Oriental Rug Aesthetics in 1910
Dear folks –
For most of my early life, until age 12, I lived in a very small town in Pennsylvania’s Susquehanna Valley, Avis by name. Going to school we used to walk by a small, white clapboard Church of Christ one block off Main Street.
Two weeks ago while traveling I discovered that this church has, for no reason I can discern, become a sizable used book store full of shelves 10 feet tall. In it I paid a little too much for an interesting rug book.
Its simple cover belies the contents of this volume. It is a “Catalogue of a Loan Exhibition of Early Oriental Rugs” that opened in New York City’s Metropolitan Museum of Art late in 1910. Its 41-page introduction was written by Wilhelm Valentiner, then the “Met’s” Curator of Decorative Arts. My copy is one of 1,000 printed.
Mr. Valentiner’s views seem likely to represent those of those who saw themselves as the foremost experts on oriental rugs at this time. He says some interesting things.
First, he says that the exhibition had to draw on loaned rugs mostly in private hands because in 1910 “no institution in this country has as yet a collection of old rugs equal in any way to the collections in nearly every large European museum, especially those of London, Paris, Berlin and Lyons.” The Textile Museum here in Washington, D.C. was not to be founded for another 15 years.
The list of lenders to this 1910 "Met" exhibition contains some names with which many of us are familiar.
Second, Mr. Valentiner has strong views about what rugs are “worthy” of consideration by folks serious about them. Here are a few passages (they are too good not to quote at length):
He says (and claims that Bode, Martin and Sarre agree with him) that “three centuries are especially distinguished by their excellence in rug weaving, namely, the fifteenth, sixteenth and seventeenth.” Valentiner says that the “superiority” of the rugs of these centuries is “distinguished by their infinite varieties of pattern, their clarity of design and wonderfully rich harmonies of color.”
Sadly, Valentiner goes on, the superiority of 15th, 16th and 17th century weaving “is not yet recognized by the educated public.” They seem not to see that “seemingly old types such as Ghiordes, Ladik, Meles, Kula, Bokara and other fabrics which, as a rule, are not older than 1750 and usually show a lack of clearness in design and a weak sense of color relations typical of periods of decadence.”
I love "decadence." A real rug world "swear word," signaling something likely more general and even worse than "degeneration."
Valentiner quotes Martin on this point: “It is incomprehensible that collectors who know splendid oriental carpets can be so fond of such poor work as the Ghiordes and Kula carpets which one sees now in almost every collection.”
Valentiner allows that there may be some 18th and 19th century rugs that might serve as “appropriate and charming floor decoration, but they never stand comparison with those made in the earlier periods.”
He goes on “The eighteenth and nineteenth century rugs of different manufacturies repeat the same pattern over and over again; a pattern which is generally a misunderstood imitation of a design of the sixteen or seventeenth century, and one that has often lost the meaning of the older conventionalization of natural forms. Every old rug, on the contrary, has a marked individual character showing a design that has never been exactly repeated and is alive with the personal quality of every great work of art. In more modern types it is seldom possible to determine the meaning of a single motive, to know whether flowers, animal forms or purely geometric designs were intended. It is even difficult to decide which is the ground and which is the pattern, and in what connection the border stands to the centre field, or, in fact, even where the border begins. The pattern is always overcrowded, lacking in the noble simplicity which is characteristic of the old rugs as it is of all really great works of art. And the modern color-schemes are as restless as the designs; they lack the power and sobriety of the old rugs of the fifteenth and sixteenth centuries or the subtlety and delicacy of such fabrics of the seventeenth century as the so-called Polish carpets.”
Valentiner and his fellows are particularly exercised about collectors choosing eighteenth and nineteenth century rugs over older, more worth worthy ones because “It is seldom a question of price, as the sums paid for Ghiordies rugs and other weaves of the late eighteenth and early nineteen centuries are often greater than those paid for good Asia Minor rugs of the fifteenth and sixteen centuries, which belong to a period of the highest artistic feeling.”
So such purchases were not only aesthetically impoverished, they were financially inexcusable as well.
Now it is hard not to smile at these words since in 2006, rugs woven in the 19th century are the center of those seen as worthy of collection and a collector who owns a rug that was demonstrably woven in the 18th century is likely display that fact in a prominent way. But a late 19th century rug was essentially “new” in 1910 and fifteenth, sixteenth and seventeenth century rugs were, apparently, not only available, but priced (if Valentiner is to be believed) at levels accessible to those who could buy a Ghiordies at that time.
But is seems clear that the arguments being made by Valentiner and his fellows are mostly aesthetic and are very similar to those being made nowadays to distinguish pieces woven in the 20th century from the more meritorious ones now said to have been woven in the eighteenth and first half of the nineteenth century. This similarity demonstrates to me again how subjective, or at least socially constructed, the aesthetic values of a given age seem to be and make me smile about the certainty with which many experienced folks in the rug world believe that they (and perhaps only they) have “a good eye.”
I am not sure we will want to pursue this issue much in this thread, but one could attempt to collect some images of rugs from the 15th, 16th and 17th centuries and to compare them with similar pieces from the 18th and 19th centuries to see if we can discern the differences that so exercised Mr. Valentiner and his peer experts in 1910. The catalog itself provides some 50 examples, but they are all in black and white photographs.
R. John Howe
Thanks for demonstrating what the French say: plus que ça change, plus que c'est la même chose.
On the other hand, I wonder how good my collection will be considered in a couple of centuries.
Well, let’s make three centuries, to be on the safe side.
Thanks for posting excerpts from this book. Fascinating! And the comparisons you suggested are very worth while, in my opinion. Some of my books I purchased exactly for this exercise. I am not sure, however, if I agree with your conclusion.
But is seems clear that the arguments being made by Valentiner and his fellows are mostly aesthetic and are very similar to those being made nowadays to distinguish pieces woven in the 20th century from the more meritorious ones now said to have been woven in the eighteenth and first half of the nineteenth century. This similarity demonstrates to me again how subjective, or at least socially constructed, the aesthetic values of a given age seem to be ...
Hi Tim -
I would not hold that there is not such a thing as an ugly rug or that conventionalization cannot be noted and evaluated.
What I object to is the seeming conviction by Mr. Valentiner and his peer experts that they have discerned the basis for determining what rugs have aesthetic merit.
Take conventionalization, for example. Sometimes (often?) it can have results that are not pleasing, but not always. Jerry Silverman spoke up recently about a Caucasian rug precisely to indicate that, for him, the conventionalization of the design elements had moved to a very lean position indeed. His seemingly approving description suggested that, for him at least, there can be Mondrian-like versions of some rug designs that are very pleasing.
My own take on Valentiner and his expert peers is that they seemed to be "hooked" generally on "old" and are insisting, not always consistently, that there was a kind of aesthetic "cliff" in about the 18th century and the world of rugs went roughly to hell then, and that the only viable solution is to go back to the "golden" years of the 15th, 16th and 17th centuries.
Now it is easy, and a bit unfair, to pick on their language, so I am going to take the assertion of Valentiner and his expert peers seriously and put up images of some of the sorts of pieces they seem to recommend (I have also found at least one example of a dreaded Kula). This may permit others to offer counter examples from later periods that they feel have real merit in the aesthetic world.
R. John Howe
I think I understand that you are going to post some excerpts from the catalog. Excellent. It seems that some of the reasons for which Valentiner damned the 19th century rugs, e. g., ammbiguity between what is background and what is design, or the repetition of motives where the original design was forgotten, are some of the reasons for which connoisseurs today prize the same rugs.
Given Valentiner's comments, are there any rugs shown in the catalog that you were surprised to find there? Anything we would today call 13th or 14th centuries, or earlier? Seldjuk, as contrasted with Ottoman?
By the way, I see him invoking the term "decoration," to be contrasted, presumably, with "art." The use of the phrase, "decorative carpets," usually damning them with faint praise in order to exalt "artistic" carpets, was always a pet irritation of mine.
Is it not generally accepted among carpet scholars that the 16th and 17th centuries were the high points in carpet manufacture, which has declined since then? Based on your quotes, I think this is basically what Valentiner is saying.
He may be a bit too dismissive about 18th and 19th century weavings, but if the comparison is with the most accomplished carpet achievements he has a point I think. The reason why most collectors would proudly show off their 18th century rugs is because the point of reference nowadays is mostly 19th century.
The only pharse I find objectionable is, "Every old rug, on the contrary, has a marked individual character showing a design that has never been exactly repeated and is alive with the personal quality of every great work of art." To say that every old rug is great is probably a bit too strong. Similarly, it would be too strong to say that every rug of the 18th and 19th century is inferior to older rugs, and I would expect Valentiner would agree. I think he is talking about gerneral tendencies.
I don't suppose John has much quarrel with Valentiner's opinion as to the superiority of the older rugs. The remarkable thing is his near contempt for the 18th and 19th century rugs that are so prized today.
We do have to concede Valentiner's point about the vogue of certain Turkish prayer rugs in that time. They went on a kind of "tulip madness" binge, if I've got it right. And I think some of the highest fliers weren't even the best examples.
Dear folks -
Valentiner presents and describes in this catalog rugs from each of the centuries that he and his expert peers see as the "worthy" ones. For the 15th century he cites rugs from "Asia Minor and Syria," this include classical Turkish rugs and what we now call "Mamluks" from Eygpt. His examples from the sixteen century seem to center on Persian pieces and those from the seventeenth feature the Persian "Polish" rugs and some Mughals.
If you look again at Valentiner's list of lenders you will see that a Mr. Williams is among them. Williams' rugs eventually came to the Philadelphia Museum of Art and were analyzed and published in color by Charles Ellis. So we have some color images of rugs in the Valentiner exhibition. I will supplement that with images I have found in other places that seem to be similar to other rug types included in this 1910 exhibition.
And I want to acknowledge before I begin that Valentiner and his peers have called attention to some very beautiful rugs. My objection to their thesis is that that they seem to suggest that there are not similiarly meritorious pieces woven after the 17th century.
I have turned most of the following images on their sides in order to use the shape of our monitors, but that is not the orientation in which most of these pieces are presented when published and you may wish to print off an occasional piece in order to see it in a more advantageous orientation.
Let's start looking at 15th century rugs with a Mamluk rug that was in this exhibition.
Despite the fact that the Mamluks have a clear link to Central Asian they have not usually been among my favorites. They have a narrow color palette and while their designs show great intricacy and discipline, they often seem simultaneously busy and vague to me.
But I think this is one of the most unusual and interesting Mamluks of which I know. I think the variations in the design and the use of cross-panels and cartouches is graphically very effective.
Here are some Anatolian examples from the centuries Valentiner admired. It is not always clear what century each of these pieces is assigned to, but they are all 17th century or earlier. (Ellis dated some of these rugs later than they had been previously seen to be. None of the redated rugs I will present here are seen by Ellis as later than 17th centiry. I will not tussle with whether given rugs are 15th or 17th century, but simply include them as examples within the claimed meritorious period.)
The first is a "Holbein" rug from the Williams collection that was in the Valentiner exhibition. Ellis sees it as 17th century.
It unquestionably has great graphics and color.
A second Analatolian piece (below) is not one from this exhibition but is instead what we now call a "Turkish village rug," this one from the Christopher Alexander collection and estimated to have been woven in Konya in the 15th century.
It has wonderful stark and powerful graphics and a blue that has retained its vividness in an impressive way. I think from their description of "Asian Minor" pieces from the 15th century that Valentiner and his expert peers would admit it to their group of worthy rugs.
The Anatolian piece below was not in this exhibition. It is from the McMullan collection and when published was described as an "extremely rare variant of the 'Lotto' group. (I have not included a number of the more frequent classical Anatolian designs that could be cited as examples: the Lottos, the momumental Usaks, etc. but have tried to give you for the most part more unusual examples from these three centuries.)
The colors are intense and I find the large scale of the devices in the wide borders interesting and effective.
Now let's look at two Persian examples. The first, below, does not seem to have been included in this exhibition but was part of the Williams collection donated the the Philadephia Museum.
Ellis calls it "one of the best known and highly respected examples of the classic era of Persian weaving in the sixteenth and seventeenth centuries. For me the spacious elaborateness of the wide framing boders, the subtle, detailed drawing of the field and the very skillful use of color make this a truly great rug.
The fragment of a classical Persian carpet below was not in this exhibition but is another that was part of the published McMullan collection. It is described as one of the 16th century designs of the highest level of sophistication, complexity and nervous energy" that have "survived."
Among it merits are the fact that "The basic expression of the field is secured by placing one series of forked arabesques over another. Either system in itself is a complete and satisfactory design. However, the blending of two such arabesque designs is an outstanding achievement." The description of the virtues of this piece goes on in considerable additional detail, but despite having to acknowledge the design virtuosity this fragment exhibits, I do not personally care for it. I find its colors unpleasant. But it is clearly within the group that Valentiner and his fellows recommend.
As a kind of aside, I am not sure that Valentiner and his fellow experts would admit a Kurdish rug to their recommended group but there were some that might meet their age criteria. The piece below is all I could scan of a detail of another rug in the McMullan collection. The description says that rugs with such designs were produced from the 17th through the 19th centuries.
This is the sort of Kurdish rug that demonstrates what Kurdish weavers are capable of with regard to adaptation of classical Persian designs and their very talented use of saturated colors.
The last rug I want to show you in this first series is a very dramatic example of a "Caucasian" dragon rug. Again from the McMullan, such rugs are now dated from 17th century forward.
Valentiner did include some "dragon" rugs in this exhibition, designating them "Armenian" and estimating them as early as 14th and 15th century.
Now I need to consult another book or two to give you some Mughal examples or some good Polonaise pieces, but I think you ge the idea. These are the sorts of rugs that Valentiner and his expert peers felt should be collected.
They are certainly objects of great beauty. My only dissent is that I think there are things produced in the 18th and 19th centuries that are of different, but equal aesthetic worth. This, speaking in 1910, they would not allow.
R. John Howe
One thing worth bearing in mind is that these guys were writing and collecting about 100 years ago. The 19th century stuff they saw included the mediocre, which was most likely the vast majority (98% of the rugs are not in the top 2% for their period). The older pieces probably had few of the original mediocre group - they would have nearly all been trashed.
Fast forward to today. Most of the mediocre 18th and 19th century stuff is gone, some of the very best of it remains. So, 18th and 19th century rugs look pretty good to us. Nearly all of the mediocre mid-to-late 20th century pieces are still around (if you don't think so, browse eBay for a little while). While some modern things are very good, indeed, the trash hasn't been discarded yet. For that reason, most contemporary (say, less than 75 years old) rugs aren't as good as most antique rugs are. But, wait 100 years or so, and the remaining late 20th century rugs will include a small proportion of trash and a large proportion of good stuff.
While I would not argue with the very general rule that "older is better" in terms of rugs, I think that we should consider a couple of issues.
First, it would depend on which weaving groups are being compared across time. Many of the rugs that we appreciate today are "tribal" Turkmen, Kurd, Baluch, S. Persian, etc. Rugs from these weaving groups have developed a following over the past several decades, in some cases rather recently. Perhaps it is because of availability, but it might also be due to changes in aesthetic sensibilities. In any case, based on my limited knowledge I have doubts about whether the authors were referring to 15th to 17th century examples from these weaving groups. Still, I think many of us still find that within these groups, older is often better, though we seldom are able to look earlier than 150-200 years ago.
Second, I still wonder if there is a possibility of "survival bias" in rug appreciation and collecting. Current practices suggest that many of us are participating is the selective preservation of better examples of rugs made in the past 75 to 150 years. For example, a pedestrian Jaf Kurd or S. Persian bagface with poor colour is less likely to be preserved on a wall than an exemplary version. Similarly, I expect that the relatively small percentage of rugs that survive 200-300 years are not a representative sample of all rugs woven in their time. Has anyone else noticed how often the versions of old rugs that are depicted in paintings (i.e. from the 15th to 17th century) often seem simple and uninteresting in their design and proportions (Gantzhorn is a good reference for this)? Perhaps the painter didn't spend as much time focusing on reproducing the rug or carpet, but if they are accurate then many of the painted examples look like cartoons compared to the real, preserved examples. Could it be that there were also less inspired examples of rugs and carpets woven during those eras as well?
Just some scattered thoughts from the "peanut gallery".
Do you think the visible range of 19th century rug production was significantly, or perhaps qualitatively, different in 1910 than it was in, say, 1975 (when some of us fossils were looking around)?
Unless we have some reason to believe that people are just as likely to preserve the worst as the they are to preserve the best, the average quality of the surviving pieces from any period has to improve with time.
So, I guess the short answer to your question is yes.
Dear folks -
Two of the types of rugs that Valentiner and his expert peers complain about collectors purchasing are the Ghiordies and Kula varieties.
Here to give you a concrete picture of what they are talking about are one of each of these types.
First, a Ghiordes niche format from the Smith collection in Springfield, MA.
I heard, when I first began to collect rugs, that there was a time when a rug collection was not considered "complete" unless one had at least one Ghiordes. The Smith catalog referred to above places two Ghiordes niche format pieces at the beginning of its Turkish section.
Ironically, this is one type about which the views of Valentiner, etc. have ultimately held sway. Ghiordes prayer rugs are not seen nowadays as particularly desirable, although there are lots of niche format pieces that are, for example, those in Kaffel's volume on Caucasian prayer rugs.
A second type objected to in 1910 was the Kula. Now in truth it may be that the particular Kulas being pointed out then might have been those that resembled the Ghiordes prayer rug or other niche formats with columns in their design. Here, though, is a 19th century Kula of a different sort from a Brausback catalog.
I personally don't find either of these pieces particuarly attractive.
R. John Howe
rugs come and rugs go
Of course, you're right. Still, I wonder whether the difference as perceived at that time was such, per se, as to compel a fundamental difference among the observers about what was worthy and what was not. I wonder whether Valentiner was familiar with, say, what we would consider the very best of extant Turkoman main carpets. I am thinking of something like the excellent Yomud kepse gul carpet in the collection of the Textile Museum, or other things of that standard. I have a feeling he was disposed at a basic level to disregard weavings of this kind, or perhaps one should say he was indisposed to accept them as meritorious. They would be OK to put on the floor.
As James mentioned, we are looking at changing aesthetic sensibilities. Valentiner seems to have been unable to take seriously certain classes of rugs his succesors in connoisseurship now covet. I was saying in another thread that latter day New England farmers were overlooking the artistic merit in a handmade pitchfork.
On a side note, three cheers for the Kurdish weavers of the world. John's example is my idea of a good looking rug.
"...[T]heir very talented use of saturated colors." I should say so!
I was a little hurt by his comments about the Kula, as I happen to have one. It's double or so the length, and pretty beat up, but it's old, I own it, and there you go. (i'd put some sort of smilie in here, but they don't seem to be available on this screen for some reason.)
I notice that your post indicates that you are a "Guest". That means that you weren't logged in, which is why you didn't get the smilies option. Your user name is Richard Larkin, so if you enter "Rich Larkin", it treats you as a stranger.
I'm sure changing taste is important. Somebody (maybe you, I don't recall) pointed out that Belouch rugs were uninteresting to collectors until about 30 or 40 years ago, and collector interest in Turkmen weavings, especially the bags and trappings, didn't enter the mainstream until after World War II.
The fact that Valentiner, et al., actually objected to collectors buying things of which they disapproved is an arrogance that some people never seem to outgrow. One moron has spent more six years trying to muzzle our insistence on discussing things of which he doesn't approve, and even opened a website that he's maintained for more than four years devoted largely to that end. I am mystified by the notion that one form of the collector neurosis is morally or intellectually superior to all others.
Thanks for the heads up. I've typed up two somewhat lengthy responses that have been lost in space. The system seems to be telling me I don't know my password, so I'm trying another one.
How frustrating. Not your fault. I won't recreate the pearls of my earlier responses that didn't make it. I explained ad nauseam what I had done about trying to post images by emailing them (I think) to you and Filiberto. One point was that the first try (to you) included very large images, then I learned (from John, I believe) that you preferred small. So I emailed (I think) smaller versions to Filiberto. Am I getting anywhere on this?
Sorry for the trouble, and to be cluttering up the main pages with this. I'm not even sure I'm emailing to the right places.
I didn't receive anything.
Click here to email me
I don't recall receiving anything from you in the past 24 hours, but my memory for such things is pretty bad.
The image files that we post seldom exceed 100 KB, and are seldom wider than about 550 pixels. When they come to me (or Filiberto) bigger than that, we resize them before putting them into the server. We prefer to receive them already appropriately sized, because that makes less work for us. Filiberto loves to work, but I'm incredibly lazy.
You can send e-mail to me at this address.
I can set your password to anything you like, and you can then change it through User CP (button on left side of screen, just below the introductory paragraph) to something I don't know.
Whew! I'm almost ready to yell, "Uncle!" How pathetic on my part. I'll try to resend from home shortly. Sorry for the trouble.
Filiberto: I'll try emailing to you again, too. I don't mean to make more work than necessary for you and Steve. Thank you for your patience.
I've lost things in cyber-space too, trying to do something long directly in the Turkotek post program. In fact it happened to me today.
What I'd recommend is if you have a long post, compose it first in Word and paste a copy into the Turkotek posting area. That way you always have the original.
It's very discouraging to lose something you've worked an hour on. I'm not even going to try to redo my lost post from earlier today. Fortunately, the conversation has moved on a bit and western civilization is not breathless over what I might have said.
And I have password problems too. Just ask Steve and Filiberto.
R. John Howe
Maybe not breathless, the whole of Western Civilization, but a good chunk of it is very interested. If you settle down and reconsider, let's hear it.
As for my password, I mixed them up, but I've regained my composure on that. Thanks for the encouragement. And the ever fertile postings.
Let's be fair
From Valentiner’s quotes that John posted I do not infer that Valentier regards all of 18th and 19th century rugs as inferior to earlier ones. He first states that the high point of rug manufacture was reached in the 15th-17th century. Then he says that collectors do not seem to recognize this fact, as they tend to prefer collecting rugs from later times. He does not write that later rugs are all bad. He writes that they “usually show a lack of clearness in design and a weak sense of color relations.
Then he goes on to argue that there is no design innovation in rugs during the 18th and 19th centuries. Instead, later rugs merely repeat already known designs. Valentier again uses words like “generally” or “seldom,” indicating that he does not dismiss all later rugs.
His most exclusive statements are that 18th and 19th century rugs “never stand comparison with those made in the earlier periods,” and that “Every old rug, on the contrary, has a marked individual character …” Here I would disagree. But as these statements are inconsistent with his other statements, I would discount them somewhat and not infer that Valentier dismisses all later rugs.
BTW, Valentiner does not use the word decadence to describe rugs, but to describe a general attitude of people. So, I think it is not fair to label this as a rug “swear word.”
Hi Tim -
Well, you're a scholar and capable of close distinction, but I think you're WAAAY too easy on these folks. For me, they are making the argument that only they can discern truly great aesthetics and that both the "educated public" and most collectors are to be pitied for not seeing what they see and as a result buying less worthy material.
It's still going on today, only the dates have changed. The arguments are identical and the confidence of aesthetic experts is breathtaking.
My root argument is that since it seems that there is no objective basis for any theory of aesthetics so far (the formalists claim we're "hard-wired" but we're obviously not; we tested it modestly once a bit, way back http://turkotek.com/salon_00011/salon.html ), there is no reason why any of us should treat seriously the aesthetic judgments of any of the rest of us. Too often, "it is good" is indistinguishable from "I like."
I think the reigning aesthetics in any field are socially constructed and therefore manipulable (and manipulated) by the dominant elites of the time.
R. John Howe
You are cutting the man plenty of slack by emphasizing his qualifiers in those statements in which he essentially dismisses 18th and 19th century rug production. Then, in those places where he comes in with the punch line, unqualified, you discount the statement as inconsistent with his other comments. Don't we have to concede that he has little or no regard for the later rugs, for whatever that's worth?
Ooh! I took a look at that "aesthetic" link. I didn't know about that. It looks like it's going to be fun.
Of course, I know that salon is ancient history.
I see that the Holbein you showed, just under the Mamluk, has a box-type border of serrated leaves surrounding a diamond design. Quite a while ago (decades in computer years, but probably only a couple of years ago) I suggested here that the ubiquituous leaf-and-wine-glass border found commonly in Caucasian and also often in SW Persian tribal rugs was possibly derived from a box of this type. If this box is sliced in half, then you have the leaf and wine glass.
Just to speculate even more, it would seem likely that an old rug with the box border may have lost several columns of knots along the selvages and several inches of length, producing a leaf and wine glass border.
Hi Pat -
I think the thrust of your observation about the "wine glass and leaf" border is correct.
In fact, Wendel Swan made a presentation at an ICOC conference a few years ago in which he held that many devices that we see or interpret as representational are in fact instances of a largely geometric Islamic design tradition. This particular border was one of his examples.
I don't think we have to resort to "knots wearing off" to suggest a source for the "wine glass and leaf" border. Even more likely, it seems to me, is that a weaver somewhere decided to use only half of this broader one.
R. John Howe
One Example of Design Change Over Time
Dear folks –
In 2002 Walter Denny, the UMass-Amherst rug scholar, curated an exhibition at The Textile Museum here in Washington, D.C. entitled “The Classical Tradition in Anatolian Carpets.” Denny also wrote and the TM published an associated full-color catalog with the same title.
Denny’s chief strategy in this exhibition was to take some key types of Anatolian carpets and to provide various interpretations of these designs over time. He used the “Holbein” varieties, the “star” and “bird” Ushaks and several others in this way.
One of the classical designs that Denny selected was the niche format carpets with “coupled columns.” He began with the one below:
This rug is among the most famous of its type. It came from the Ballard collection, has a silk foundation and the twist of its wool suggests that it was woven either in Istanbul or Bursa rather than in Cairo. It is dated to the second half of the 16th century.
Another example is the piece below thought to have been woven in western Anatolian in the mid-17th century.
Denny asks that we note the changes that the area above the niches have undergone (they have become a kind of “fish” design) and the fact that the curvilinear nature of the border in the fist piece has been entirely lost.
A third example is from the TM’s own collection and is said to have been woven in central Anatolian in 18th century.
Denny points to the fact that the “flowers of the parapet have been doubled, the spandrels exhibit a regularized border design in their lower parts and the columns are very simple stripes with identical capitals and bases. But he also says that the use of color in this third example give it a “presence and liveliness lacking in many early examples” of the same design.
Denny’s next example, below is coupled-column torah curtain woven Ghiordies, western Anatolia in the 19th century. (Denny has a thesis that this coupled-column design likely originated in Spain in the 15th century because the architectural shape of the arches in the first example occurred only in Spain at that time. He feels it likely that Spain’s Jewish weavers, driven out of Spain by the Inquistion, but welcomed by the Ottomans, brought this design with them often in their torah curtains.)
Denny said that this is the most lavishly inscribed torah curtain of this type that has survived. He notes that the fine Ghiordes weave has permitted the design here to echo and retrieve some aspects of the first rug (the border, for example, is closer again to the curvilinear) but that “the bases of the columns now float weightlessly in the field.”
Denny’s next example, below, is an Ushak rug from the 19th century.
He describes it as a coarsely woven, simpler, more color-dominated version of the original in which the area above the niche has become more prominent and the columns have become “an almost insignificant part of the design.”
Denny next presents a fragment from a coupled column saff.
Likely woven in Ushak in the 18th century, it exhibits large scale arabesques in its spandrels, niches with cusps and alternating red and blue fields. The plain columns have capitals but no bases.
Denny describes the piece below (19th century, northwestern Anatolia) as the “ultimate evolution of triple-arched and coupled-column ‘sajjadah.’” The parapets have been reflected vertically and the columns have disappeared entirely.
Denny says “The artistic result is as beautiful a village rug as one could imagine; a work of art that combines striking originality of design and color with a design tradition which can be traced back almost four centuries.”
A final piece, below, is described by Denny as completing the “circle,” since now the reflected parapets have been moved into the field of a niched design.
Denny describes this 19th century fragment as “free and spontaneous, if not exactly precise and well-planned…” Again color draws one to this piece.
So there are some versions over time of a basic classical design. What do you think? Does aesthetic excellence reside only in the first two examples in this sequence? Has the influence of Ghiordes been visibly pernicious? Does color use weigh with design complexity and articulation? Do the last four examples simply not weigh with the first one? I leave it to you.
I will say that this was a fairly large exhibition with several other design types included. Denny volunteered during his walk-through that if he could take only one piece home from this exhibition it would be this one.
Valentiner and his expert peers would cry.
R. John Howe
The examples you posted show not only the evolution (or let’s call it development) of a classical design over time, but also its transfer from an urban workshop to an increasingly rustic milieu.
I bet that Valentiner would have considered the 19th specimens above as decadent and degenerated. I find them charming.
I think the notion that designs of the two village rugs are evolutionary descendents of the older ones is an interesting speculation, but that it isn't supported by very compelling evidence. An awful lot of what appear to me to be ad hoc assumptions are involved in the argument.
Hi Steve -
I have stepped out of one argument into another momentarily to show how Denny presented different versions of what he sees as a particular type Turkish classical design.
Although, he does make at least one seeming evolutionary reference in his description of the next to the last example, for the most part his discussion is taxanomic.
I am with you that we are not usually in position to say really whether a given design is based on another. What we can, I think, say is that particular designs resemble one another and then point out the differences as well.
Gayle Garrett, a local rug specialist who presents sometimes at the TM, entirely objects to any evolutionary or developmental language when talkng about rug designs, saying that we should talk only about differences. She also rejects any evaluative language in talking about these. I don't go quite that far.
Her argument sometimes looks like that of an interested party because she is active in the DOBAG project, but it is likely still basically sound.
R. John Howe
I think what led me to believe Walter was saying that this is an evolutionary sequence was this,
Denny describes the piece below (19th century, northwestern Anatolia) as the “ultimate evolution of triple-arched and coupled-column ‘sajjadah.’”
This was reinforced in my hopelessly linear mind by
... described by Denny as completing the “circle,” ...
I share Gayle Garrett's misgivings about talking in terms of design evolution, at least about doing so without explicitly recognizing the speculative nature of the topic. This is not to deny that it provokes thought and is kind of fun. I am bothered by Denny's analysis mostly because it involves so much moving things around, adding elements and deleting elements to get from one point to another. If you do enough of that, you can get from any design or motif to any other design or motif, whether they are historically related or not.
You wrote that Gayle Garrett "...objects to any evolutionary or developmental language when talking about rug designs, saying that we should only talk about differences." I'm not sure I appreciate what that means. Could you elaborate? Also, she rejects evaluative language. I don't get that either.
How is her approach essentially different from Walter Denny's?
It's always a bit dangerous to make someone else's argument for them, but since I share some of her position, here goes.
A great deal of the literature on oriental rugs is taken up with describing either how the designs in older rugs change as they come forward in time or with arguments for the likely sources of particular more modern designs. You can find the former in almost any old rug book you pick up. Christine Close, provides an example of the latter in her analysis a few years ago of the likely sources of the central devices in the "eagle" Kazak. Much of this analysis is presented in the language of "evolution," of "development," and there are often comments alleging "design degeneration."
The problem with making evolutionary arguments is that we do not know the weaver's intent. We can cite similarities and point out seeming conventionalizations, but the real connections between two similar designs are not made in this way, excepting through inference.
So some have suggested that we drop the claim that we know much about evolutionary paths of rug design and restrict ourselves to what we can in fact do: taxonomy.
So that is the first part of what I see as Ms. Garrett's argument. Give up any suggestion that we know much about the "development" of designs because it is likely that we do not.
But she seems to go further. The old rug books often denigrated newer versions pointing out what has been lost in conventionalization. Ms. Garrett will admit to "differences" but not to language that suggests, for example, that more simplified versions of seemingly similar designs are lower on some aesthetic dimension than the older, more elaborated ones. She objects to any suggestion of either aesthetic "development" or "degeneration" in description of differences between rug designs. She will only acknowledge "differences."
I hope that is clearer...and accurate.
R. John Howe
Thanks, John. Very helpful, and very interesting.
I know Walter Denny has been making the arguments for design development in Anatolian carpets much along these lines for quite a few years now. Certainly, there is some connection between, for example, the first two examples in your post. But as you say, we only know that through inference.
Are you impressed with the observation that the weaving medium, operating as it does on vertical and horizontal lines, enforces a straight line geometry that weavers (collectively over time) tire of fighting? In other words, curvelinear designs are achieved by a special discipline, and angularity is ultimately self-fulfilling.
I can't say I've heard that particular thesis, but as you say it it seems plausible. I'm sure you have heard that some more restrictive techniques tend to produce particular sorts of designs.
I do know one collector who is impressed with the notion of being able to "circle" the basic rectilinear character of weaving. He is an actuary and so a pretty good math student and has said repeatedly that one thing that drew him to oriental rugs and that continues to attract him is the small "miracle" of producing curvilinear designs on a rectilinear grid.
It is not an accident that he collects only silk rugs, since their high knot count makes it possible to draw quite "smooth" curves, even circles, on such a matrix.
R. John Howe
Circles and smooth curves are almost always the result of weaving from a cartoon. That's why you hardly ever see them on rustic or tribal weavings. Your friend's silk rugs, like most silk rugs, are probably urban workshop products made from cartoons that show the placement of every knot. Don't tell him this - let him enjoy the illusion that the weavers were performing miracles that he now holds in his hands.
Oh, he's quite aware. He's not naive about what he is admiring. He's an astute enough mathematician to see clearly how the "miracle" is achieved.
And he is under no illusion that he is buying pieces woven from memory. The "perfection" of the silk rugs is another thing he admires, so cartoon-based designs do not diminish his enjoyment.
And he is not put off by a rug being new either. There are some very nice silk rugs being woven these days. The Chinese are sometimes doing so well that some Turkish weavers are buying Chinese rugs, weaving a "Hereke" designition into them and trying to sell them as Herekes.
Some experienced collectors don't approve of his collecting, but he's steadfast in what he is about.
R. John Howe
Thanks John for posting that great sequence of rugs. Walter Denny also wrote
the little Smithsonian sponsored book on oriental rugs- they must have printed
lots of them as they are available for $5.00 at the usual on line book sellers-
I include a copy along with every weaving I give to somebody.
Oh, I wasn't expressing disapproval. In fact, I own a silk Hereke prayer rug, circa 1960, that I like very much. I just had the impression that he thought the weavers who could make circles must have been technical masters.
Hi Doug -
Yes, Denny's little Smithsonian "chestnut" you mention is an ideal "give-away" book for someone new.
Another is Preben Liebetrau's "Oriental Rugs in Color," which not only has a handy size and lays out the typology that Jon Thompson popularized in his own basic book, it has remarkable color for a book first published in the U.S. in 1963.
And for some reason you often find the dust jackets still intact on the Liebetrau volume.
R. John Howe
I wasn't aware of Gayle Garrett's argument, but I like it.
I've gotten to much the same place in a slightly different manner. Evolution (or devolution as is sometimes claimed) requires knowledge of what came before. So if an atelier produced rugs over a period of time and made V1.0, V1.1, V1.2, V2.0, etc. of a similar design, evolutionary tendencies could be confirmed. But that's not the way it worked. We're sitting here hundreds of years and thousands of miles away comparing one rug from one place with another probably made somewhere and somewhen else - and then stringing them together in a design evolution.
In science this is called "dry labbing".
That Liebetrau book was the first one I got when I initially contracted the disease. I had the impression back then it probably covered about 60% of known production. Turned out I was mistaken.
P. S.: The dust jacket's still on it. Must be a good grade of paper.
Thanks for posting the coupled columns niche format rug sequence. Very interesting, indeed. Among the first five rugs the choice is obvious (based on just the digital images). The first piece is truly outstanding, and I am surprised by Denny's decision to go for the third one. But he presumably saw all pieces in the wool, and we have only the digital images.
The comparisons with the last two village rugs are more difficult. The difficulty is to disentangle quality from personal taste. I might go for the last one, because of my personal preference, but I'd still consider the first rug as the most accomplished one.
On the matter of weaving curved lines from cartoons. Some rugs we consider tribal, such as certain Qashgais, show some fairly sophisticated (and well drawn) designs. There are some in a pretty elegant version of the herati pattern, even some of the boteh we were recently looking at on one of these threads. Would you think they are drawn from cartoons? Or are you referring only to the more elaborate designs modeled after Persian and classical carpets, etc.?
I wasn't thinking of Qashqa'i when I wrote, and would call them exceptions to the rule. I left a little wiggle room for exceptions in my text. Note the hedging in the boldface type (added for emphasis here):
Circles and smooth curves are almost always the result of weaving from a cartoon. That's why you hardly ever see them on rustic or tribal weavings.
My momma didn't raise no dumb boys.
Richard, Steve -
It's remarkable how well some weavers can draw credible circles on pretty coarse weavings. I've seen some Ersari examples that can't have been much beyond 60 KPSI. And some of those might well have been woven from memory.
And one of the real technical achievements of the Indians who "wove" the Chilkat dancing blankets is that they learned how to combine species of braiding with weft twining to make circles on flatwoven fabrics done on a warp-weighted loom.
So there's a lot a weaver ingenuity that's been applied over the years to circle drawing.
(By the way, I put the term "wove" in quotes above because folks like Marla Mallett do not see weft twining as a species of "weaving." It does not meet their standard for "interlacing" of the warps and wefts, since no shed is involved. Just an aside.)
R. John Howe
Weavers of Chinese rugs were masters of making curved lines and circles with <30 kpsi. Look at some pictures of "wheels", lotus blossums, or clouds in old Chinese rugs.
Whether one calls it evolution, devolution, degeneration, transition, development or Garrett’s meaningless “differences,” various processes of copying are at the heart of weaving and design proliferation. Any pattern that is repeatedly copied is going to vary, change, devolve or evolve (you pick the verb).
Walter Denny’s sequence illustrates that one long-term result of copying is that certain elements are extracted and become independent of their earlier contexts.
The following rug could be inserted after the saph to show another possible step in Walter’s progression. In this case, you must view the ivory as the columns, not merely as the ground for what some will want to see as green cypress trees. A few inches are missing across the center.
All of the rugs Walter shows are just a few of the dozens or hundreds of similar rugs. And there must have woven dozens or hundreds more in various transitions. I don’t believe that it is speculation to observe such a continuum.
John, you wrote: “The problem with making evolutionary arguments is that we do not know the weaver's intent.” I think we do. The weaver(s) intended to copy all or part of some other rug or textile or a cartoon. Period. Variations may occur either intentionally or unintentionally. It doesn’t really matter which. In any event, we end up with a lot of variants that look similar and then another batch of further variants that look similar to one another.
In my view, here on Turkotek undue emphasis is often placed upon the work, status or “intention” of an individual weaver. We don’t recognize a Salor or Qashqai or Kurdish rug because of what one weaver may have done or what her intentions were. We can recognize and categorize them because of the enormous quantity of similar weavings produced by a community or culture.
In the past few years I’ve come to believe that most collectors and even authors view rugs myopically. I cannot pretend to have the answers, but all would benefit by trying to look at rugs in a much broader historical, artistic and cultural context.
Walter’s quite conservative approach with this limited number of pieces is precisely the kind of analysis from which all can learn. I’m surprised at how quickly it was abandoned as speculation and as not demonstrating any more than “rugs are different.”
Dear folks -
This has the makings of an interesting and potentially useful discussion.
Jerry Silverman says that Gayle Garrett's argument make sense to him and Wendel Swan, an old Chicago area neighbor, in years past, says that Garrett's argument is "meaningless."
Can we hear more from both sides?
Without diffusing any of what might be said, Wendel let me ask why you feel that the noting of taxonomic similarities and differences would be "meaningless?" A great deal of biology is taxonomy.
R. John Howe
I think I was the primary objector to Denny's presenting the thing as an evolutionary sequence. I have little difficulty with the first several examples, in which the changes from one form to the next are fairly simple and straightforward.
The going gets rough, I think when he gets to the village pieces. To get the first of these into the sequence, he says,
The parapets have been reflected vertically and the columns have disappeared entirely.
That's a lot of changes.
To get the next one in,
... the reflected parapets have been moved into the field of a niched design.
The reason I object is that any form can be morphed into any other form if you make enough appropriate modifications to it. With patience, the image of any rug can be converted into a picture of an elephant eating a head of cabbage in a (long) series of single steps. So, I get wary when someone argues that it is an evolutionary sequence if several steps, especially if some are really multiple operations presented as single steps, are needed to accomplish it. Not that it's impossible for the sequence to be correct, only that it's impossible to create compelling evidence that it is.
It may be that I do not understand the respective arguments adequately, but the two positions do not seem mutually exclusive. Rather, the two sides appear to be carrying on the tradition of the Miller Lite campaign: One side is shouting "Tastes great," the other answering, "Less filling!" Both can be true.
I refer to the first sentence of Wendel Swan's comment above and suggest it is a truism. With that thought in mind, I offer the following additional comments that I think are self evident and do not require objective proof:
-Some copying must attempt exactness. This endeavor puts the weaver in the position, in effect, of using other weavings as the cartoons Steve was alluding to above.
-In some cases, the copying must constitute less of an effort in precision; and the nature of the weaving medium being what it is, viz., a collection of horizontal and vertical elements, the weavers (and their designs), over time, surrender to angularity and give up detail.
-In some cases, weavers must size up what they are doing and make conscious decisions about their copying, frequently by way of simplification. (It has often occurred to me while reading commentators' assessments of what the weavers must have been thinking, that whatever they were thinking, they probably understood weaving much better than most of us ever will.)
I don't doubt that all of the above dynamics have been in play throughout the history of the weaving of pile rugs. The results are all around us. As an example of the third suggestion I made above, I offer two images of a design probably familiar to most of us.
I refer to the border design, which is familiarly called the "boat border." One example, articulated with an abundance of detail, is from a Yomud rug. The other is a Baluch. I will not attempt to prove who is the ultimate copyist, but most will assume that the Baluchi weavers are famous (or notorious, depending on the commentator) copyists of the Turkoman. It is not possible to say just what the connection is between these two examples. However, I suggest that somewhere in the interplay between the peoples in these weaving regions, the border manifest in the Yomud example was known to the Baluch weavers in pretty much its full form. The Yomud is probably somewhere in the nineteenth century, give or take. I would estimate the Baluch to have been produced sometime early in the twentieth century. I don't think these statements are controversial.
My point about these two rugs is that I think it is quite likely that the simplifications of the boat border design were implemented consciously by the weaver. Certain elements were preserved or suggestions of them were adopted. The result is “clean.” It seems hardly probable that the more complex elements, present in the Yomud version, fell out of the Baluch version over time, the way (I'm told) we lost our tails. (I apologize in advance for any offense taken over that remark.) If I am approximately correct in this, I suggest that the dynamic has repeated itself countless times. When we look back over the motley collection of products that have survived, we call it evolution. It's a reasonable term as long as you don't expect too much from it.
Evolution, simplification, degeneration - all these terms are being used to
imply a chronological progression from the old to the newer, from the more
complex to the simpler, from the higher "energy level" to the lower.
Call it design "entropy", if you will. It is a lens through which we see and interpret what we see.
But while entropy is a physical law, there is no reason why it has to be a design law. I have to look no further than my floor to see an Afghan rug made in 1980 that has borrowed many motifs and made them more elaborate.
I'm not saying design entropy doesn't take place. I'm just suggesting that it needn't necessarily take place.
I would like to add a comment referring to the rugs presented by Walter Denny, courtesy of John. I agree with Steve Price that the leap to the village rugs, and perhaps the one posted by Wendel Swan, is harder to accept if the proposition is anything like a direct link. The movement from example one to example two, on the other hand, is virtually incontrovertible I would think.
I do not necessarily reject the link in the later pieces, and am loath to second guess someone like Denny, a professional scholar and art historian, on something like this. It's just that the link isn't so obvious to me within the scope of the examples shown.
First, I want to emphasize that I only said that one long-term result of copying is that certain elements are extracted.
Second, I did not assert any direct link between the small rug and any others. It is virtually impossible to link any one rug with any other rug(s) and thereby prove an “evolutionary” process. My point was that we must view as much of an entire corpus of weaving as is possible. When we do, we can sometimes see similar examples of many links, not the link.
Similarly, Walter Denny was probably not trying to establish his examples as direct links in an evolutionary process. Having spent some time looking at architectural connections to weaving, I wanted to provide one example among many of one phase of the process.
In the case of the first two Denny examples, I doubt that we are viewing evolution as such. Both are undoubtedly the product of highly supervised weaving, the designs having been perfectly developed and articulated off the loom before weaving began. Both reflect a timeless interest in architectural elements as motifs, as do countless others. If they represent any type of “evolution” at all, they are variations by virtue of having been woven in different places and at different times.
One can readily see the same type of variation in the 2-1-2 design of early Turkish rugs, the small pattern Holbeins, and the later Karachov and Kagizman versions.
Memling guls come to mind as another ubiquitous pattern that is probably more than 1,000 years old. They have been interpreted by virtually every weaving society.
John, you wrote: “Garrett … entirely objects to any evolutionary or developmental language when talking about rug designs, saying that we should talk only about differences.” (Emphasis supplied.) If she is saying that we can’t assign aesthetic values to changes, I can at least understand the argument. But if she contends that we must confine ourselves to merely saying that rugs are different and that there is no evolutionary process (a neutral term), I’ll repeat that I think merely reciting differences is meaningless.
In what I think is an effort to dispute the extraction and simplification theory, Jerry has shown “an Afghan rug made in 1980 that has borrowed many motifs and made them more elaborate.” Clearly, several motifs were borrowed, but they look more jumbled than elaborated. Sorry. While cross-cultural borrowing is ancient and even interesting, this is an unfortunate example of what happens when a group manufactures a rug with no connection to an earlier example and its culture.
"elaborated" - "jumbled"
"tomayto" - "tomahto"