Posted by Filiberto Boncompagni on 02-06-2002 10:02 AM:

Some knotty questions.

Dear Sophia,

In your interesting Salon you raise, among others, the question of Turkmen knowing or not the pile-knotting technique. I hope you do not mind if I make a few speculations on the subject.

As we all know, there is a wide school of thought indicating the origin of knotted carpets among the Turkmen tribes of central Asia. As Elena Tzareva does in her conference (introduced in another thread), as a proof they propose similarities between Turkmen designs and prehistoric artifact. Or, looking at design similarities between Turkmen rugs and Anatolian, Caucasian, Kurdish and nomadic Persian rugs, they suggest that Turkmen originated such patterns
ergo invented the technique.
To be more specific: the Turkic populations propagated the art of knotted carpets in Anatolia, Caucasus and among Persian tribes.

The early example of knotted rug is the Pazyryk carpet - I will not bore you to death discussing it since we all know it. I want just emphasize the fact that although its origin is not clear (some say it is Persian, others say it is Caucasian) it seems to be a general agreement on the fact that it is a workshop, rather than nomadic, product, and it was not woven by the people who made the burial (people of which we know very little).
The knots are symmetrical.
In a site near the Pazyryk tombs archeologists found older fragments of asymmetrical knotted pile, so let us assume that both techniques are very old and we do not know which one comes first...

Now a little history: According to Opie (Tribal Rugs, page 193) Turkic tribes arrived in southern Persia in stages (from the Caucasus and northwestern Iran rather than directly from central Asia) during the first millenium AD.

Then, in the early 11th century, a Turkic tribe - the Seljuks - established in the Iranian province of Khorâsân and from there started to conquer most of Iran, Iraq, Syria, Palestine, Armenia and Anatolia.

What did they find as rug weaving art?

In Persia there was for sure a well-implemented workshop tradition: Muslim historian al-Tabari mentions the famous KHOSROW CARPET, possibly the most costly and magnificent of all time, made in the sixth century AD.
The subsequent Arab conquest had a deep impact on all the art forms, though.
The prohibition in the depiction of living beings forced the artists to gradually abandon realism and to develop the decorative aspect of fine arts.

Under the Samanid Dynasty, (AD 819-999), first native dynasty to arise in Iran after the Muslim Arab conquest, the rigid rule (about the depiction of living beings) started to soften. The potters for example employed early stylized Sasanian motifs such as horsemen, birds, lions, and bulls' heads, as well as Arabic calligraphic design.

The only early depictions I know of possibly Persian rugs are, unfortunately, of a later period: on the Shah Namah illuminations (XIV- XV cent.) visible also on Gantzhornís book (pl. 225-227). The first shows a Kufic main border, a strange secondary border (which Iím planning to discuss later) and a field with geometrical lattice motifs. No need to say that, for Gantzhorn, those rugs are Armenian.

In Anatolia IT SEEMS that Greeks, Armenians and Kurds already wove piled rugs from immemorial times. Now see Prof. M. R. Izady lecture given at Harvard University in 1993:

Following the Manzikert (the battle of Manzikert in AD 1071), the Turkic nomads gradually imparted their language to all the millions of civilized, sophisticated Anatolians whom they converted from Christianity to their own religion of Hanafi Sunni Islam. Almost everyone in Anatolia gradually assumed a new Turkish identity when converted to Islam. But, this did not mean that the old cultural, human and genetic legacy ceased to exist. On the contrary, the rich and ancient Anatolian cultures and peoples continued their existence under the new Turkish identity, albeit, with the addition of some genetic and cultural material brought over by the nomads.

In the same lecture he also states that the Kurds had learned how to weave rugs from the Mittanis (from 2000 BC):
The Mittanis settled inside Kurdistan around modern Diyarbakir, and influenced the natives in several fields worthy of note, in particular the introduction of knotted rug weaving. Even rug designs introduced by the Mittanis and recognized by the replication in the Assyrian floor carvings, remain the hallmark of the Kurdish rugs and kelims. The modern mina khâni and chwar such styles are basically the same today as those the Assyrians copied and depicted nearly 3000 years ago.

Now, to be clear, I do not know how he found out that and I do not subscribe his affirmation - just found it interesting to quote. He could be right.

Anyway, the Turkic tribes settled down in Anatolia and later founded the Ottoman empire. In the mean time they wove pile rugs, together with Armenians, Kurds, Greeks, you name themÖusing the symmetric knot also better known as the Turkish Knot.
Did the Seljuks influence the Persian city rug production? I donít think so. But they probably imported in Anatolian workshops the highly geometric decorative design born in Persia from Islamic restrictions as well as the use of calligraphic (Kufic) decorations. Probably the Armenians were more influenced than the Kurds, but I think they both retained (especially in rustic production) their own style and motifs that were probably already geometrical because they are born from flat weaving technique.

Now the knotty questions:

Why Turkic people of Anatolia, Caucasus and Iran use the symmetrical knot while the main Turkmen of central Asia use almost exclusively (with few exceptions for Yomuts) the asymmetrical, Persian knot? If the Turkic taught to Anatolians that technique, Anatolians should use the Persian knot, right?

It is possible, instead, that until the eleventh century Turkmen did not know how to weave knotted rugs (although they surely made flat-weaves and felt)?
It is possible that they learned it from Anatolians (symmetric knot) when they settled there and from Persians (asymmetric) - but somehow those who learned in Persia carried the art back to central Asia?

The similarities of design in the tribal/rustic productions of oriental rugs could be explained with:
1) Motifs born in more restrictive weaving techniques.
2) Urban influence - there is a whole chapter on this in Opieís Tribal Rugs and, as John wrote in another thread, "Murray Eiland has made a very plausible case for believing that the primary flow of design is from urban centers into the country-side".
3) Influence from India, China, Buddhism, Zoroastrism, whatever.
4) Pure copy of successful motifs for aesthetic or commercial reasons.

N.B.: these are only a few (albeit lengthy) thoughts, not an article of faith.
Please feel free to shoot them down (them, not me) if you feel so.



Posted by Michael Wendorf on 02-06-2002 11:51 AM:

Khosrow and early weaving

Dear Filiberto:

Thanks for this interesting, careful and thought provoking thread.

Regarding the legendary Khosrow carpet, I am not certain we can assume that this was a knotted pile carpet. I do not think this is critical to your thread anyway.

In Anatolia, carbonized textile fragments dating to 6000 BC are documented. They are housed in the Archeological Museum in Ankara. There, I think, we can speak of an 8000 year weaving tradition with assurance. These earliest textile fragments are made of bast fibre, probably flax. The weaving probably predates doemestication of sheep by a long time as well as the modern loom - no sheds, no heddles. The techniques are probably closest akin to basket weaving. Weft twining, transverse wrapping.

In Iran, fragments dating to at least the ninth century B.C. have been found near Lake Urmia, from Hasanlu. Other fragments of this date are also known in Anatolia around Lake Van.

The influence of the Mittanis and/or their introduction of evolution of knotted pile carpet weaving in Anatolia/Kurdistan or to Kurdish ancestors is fairly difficult to establish notwithstanding Professor Izady's remarks. I think it is fair to say that knotted pile has a long history in Anatolia, probably thousands of years and nearly so long in Persia. My own belief is that Kurds and their ancestors were right in the middle of it and that this all predates and Turkic influence.

I would not rule out the possibility that knotted pile weaving arose in several places among several groups independent of each other.

Thanks for this thread.

Regards, Michael

Posted by Filiberto Boncompagni on 02-07-2002 05:36 AM:

Dear Michael,

On the contrary, I have to thank you for your kind words.
Frankly I was expecting some heavy critics or WORST, a heavier silence (which is still likely to occur).

About the Khosrow carpet, you are right: I did not say it was a knotted piled carpet (although I think it was) because there is no evidence of that.

And - anyway - it is not essential for my point of view.

If the Persians learned the asymmetrical knot from the Turkmen (which I doubt), why Turkmen did not teach it to Anatolians too?
Because Anatolians already used the symmetrical one.

Letís suppose, instead, that Turkmen knew and used the Turkish knot but were unable to influence the Persians (asym. knot) for their long established traditionÖ

If this is the case, why most of the known production of central Asian Turkmen rugs is made with the Persian knot?
I hope my phrasing is clear enough. Iím not so sure.


Posted by Sophia Gates on 02-07-2002 05:03 PM:

Dear Filiberto

This is an excellent question. It begins to appear that the Turkmen may have learned pile weaving from more than one source. And, that pile weaving has been around for thousands of years. Oh - for a time-travel machine!

A friend of mine has suggested that carpets are a type of historical document. I think that's an excellent thought - I wish, though, we had more really old ones.

Another thing this post reveals: our history - the way we write it and understand it - is sadly lacking in many respects. We often mark the past by naming battles and the beginning or ending of powerful dynasties. Sometimes that's all the "hard evidence" we have to go by - but we're left with an opaque lens when it comes to understanding real life - the ebb and flow of the rivers and the silt they bring to the fields; the harvest songs; the birth of a child; the spinning of a skein of wool. Who were these people? How can we touch them?

I'm hopeful that DNA testing can reveal some truths, as Greg has suggested. Already we have learned from DNA analysis that when we speak of people migrating - we often mean WOMEN. I will have to try and track down the article I'm thinking of. But apparently some testing has revealed that men tend to remain in the tribal homeland but women go away. Men near China, for example, show the same DNA as their ancestors but women as far away as Greece share it. Does this make sense? To me it means that one way designs and craft techniques travel is quite literally: they go with their inventors or practitioners in the suitcase! So it's a much more direct means than the famous anecdotes about people seeing rugs in bazaars, so forth. And it makes more sense. It would also explain why some Salor rugs are left-hand knotted and others right, for example.

Marriage of course is only one way people move. Another way is through slavery. Tribal people like the Turkmen often captured townspeople and probably vice-versa. Also, I read something on the 'net the other night, about nomads in general I think, which again stated something logical: that many nomads, given sufficient wealth & opportunity, move into towns. This would explain my theory that designs come into town (or up, if you must) and not just "out" or "down".

Posted by Steve Price on 02-08-2002 06:32 AM:

Hi Sophia,

I wonder if you would be good enough to track down the article to which you refer, the one that provides DNA evidence that some Chinese women migrated (voluntarily or otherwise) to Greece while the Chinese men stayed home. I'm having trouble figuring out what kind of data would demonstrate such a thing.

Thanks, and regards,

Steve Price

Posted by Sophia Gates on 02-08-2002 11:18 AM:

Steve: I'll do my best. And I thought I explained it to some degree: women leave home. Men stay put. This of course in patriarchal societies. Does this help?

Posted by Steve Price on 02-08-2002 11:43 AM:

Hi Sophia,

As the son of an immigrant father (my mother was born in the USA, although her parents were immigrants and so were her older brother and sister), I'm afraid I don't understand the assertion that women leave home and men stay put. My observation - only anecdotal evidence, but the best I can offer right now - is that some men stay home, some leave; some women stay home, some leave.

My last post, though, was simply a request for the source in which DNA analysis is cited as evidence that a particular group of women left and the corresponding group of men stayed home. The reason I asked, which I mentioned in that post, is that I can't think of any DNA analysis results that might lead to that conclusion even if the event occurred. So, I'd like to see the source or at least some summary of the findings and arguments. Perhaps I'll learn something from it.


Steve Price

Posted by Sophia Gates on 02-08-2002 12:12 PM:

Link to DNA Article


Here is the link to the article I mentioned:

Hope this helps!


Posted by Steve Price on 02-08-2002 12:54 PM:

Hi Sophia,

Thanks for the link. I had forgotten that mitochondrial DNA comes entirely from the mother. At least now I understand how such data can be obtained and used.

I didn't see any mention of women from China going to Greece and leaving the men behind, though. Nor did I see anything in it that would justify a sweeping assertion like "women leave, men stay home". I think the article's take-home message can be summarized by saying that the study showed that in most cultural groups the percentage of women who emmigrate exceeds the percentage of men who do so, and that this finding was unexpected.


Steve Price

Posted by Sophia Gates on 02-08-2002 01:34 PM:


Dear Steve:

I was talking about a process that occurs over generations. And I was trying to make a point.

But I'll bet, if you looked, you'd find plenty of Mongol DNA in Greece. Greece used to be part of the Ottoman Empire. And we know where they came from, right?

In fact there's been a great deal of cross-fertilization from Greece to Asia and back again, both of people and of culture, art and religion. Maybe even weaving techniques!

Posted by Steve Price on 02-08-2002 01:59 PM:

Hi Sophia,

Like you, I'll bet there's China-type mitochondrial DNA in women in Greece. I'll bet you'd find China type nuclear DNA in men in Greece, too. Your statement had to do with gender-specific migration of women, and the article was cited in support of your position - "women leave, men stay". In fact, what the article says is that men and women BOTH leave, but contrary to what people have long believed, women do so more often in many cultures. The article doesn't mention the extent of the difference.

Nobody has suggested that there was little or no exchange of culture or peoples between eastern Europe and Asia. That dreary old history stuff is full of accounts of significant exchanges and the men who were major players in them, and there's little reason to doubt that exchanges occurred on smaller scales all the time.


Steve Price

Posted by Yon Bard on 02-09-2002 07:44 PM:

Steve, if it is the case in most cultures that brides move in with their husbands' families whereas the bridegrooms stay in or near their own family's homes, then this would result in a faster diffusion of the females' genes than the males.' So the claim that women migrate more than men seems reasonable.

Regards, Yon

Posted by Yon Bard on 02-09-2002 07:54 PM:

Turkmen knotting

On the subject of why the Turkmen use the specific knotting that they do, I have on another occasion raised the question of how the marvelous symmetry in which the different types of knots among the tribes could have arisen. To recapitulate, we have the following division:

1. Tribes with fine weave and predominantly red ground color:
- Tekke: asymmetric right
- Salor: asymmetric left
- Saryk: symmetric
2. Tribes with coarse weave and predominantly brown and aubergine ground color:
- Chodor: asymmetric right
- ArabachI: asymmetric left
- Yomud: symmetric

The only reasonable explanation is that at one time a general council of tribal chiefs was convened, and it alloted the available weaving modes to the various tribes.

Yes, I know that there are many exceptions and irregularities, but that's to be expected. The important thing is that there is much truth in this scheme.

Regards, Yon

Posted by Steve Price on 02-09-2002 08:52 PM:

Hi Yon,

The argument,
if it is the case in most cultures that brides move in with their husbands' families whereas the bridegrooms stay in or near their own family's homes, then this would result in a faster diffusion of the females' genes than the males is correct only if the bridegrooms and the brides are in different geographic areas. Moving from one tent to another in the same community would have no effect at all.

I do not dispute the article Sophia cited, by the way. It claims to show that in most societies the female's genes migrate more than the men's, and the evidence seems reasonable. What I disputed was the assertion that this warrants the statement, "men stay, women leave", with the clear implication that this was evidence of the victimization of women by men. The article made no mention of the extent to which the migration frequency differs between men and women, and it specifically noted that men often migrate (as in military activity, for instance), which is so well recognized that the finding was contrary to expectation.

There are many possible explanations for the difference in dissemination of men's and women's genes. One could be that the women get carted off as slaves pretty regularly, but it doesn't happen to the men as often. For all I know, this is a factor, but that information isn't in the cited article. Another could be that some women follow the men who go to war, the men get killed in battle and therefore make no contribution to the local gene pool and the women stay and have children. Again, I don't know the extent to which this occurs, if at all.

Anyway, I have no reason to doubt the findings reported in the article. The importance of the various causes of the phenomenon remains to be determined.


Steve Price

Posted by Sophia Gates on 02-09-2002 09:47 PM:

Women and Their Work

Dear Steve,

First of all - I was NOT intending to demonstrate that women get victimized by men, at least not in this article! What I think it DOES demonstrate is that, over a period of time, a woman and her offspring might migrate, first to a neighboring tent, then to the neighbor of that one, and so forth, until in a few generations her offspring and presumably their knowledge and skills will have been transferred as well, over a large distance and into alien territories.

This goes a long way toward explaining, I think, how weaving traditions including structural practices and certain design characteristics might wind up hundreds or even thousands of miles away from their origins.

Seem reasonable?

Posted by Yon Bard on 02-09-2002 10:09 PM:

Steve, you say "Moving from one tent to another in the same community would have no effect at all. " True; but who says all moves are within the same community? If they were, all humanity would have broken down into a million isolated interbred 'communities.' This is clearly not the case.

Sophia, you describe the diffusion process very well. However, if it also applied to weaving techniques, then the observed technical homogeneity within each community could not stand up. It may account though for some of diffusion of motifs, i.e., the woman from tribe A who moves to tribe B will usually adopt tribe B's technique (things like knot type), but perhaps incorporate some of her native motifs in her weavings, which will then be copied by her new relatives and neighbors. She might also teach them some entirely new skills, perhaps soumac instead of plain weave.
Of course, this is pure speculation, and hardly original to wit.

Regards, Yon

Posted by Steve Price on 02-09-2002 10:43 PM:

Hi Yon,

I'm no expert on marriage customs throughout the world, but it's my impression that the overwhelming majority of marriages are, and have been, between people living in the same community. In such cases, there is no migration in any meaningful sense, and the issue of whether the husband moves into the bride's family home or the bride moves into the groom's family home makes no difference at all in the context of this discussion. Most migrations with which I'm familiar involve families, the husband and wife moving together. And, please note, that in Sophia's preceding post she specifically refers to the wife remaining in the neighborhood, albeit somewhat further from her home than the husband is from his.

How, then, to account for the greater apparent mobility of women's genes? Well, I can think of ways that have nothing to do with who left home to live elsewhere. For instance, consider a population of immigrant families. Men die much younger than women do. Perhaps the surviving women tend to remarry and have children. Perhaps some women who are wives of soldiers follow their husbands into foreign lands and when the husband gets killed in battle, the wife stays either voluntarily or as a captive. There are other scenarios as well. I just don't know how common it is in most of the world's population for women to marry outside their communities and leave, compared to the frequency with which men marry outside their communities and leave. My suspicion is that it is, percentage-wise, a very low number in each case if we consider the whole world's population. Most people never actually get out of their home area.

The basic finding is very interesting, and I'd love to see an analysis of the mechanisms involved.


Steve Price

Posted by Yon Bard on 02-10-2002 09:41 AM:

Steve, as a chemist you should know that if two substances diffuse out from a center, even a very slight difference in their diffusivity will, in time, bring about a substantail seperation. That's the principle of most purification processes. Even if 99% of marriages are 'within a community' this is not enough to stop the differential diffusion over many generations.

Regards, Yon