic flatweave pieces were new to most people. Of the 112 pieces in that catalog, only 10 kilims and bags were Turkish. A great many pieces in the catalog were erroneously identified, because so little was known about them, and question marks appeared routinely in the labeling. Caucasian and NW Persian pieces dominated the exhibition, especially small soumak bags from the Caucasus and NW Persia. Pieces at that time had been gathered up by a few collectors only; they were strictly ethnographic items, just as were the wooden farm tools, spinning wheels, etc. that they also collected. Ethnographic museums in Europe were also beginning to put together collections. Josephine Powel, for example, gathered materials for Amsterdam's Instituut fur den Tropen and Rotterdam's Ethnographic Museum.

In May of 1977 David Black and Clive Loveless opened an exhibition of kilims in London and published a catalog, THE UNDISCOVERED KILIM. It included 55 kilims--Persian, Caucasian and Anatolian. It was an eye-opener for most of us. They stated in their introduction that they had often paid ten to twenty pounds for each, because in the trade, kilims were considered inferior. Bales of pile carpets were shipped to the West in the nineteen hundreds wrapped in old kilims, and these pieces were often unraveled to use for re-piling old knotted carpets.

Old kilims turned up during these years only in odd places in the US. I occasionally encountered a piece in the back of an Armenian rug dealer's office--a curiosity that he had picked up on his travels. I remember seeing a Caucasian kilim in the background of an old photo of sculptor Louise Nevelson. I, myself, bought a large crock for $15 in a Minnesota auction in 1964 that had an ancient Caucasian kilim stuffed inside. In about 1975, I found a very ragged northeastern Anatolian prayer kilim in an Atlanta flea market. But these were rare events. Rug dealers, for the most part, knew nothing of kilims. Shortly after that, I found a Turkish rug dealer and a Persian dealer in New York who had private collections of old kilims, and who were starting to sell a few late 19th century kilims. But there was, at that time, NO commercial kilim production on the market.

David Black, himself, says , "I first came upon them (kilims) in Greece in the sixties when I had just left school. I stumbled upon them almost by accident, walking through the flea markets of Athens at a time when I was experiencing an increasing and I suppose, somewhat romantic attraction for the Orient." He goes on to say, "I did not find a great quantity of kilims there. They were few and far between. Moreover one often had to seek them out in some very unlikely places such as the haunts of Gypsy traders who crisscrossed the countryside exchanging new for old, in an age when the cult of the acrylic blanket reigned supreme! However, the kilims one did find there were usually very good. A quirk of history, namely the influx of immigrants from Turkey in the twenties, meant that a number of wonderful kilims found their way to Greece. "

David Black continued (in the 1977 catalog), "Very few collections of kilims exist, and most of these have been built up in the last ten years. In Tehran they were not considered luxurious enough to be put on the floor of town houses."

When I first went to Turkey in 1980, there were plenty of crude, ugly synthetic-dye kilims available, but NO commercial production. I visited semi-nomads in the countryside who were still producing a few kilims for their own use--mainly for sedir covers and wall hangings. Commercial production only began a few years later when a couple of Turkish entrepreneurs started very small productions of natural-dye kilims, usually with small prayer-rug formats, in an attempt to cash in on the new interest in kilims in the West. Harald Bohmer tried to interest women in the Ayvacik area to weave brocaded cicims for his DOBAG project in the 1990s, but he was only able to find one elderly woman willing to do so. She is pictured on my website; one of her cicims is on the floor of that photo. Others complained that brocading was too difficult and too time-consuming.

To sum up, there are indeed lots of crude, synthetic-dyed kilims now on the market--pieces made in the 1900s. But these were ethnographic pieces made by nomads and villagers for their own use, not for commerce--because there was virtually no market for such things until recently. We do better to speak of Pre-Synthetic Dye Kilims, rather than Pre-Commercial Kilims.

Best,
Marla

Steve Price · December 31st, 2009, 01:44 PM

Hi Marla

Many thanks. I didn't take Cassin's statement any more seriously than I take most of his other statements. But, you never know - a stopped clock shows the right time twice every day.

Regards

Steve Price

Joel Greifinger · December 31st, 2009, 02:27 PM

Hi Marla,

Thanks, once again, for sharing your knowledge and experience.

Happy New Year,
Joel Greifinger

Horst Heinzlreiter · January 1st, 2010, 10:30 AM

To feed the thread with pics.

An other old one - after cleaning and before it was mounted:

And in show condition:

Best regards Horst&Helga H.

November 25th, 2009, 03:21 PM	1
Horst Heinzlreiter Members Join Date: Nov 2009 Posts: 7	kilim fragment Dear forum, let me show you a kilim frag, which is not mounted professionally yet. It comes from the late Tom Weisbuch (California). Would be great, to get some ideas about this stuff, origin, age etc. Thanks in advance Horst

November 27th, 2009, 02:01 PM	2
Doug Klingensmith Members Join Date: Jul 2009 Posts: 1	Wonderful old fragment Horst. Thanks for the look. My uninformed amateur first impression is Central Anatolia (Konya?) d.k.

December 2nd, 2009, 09:50 AM	4
Horst Heinzlreiter Members Join Date: Nov 2009 Posts: 7	Thank you for the pictures and your guess about the origin of this kelim. The striped border on the pics,which surrounds the main motive, is really similar to our kelim, but I have not seen this big hooked ends on any other kelim, did you? Once again,thanks to all! Horst

December 8th, 2009, 05:27 PM	11
Michael_Wendorf Guest Posts: n/a	central Anatolia Dear Mr. Heinzlreiter, I was just about to add some of your most recent references when you posted. I think the most one can say is Central Anatolia perhaps closer to Sivas. There are lots of ivory ground kilims from greater Central Anatolia so one should be careful, not all are necessarily related in design. The main point I wish to offer is a comparison to the piece from the collection of David Lantz of New York. This is plate 33 in Rageth's Anatolian Kilims and Radiocarbon dating. Also there are close comparisons with three other kilims including plate 34 from a "private collection". In these pieces the C wrenches extend out like yours. Lantz's piece is dated 1427-1692 with 86.4 confidence limit. Characterized by large C-shaped pendants at the hexagons Rageth writes the secondary motifs (small hooked hexagons) always form a 4 + 1 composition together with the primary motif while the large hexagons form a center on the vertical axis - as does your piece. The border design of the Lantz piece corresponds to the secondary field design - also as yours. I think your piece could be 17th century though it is hard to see the colors clearly on monitors. In any event Tom Weisbuch was a strong collector with a keen eye and you must be very pleased to have such a evocative kilim from his collection. Best wishes, Michael Wendorf

December 9th, 2009, 05:03 PM	14
Horst Heinzlreiter Members Join Date: Nov 2009 Posts: 7	Thank you all so much for your inputs. @ Marla you were the guide - you made my day! @ Michael: Yes, it would have been a great thing to see all the pieces of Mr. Weissbuch's collection. We are proud of the kilims we were able to catch. To legitimate my reply, please let me show an other kilim fragment: It is mounted, dimension - 222x157cm and probably Yoruk? Best regards Horst@Helga

December 18th, 2009, 02:48 PM	38
Steve Price Administrator Join Date: May 2008 Posts: 80	Hi James Michael Wendorf's post #29 describes something similar to your suggestion about experts making independent age attributions of what they agreed were pre-1800 Turkish kilims. Regards Steve Price

December 19th, 2009, 06:40 PM	45
Marla Mallett Members Join Date: Jul 2008 Posts: 27	One more thing: Georges Bonani, who has written the technical section of the Ragath book, emphasizes that "Whenever calibrated data is reported, ALL true time intervals have to be quoted. It is NOT permissible to only report the interval with the highest probability and to omit the intervals with the smaller probabilities." Marla

December 20th, 2009, 09:22 AM	52
Marla Mallett Members Join Date: Jul 2008 Posts: 27	James, Why would anyone embark on a project to carbon-date pieces? To SELL them. Marla

December 20th, 2009, 10:10 AM	54
Steve Price Administrator Join Date: May 2008 Posts: 80	Hi Joel I think the fundamental difference of opinion is whether we can attach a time line that has reasonably well defined points of known states in design evolution, assuming that design evolution is unidirectional. Marla (and many, many others) believe that we can. I don't, nor do some others. Regards Steve Price

December 20th, 2009, 12:32 PM	55
Marla Mallett Members Join Date: Jul 2008 Posts: 27	Of course design evolution is NOT unidirectional-- not in kilim weaving or any other of the visual arts. Marla

December 20th, 2009, 02:47 PM	58
Steve Price Administrator Join Date: May 2008 Posts: 80	Hi Marla Is it correct that the term "design evolution", used throughout this thread and elsewhere, actually means "style evolution"? Regards Steve Price PS: I think the focus collectors often place on motifs is an outgrowth of the common notion that the "magic" in a tribal textile has to do with the meanings of motifs.

December 20th, 2009, 05:51 PM	60
Steve Price Administrator Join Date: May 2008 Posts: 80	Hi Marla Thanks for the clarification. Regards Steve Price

December 21st, 2009, 07:59 PM	70
Marla Mallett Members Join Date: Jul 2008 Posts: 27	Steve, Also...keep in mind that in choosing my random sample I EXCLUDED kilims in which all of the date intervals given were PRE-1900. Best, Marla

December 23rd, 2009, 06:57 AM	95
Steve Price Administrator Join Date: May 2008 Posts: 80	Hi Marla Many thanks. Your structure-related perspective helps a lot. Regards Steve Price

December 23rd, 2009, 12:02 PM	97
James Blanchard Members Join Date: Jun 2008 Posts: 16	Thanks Marla, This is very enlightening. What approach to you take to estimate the time lag (in years or generations) in the evolution of designs? Are there any specific reference points that can be used to know how to peg a particular stylistic expression to a particular era? James

December 24th, 2009, 01:07 AM	99
Horst Heinzlreiter Members Join Date: Nov 2009 Posts: 7	Wow, what an input all around here! Thank's so much MARLA for your object-lesson, your "kilim critique" is an eye-opener, I have never got such interesting detail informations. Since I have posted my fragment I have learned quite a lot in doing so. Great forum, bravo TURKOTEK Best regards Horst

December 31st, 2009, 01:44 PM	102
Steve Price Administrator Join Date: May 2008 Posts: 80	Hi Marla Many thanks. I didn't take Cassin's statement any more seriously than I take most of his other statements. But, you never know - a stopped clock shows the right time twice every day. Regards Steve Price

December 31st, 2009, 02:27 PM	103
Joel Greifinger Members Join Date: May 2008 Location: Massachusetts Posts: 14	Hi Marla, Thanks, once again, for sharing your knowledge and experience. Happy New Year, Joel Greifinger

January 1st, 2010, 10:30 AM	104
Horst Heinzlreiter Members Join Date: Nov 2009 Posts: 7	To feed the thread with pics. An other old one - after cleaning and before it was mounted: And in show condition: Best regards Horst&Helga H.