Posted by Wendel Swan on 01-13-2002 09:48 PM:

The accuracy of dates

Dear all,

Some of my long held beliefs about Caucasian rugs from the mid-19th Century recently came into question. Below is a rug belonging to the late Richard Markarian that appeared on the cover of "Oriental Rugs in Cincinnati Collections," the catalog of an exhibition at the Cincinatti Museum of Art in December of 1976.

When I saw this 3'3" x 4'9" beauty, I made a few calls to see if it might possibly be for sale. It wasn't, but it remains one of my very favorite rugs.

The inwoven date of 1857 seemed plausible and may accurately the year of weaving, but two recent offerings on eBay have caused me to question the accuracy of the date and the accuracy of other dates on Caucasian rugs.

The following rug at first presented a temptation for me to try to get what I couldn't have so many years ago.

However, this rug has some synthetic dyes (why are they always called "early" synthetics by the sellers?) that ultimately stymied my brief interest in it. While it obviously wasn't woven by the same masterful hand as had done the Markarian rug, the style is remarkably close. Absent the date and/or the synthetic dyes, one would be hard pressed to make an age distinction between the two. They are, in my view, virtual contemporaries.

This rug on eBay is clearly less desirable, in large part due to its horrendous synthetic dyes. It bears two dates: the one on the right is 1837 and that on the left is either 1867 or possibly 1817 or even 1827. None of these alternatives could have been the year of manufacture.

We cannot assume that the Markarian rug is the oldest simply because we find it to be the most beautiful. Are we justified in relying on its date? The format and elements of the "1837" rug are quite close to those of the Markarian example, just not as well done. What is the purpose of these dates? To commemorate someone?s life or marriage? To mark an anniversary of some sort?

The only safe assumption is that the rug was not woven before the inscribed date.

Incidentally, the medallion on these rugs is commonly seen on the so-called Fachralo pieces. It must have been a popular device.


Posted by Steve Price on 01-14-2002 09:10 AM:

Hi Wendel,

A few years ago I collected a little database of inscribed Caucasian rugs and the dyes in them. Specifically, I noted the presence of corroded black/brown and of a tip-faded fugitive purple - let's called it fuchsine for convenience.

Here's what I found, with 31 rugs.
1. Every rug dated 1907 or earlier (21 out of 21) had corroded black or brown. Of the other 10 rugs, dated from 1913 to 1957, only 2 had corroded black or brown.
2. Of the 12 dated 1875 or earlier, not a single one had "fuchsine".
3. Of 7 rugs dated 1877-1901, 2 had "fuchsine".
4. Of 12 rugs dated 1904-1957, 9 had "fuchsine".

As you can see, this (admittedly small) body of data suggests that the inscribed dates are generally consistent with the dating we'd make on the basis of the dyes.

They are not a perfectly reliable basis for date attribution. But, then, neither is anything else. I think that the inscribed date is pretty close to the date of weaving most of the time, and ought to be given considerable weight in making an attribution. In fact, the only times I'd summarily dismiss them as being accurate is when it is clearly impossible for the weaving to have been done at that date. For instance, a fugitive synthetic dye cannot possibly have been used prior to 1858 and is unlikely to have been used often before about 1880 just because of the lag between first laboratory synthesis and wide commercial availability.

That is, even though inscribed dates are not always accurate, I believe them to be the most accurate age indicators we have, and for any individual inscribed rug, I take the date inscribed as accurate by default in the absence of clear evidence to the contrary.


Steve Price

Posted by Michael Wendorf on 01-14-2002 11:38 AM:

color quality

Dear Wendel:

The synthethic/non-synthethic issue aside; the color
qualities of the first rug compared to the others seems to
be completely different based on the colors on my monitor.
The green is softer, the red mellower. The Markarian rug
looks a lot more beautiful to my eyes in its coloration.
The other rugs, I do not imediately see synthetics on the
second rug, seem more monochromatic to me by comparison even
if they have the same number of colors. It a distinction of
color quality.

You may recall that a related rug was auctioned at Wechslers
several years ago. It also did not have the same color quality.
I saw another in 1993 at the ICOC in Hamburg. I still remember
discussing the piece with Jim Burns. Burns was of the opinion that
these pieces were made over a period of 100 years or more and that
most were made later, including the piece then at hand.

I think the inwoven dates should always be viewed with skepticism. However, in some cases they could well be real.
I doubt it should be seen as either much of a criterion for collecting nor an anchor for dating a piece. Color and weave are much more diagnostic.

Thanks, Michael

Posted by Rudolf Hilbert on 01-14-2002 01:07 PM:

Dear all,

inwoven dates seem to be encountered primarily in later 19 ct. Caucasian rugs.
As to the validity of these dates: I suppose many of us have seen rugs with seemingly too early dates.

But what about the opposite:
Has anyone seen a Caucasian or other rug with a date that is obviously too late ?



Posted by Steve Price on 01-14-2002 01:39 PM:

Hi Michael,

Whether inscription of a date (or anything else!) ought to be a criterion for collecting is a topic I won't go near. It's kind of like making a judgment on the desirability of various neuroses.

As for the usefulness of inscribed dates as age attribution criteria, I understand exactly what you're saying about dyes and weave. But the old bugaboo, epistemology, keeps creeping into my thinking. How can we be sure that the qualities we associate with rugs made at particular times are really accurate? What are the anchors? In short, what is the foundation on which the edifice rests? Is it really more solid than inscribed dating?


Steve Price

Posted by Michael Wendorf on 01-15-2002 12:04 AM:

to Steve

Hi Steve:

I think the color and weave foundations are more solid than inscribed dates when based on significant experience and some knowledge of what was being woven when. But I agree that the anchors are relatively few and not particularly well sized relative to the drag.

For example, in the Caucasus we know what Kustar production looked like, when it started and how it progressed. It was a race to the bottom with the colors increasingly red, white and blue. The subtle or painterly use of numerous shades of red or other colors becomes extinct. At some point it seems certain colors just seem to fall off the palette. We also know when some synthetic colors became available, at least within a decade or so. I think these are all interrelated. In general, we also know enough about color to be able to talk about traditional Azerbaijani colors or Caucasian colors. Ultimately, the real anchors we have are those few carpets depicted in paintings, photos or documented in other ways - inventory or acquisition records, for example.

Thanks, Michael

Posted by Sophia Gates on 01-15-2002 11:33 AM:

The Medallion - A Star?

My question has to do with the bold central medallion: could it have evolved or been meant to represent, a star? Maybe this design has a long history and even the Makarian piece, beautiful though it is, might be a late example of its evolution? Overall the rug does somewhat remind me of the "Star Kazaks" in its boldness and power.

I wish we could see these rugs in person! It's frustrating to try and judge their relative quality or age from jpgs! Wish we could feel them, smell them, taste them! Rub them against our cheeks! Oi - now I'm starting to ACT like my cats...but I think they have a point. . .


On dating: I think corrosion and presence/absence of dyes, as Steve has pointed out, is certainly a logical way to corroborate or challenge an inscribed date. However many of us have probably seen the Heriz-type example in Thompson's book "Carpet Magic", which seems to have a really early date; I've heard of dealers removing one or two little knots to change a date by a century or more. I think I would be a little cautious, yes? On the other hand if a piece has all Natural dyes and "feels" old - well, then what?

Among Caucasians, some of the most interesting dated pieces I've seen were woven to commemorate some Czarist event or other - I used to really hate them but recently their folkloric drawing and brilliant color has been making an impression on me - some of them have been featured in past Halis - I think they're probably an acquired taste. But - one can't leave rugs like that, figurative pieces, completely out of a discussion of Caucasian design. Who knows, they might represent one of the really characteristic aspects of Caucasian art - they remind me of those Russian dolls.

Any thoughts about the "star"? The "folkloric" rugs?

Posted by Steve Price on 01-15-2002 06:41 PM:

Hi Again Michael,

The accuracy with which we can make date attributions from properties like color and weave cannot be any better than the information relating weave and color to particular periods. That, I submit, is not very good at all - usually just marketplace folklore handed down from one person to another and getting credence from sheer repetition.

As for the real anchors being
those few carpets depicted in paintings, photos or documented in other ways - inventory or acquisition records, for example, this would be really useful if there were more of them and, especially, if they covered a wider range of textiles. The earliest documented Turkmen pieces were collected in the late 19th century, and the Rickmers collection, applauded as the best documented early collection of Turkmen stuff, was collected between 1902 and 1906. We have no Turkmen anchors at all with known dates before about 1890! Not only that, but there are probably no Belouch or NW Persian (or Caucasian tribal, if we are to make the distinction between them and NW Persian) that can be documented to have been collected as early as 1925. So, this is a real shortcoming.

But wait, there's more. The rugs in old paintings tell us that a particular pattern already existed at the time the painting was done, but can't show that a specific rug with the same design was made during that period. And, of course, the paintings give us no structural information at all, only design.

Shortcomings and all, rugs with inscribed dates are usually the best anchors we've got. Not perfect, but much better than nothing.


Steve Price

Posted by Marvin Amstey on 01-16-2002 10:02 AM:

Good morning, all. Have any of you seen and handled the Markarian rug? Are all of its dyes "OK"? Perhaps it is a contempory of the second two rugs.

Posted by Steve Price on 01-16-2002 10:47 AM:

Hi All,

I noticed that in an earlier post Rudi asked if anyone has seen a rug with an inscribed date that's obviously too recent. I saw one that is at least close to meeting that description. It was a Fachralo Kazak at a Sotheby's sale a few years back. Gorgeous piece, in great condition with terrific colors. And an inscribed date in the 1930's. My recollection is that it didn't sell, and I thought that in the absence of the inscription it might have gone into the $20,000 range. But, who knows?


Steve Price

Posted by Deschuyteneer daniel on 01-16-2002 10:59 AM:

Dear all,

I am thinking the inwoven dates in THIS group of rug has something to do with the commemoration of one event.

They look so much similar : 1817, 1827, 1857 ... that I can't find any other reason.


Posted by Wendel Swan on 01-16-2002 11:29 AM:

Dear Marvin,

I haven't had the pleasure of handling the Markarian rug. But the green medallion, the ivory field, the wonderful botehs, the crisp drawing, the use of and contrast in color all make nearly an ideal rug, in my opinion.

The fact that a master weaver may have woven it doesn't necessarily make it older than the other pieces. The close comparison with them made me question whether the Markarian piece was actually woven in 1857, as many of us may have assumed.

Donald Wilber's monograph on dated rugs (which I don't have) demonstrates, I believe, that the dates on rugs are not evenly distributed, as we would expect them to be if they merely denoted the year of weaving. The rug world knows very little of the reasons for dating rugs, whether Caucasian or other. Our western perspective may be that it is to mark the year of weaving, but these anomalies may give us reason to question that assumption.

My further impression is that many Caucasian rugs have dates using the numerals 1, 2, 3 and 0 much more often than any other numbers. I can understand how the first two digits would be 1 and either 2 or 3, but the last two should demonstrate more even distribution.

Does any reader have Wilber's work and could someone share conclusions to be drawn from it?


Posted by Wendel Swan on 01-16-2002 12:04 PM:

Dear Rudi and Steve,

On the topic of late dates, I remember seeing a Caucasian kilim at a local dealer's shop a few years ago with an inwoven date of something like 1945. It may also have had an inscription in Hebrew as well, but there was something else besides the inwoven date to suggest age and origin.

Absent that date and the inscription, I would have placed it with all those other circa 1900 kilims we see. One might conclude that a lot of the Caucasian material is much later than we think.

The concept of prospective dating doesn't make any sense to me.


Posted by Michael Wendorf on 01-16-2002 02:58 PM:

more late dates


I recall Saul Barodofsky once showing me a large Shirvan kilim he
had acquired in Turkey and asking me how old I thought it was.
Though it was a little dark with a great deal of deep blue, the colors
and drawing all suggested a conventional turn of the century date at the latest. He then showed me a large Armenian inscription woven into
the kilim that clearly stated (if memory serves) the kilim had been woven in 1922 to mark the occasion of someone's wedding. Yes, many of these pieces could be a lot younger than we imagine. Yet, there is
other evidence in photos and inventory records that we know some of these rugs and kilims do date to the 19th century at least.

I still would not trust an inwoven date for the most part.

Best, Michael

Posted by Rudolf Hilbert on 01-16-2002 05:06 PM:

Dear Wendel, Michael and all,

"one might conclude that a lot of the Caucasian material is much later than we think".


If you take a look at old auction catalogs and dealers catalogs from the 1970s until at least mid 80s you may come to the conclusion that many of the Caucasian rugs were dated notoriously at least 50 years too early, according to our present knowledge. Nonstandard (e.g. whiteground) Cauc. rugs were almost invariable put into 18th ct. by auctioneers and dealers.

Now we are much wiser and more conservative in our dating guesses - mainly because in the last decades a huge amount of Caucasian material has shown up in the trade.

Presently - for a somewhat attractive looking Cauc. rug - last quarter of 19th ct. is a fairly safe and generally accepted guess.

Personally I think that in the future dating guesses will continue to tend into the direction of younger dates, but not so drastically as you suggest. Perhaps 20 - 30 years at most.
Simply because we cannot expect much more fresh material to show up and the other sources for dating are not very reliable either.

Perhaps some technical invention in the future will allow a more or less exact dating of rugs < 200 years old.

What do you think ?



Posted by Patrick Weiler on 01-17-2002 12:27 AM:


I think it is the hope of many rug buyers that an accurate dating method can be found. It is the hope of some rug sellers that such a method never be found. Interestingly, the premium paid to top dealers exists, in part, because their familiarity and expertise in dating and identifying rugs is a valuable professional ability.

Patrick Weiler

Posted by Marvin Amstey on 01-17-2002 08:36 AM:

Good morning, Patrick. "Professional ability" or a top drawer con act? My vote is for the latter. The comments made in this thread, IMHO, speak to the most likely conclusion, that at least Caucasian rugs are younger than we think, and the dates are not trustworthy. As for scientific dating, we discusssed that at length last year, but in reply to your buyer/seller differentiation: among most collectors (we're not talking about the one time decorative purchaser) a buyer frequently becomes a seller, if only to better or change his/her own collection or get out of an obvious mistake. We've all done that, haven't we?
Best regards,