Posted by Gene Williams on 03-28-2006 03:20 AM:

Dates on Carpets - what a difference language makes?

My brother recently bought an old Caucasian on the internet advertised as a "Darbend" but which I think is some sort of Kuba - Shirvan, Afshan, Karabagh? - not sure. There is a date woven into it (photo below). In checking with our Persian speakers, they identifed it at 1313 "hidri shamsi"..that is the old pre-islamic Iranian solar year calendar. In the solar calender 2005 AD is 1385 hidri shamsi.

There is another calendar used based on the lunar year in the Arabic-Semetic world which is more commonly known in the West "hidri kamri". 2005 is about 1412 (or 1415- we had a dispute) in hidri kamri.

The Persian speakers here claim that the Iranian solar calendar is used by most Indo-European speakers in the whole Near East-Central Asia area. Indo European speakers would include urdu, Baluch, Pashtu, Dari, Farsi, Persian, etc. Well from the below darbend-kuba (or whatever?) it would appear that some Turkic speakers in the old Persian caucasian Khanate of Kuba may also use hidri shamsi.

The question is, how does this affect other dates seen in carpets? How widely used in fact is "hidri shamsi"? Which calendar is normally used by Turkoman tribes? By Caucasian Turkic ethnics? By Anatolian rug makers? I will be talking further with linguists here to try to obtain some answers.

Posted by Filiberto Boncompagni on 03-28-2006 04:22 AM:

Hi Gene,

Nice rug. That pattern is called “Kuba Afshan” on the market.
As for your calendar-related questions, we discussed about that a few years ago.
I need a little time for retrieving and re-digesting the discussion.


Posted by Filiberto Boncompagni on 03-28-2006 05:54 AM:

OK – More often than not, dates on rugs appear to be rather untrustworthy for a series of reasons that I won’t mention here.

On the other hand, most - like this one - appear plausible.
To be specific, your brother’s rug is dated 1313 A.H. (year of Hegira), that corresponds to the period between June 24th 1895 to June 11th 1896 A.D. Which seems correct.

The Islamic calendar is a lunar one (hijri qamari).
The hijri shamsi you mention is the solar Islamic Iranian calendar that was legally adopted by the Persian parliament on March 31, 1925, specifying the origin on the calendar (Hegira of Muhammad from Mecca to Medina in 622 CE) mentioning that the beginning of the year is the first day of Spring. It was used also in Afghanistan and by Muslims in other parts of Central Asia.

Notice that the hijri shamsi for Persia was nothing else that a re-introduction of a pre-Islamic (Zoroastrian, I think) calendar. But the Islamic lunar calendar was widely used since the Arab conquest of Persia until 1925.

To convert the hijri shamsi to A.D. suffice to add 622. Converting hijri qamari is much more complicated. Luckily there are tables of conversion.

In any case:
1 - In Caucasus Muslims used the classical Islamic lunar calendar (albeit sometimes they used A.D. dates too, with Western numerals).

2 - For dates antecedent to 1925 the problem doesn’t subsist.


Posted by Filiberto Boncompagni on 03-28-2006 06:22 AM:

Further note

Just found on the web:

The current calendar has been used in Iran since 1925 and in Afghanistan since 1957. However, Afghanistan used the Islamic calendar in the years 1999-2002(Taliban period).


Posted by Steve Price on 03-28-2006 06:58 AM:

Hi Filiberto

You wrote, To convert the hijri shamsi to A.D. suffice to add 622. Converting hijri qamari is much more complicated. Luckily there are tables of conversion.

This is true, but for ruggies, the important period is from about 1800 AD to about 1925 AD. This doesn't require anything complicated if an approximate conversion is good enough. All you have to remember is that 1900 AD = 1318 AH. You can then add or subtract the same number of years from each to make conversions. For example, to convert 1850 AD to the AH calendar, simply subtract 50 years from 1318, which gives you 1268 AH. The error introduced by this simplification is only 3 years for each century. For most purposes in the rug world, this is adequate.


Steve Price

Posted by Filiberto Boncompagni on 03-28-2006 08:03 AM:

Hi Steve,

Actually, we have a bit of a problem here – limited, I hope, only to Persian rugs.
Are there any indications to distinguish if a date is written according to the solar or to the lunar calendar?
Because, for example, if we find a Persian rug dated 1304, what that could mean?

It could be
hijri shamsi (1304+622= AD 1926)
but it could be
hijri qamari (A.H. 1304 = AD 1886)
as well.
UNLESS the hijri shamsi is indicated in some way.


Posted by Steve Price on 03-28-2006 08:21 AM:

Hi Filiberto

It's easy. If the rug belongs to someone else, especially someone who is trying to sell it to you, the date is hijri shamsi. If the rug belongs to you, especially if you're trying to sell it, the date is hajri qamari.

Nothing to it.

Steve Price

Posted by Gene Williams on 03-28-2006 08:42 AM:

"hay" and "shin"

Ok, the persian speakers just looked again. in this rug there is a "hay" character at the end of the date 1313 (the squiggle with a loop in the middle) which means it is the solar calendar which means 1313 = 1932. If there is no character or a "shin" character (squiggle with a sickle with 2 dots above it), it will be the Islamic lunar character date. i.e. 1313 = 1895 + or -... Is there anyway to blot out the "hay" character?


Posted by Filiberto Boncompagni on 03-28-2006 09:06 AM:

Hi Gene,

First: 1313+622=1935 and that rug doesn’t look as 1935.

Second – as I wrote few years ago:
I have several items of silverware from Caucasus. Four of them have inscribed dates. The first one is a nielloed silver belt dated 1893 with western numerals.
The second one, another silver belt, is dated 1318 in Arabic numerals. If this is an "Islamic lunar date" it should correspond to AD 1900. Otherwise it should be (1318 + 622 = 1940).
I am confident that AD 1900 is the right one not only because the condition and the style of the two belts are similar, but because they are hallmarked with the stamps used in Imperial Russia in those years. In particular, the one dated 1318 has the typical stamps used between the years 1896-1908.

The other two items I have, a Caucasian dagger (kindjal) and a kindjal belt are dated 1323 and 1333 with Arabic numerals. They have no hallmarks but, again, comparing their style and condition with the two belts they should be respectively AD 1905 and AD 1914 - they definitely don’t look as being 40 years younger.

So, I have 3 very good reasons (one of which is proofed) to believe that in Caucasus they used the normal Muslim calendar based on the lunar cycle.



Posted by Filiberto Boncompagni on 03-28-2006 09:47 AM:

Hi Gene,

If your Persian speaker friend is still available, could you please ask him/her about the date on this Bakthiari?



Posted by Gene Williams on 03-28-2006 11:37 AM:



They said it was hard to read. Both lines read from right to left except the date (four characters on the left in the second line) which reads left to right. The best they can do is as follows:

"The Person Who Ordered This Rug Haji Ali Khel Jan"
"1315 Bakhtiari"

The date they said is indeed "sin" i.e. Lunar Calendar. 1315 = 1897?

They speculated that Khel Jan might be a village name.


Posted by Filiberto Boncompagni on 03-28-2006 11:46 AM:

That’s correct.
Well, almost (it's AD 1892).
But is there a character or sign indicating the lunar calendar?


Posted by Gene Williams on 03-28-2006 11:50 AM:

Say Hey, Kid

They said there is no sign. If there is no means lunar calendar. Apparently you need a "hay" to show its solar calendar. They showed me the character on the persian alphabet I pulled up.

Posted by Filiberto Boncompagni on 03-29-2006 06:12 AM:

Hi Gene,

I made a small research on Caucasian rugs, using four books:
Bennett’s “Caucasian”,
Bennett&Bassoul's “Tapis du Caucase – Rugs of the Caucasus”,
Gans-Ruedin's “Le tapis du Caucase”
and Kaffel’s “Caucasian prayer rugs”.

That makes for more than 800 rugs. Caucasian weavers are by large the more prolific, as far as dated pieces are concerned, and I found dozens of dated rugs.

Not taking into account rugs with western numerals, illegible Arabic dates or pieces with several conflicting dates, MOST of the dates are plausible with the use of the classical Islamic lunar calendar (A.H.).
That’s especially when the dates correspond to the turn of the last century: there are a lot of such rugs and they cannot be forty years younger, as it should be if the hijri shamsi, (the Persian solar calendar) was used.
On the contrary, there are several cases where the rug actually appears older than the date itself.

I found, however, a few suspect cases that could be explained – and compatible - with the use of the solar calendar. Oddly, they are all Shirvan/Daghestan ivory-ground prayer rugs.

Let’s start with plate 428 from Bennett’s “Caucasian”

Here is a close-up of the date, the best my scanner can do:

Bennett’s caption to the plate: 428 This rug constitutes something of a puzzle. It is dated 1220 A.H. (A.D. 1805/6), although at first glance, this would seem to be impossibly early, The overall appearance of the rug is static and stiff, qualities not normally associated with early pieces. Technically, it is of very fine quality and our colour reproduction does not do it justice; it will be found more accurately reproduced in Peter Bausback's Antike Orientteppiche (p.236). According to Bausback, the piece has a wool warp and weft and is partly piled in silk. Although these are structural characteristics associated in particular with older pieces, I think the date is probably 'wrong'…
It should be noted, in addition, that the use of silk, either in the foundation or the pile, is not necessarily an indication of old age. I is arguable that we might be dealing with a copy of an old rug made perhaps in the early 20th century, but incorporating the date which appeared on the model.”

So, in this case, even with the forty years “bonus” offered by the Persian calendar (1220+622=1842), we are still far from the early 20th century.

Gans-Ruedin page 211:

the book offers also a “magnification” of the inscription:

No doubts on the age of the rug are offered here, only the observation that it’s parfaitement conservé (perfectly conserved). Well, it could be forty years younger, what do you think folks?

Last one is Kaffel’s plate 78

and the detail of the inscriptions:

Kaffel’s comment: This piece belongs to a small group of Shirvan rugs with similar design, colouring and detail. All feature 'Kufic'-type borders, usually on blue grounds; a polychrome inner border of boxed swastikas; unusual flowers in the lattice field; and inscribed dates, usually to the right of the prayer arch. This example is dated AH 1252, which translates to 1837- probably 40 years earlier than when the rug was actually made. (Most of the analogous examples bear dates between 1875 and 1900, and there is no reason to suppose that this rug is any older.)

So, this is a perfect candidate for a “solar” date – provided the third digit is indeed a 5 and not an 8. Which could settle the matter without the need of the hijri shamsi after all.

Gene, as you can see the matter isn’t at all simple.
It would be nice to know if your Persian speaking friends have any comment on the inscriptions.

Of course, other comments and especially more examples of dated rugs - Caucasian or not – are welcomed.



Posted by Gene Williams on 03-29-2006 08:17 AM:

an update


Its 1800 hrs here and we're going to supper. Later tonight we'll take a look. I've tried to get to the bottom of the story on the solar-lunar business with these guys. I am confusing myself.

Turns out it seems that much of the Caucausas went to Solar dating about the time of the Russian occupation (around 1860 AD I believe). What does this do to the dates on Caucasian rugs in the Russian areas?

And for dates in Persian rugs, they have backtracked a little and now agree with the 1925 date for the institution of the solar calandar in Iran and for the date when it began to be used on iranian carpets. As you pointed out this opens a can of worms. Every date from 1303 (1925 if its shamsi) to 1385 (2006 - note: Persian new years just passed) is therefore suspect.

Well then, I asked how to tell if a carpet date is lunar or solar (using your Bakhtiari as an example)?

Best I can understand is that its in the "feeling you get" (or maybe what you want to see?)... They both agree that if a date has a "hey" after it like my brother's carpet then it is definitely "shamsi." It it doesn't have a "hey" it probably is lunar,...but not definitively ...just probably.

And as for the Russian occupied Caucasaus...that will take some research.

I'll let you know if we can make more sense of this; one of these guys has some experience with carpets but not much so...that's either good (an unprejudiced look) or bad (an amateur's opinion) depending on how you slice it.


Posted by Filiberto Boncompagni on 03-29-2006 10:23 AM:

Hi Gene,

Turns out it seems that much of the Caucausas went to Solar dating about the time of the Russian occupation (around 1860 AD I believe). What does this do to the dates on Caucasian rugs in the Russian areas?

As far as I know, the consequence of Russian occupation was the official adoption of the westerner (Julian, at the time) calendar, while Muslims still used “privately” the Islamic one. Which was, according to all evidence I’ve found so far, the LUNAR calendar.
So they used both, the solar/Julian and the lunar/Islamic.

For example, on Bennett & Bassoul's “Tapis du Caucase – Rugs of the Caucasus”, there is a Kazak rug dated 1887 in western decimals followed by the Arabic inscription “tarikh roussia” (Russian date).

And, of course Armenians and Christian Georgians had always used their Christian calendars (Julian again, I guess).


Posted by Filiberto Boncompagni on 03-29-2006 10:55 AM:

Persian rugs

Regarding Persian rugs, I don’t have much material, with the notable exception of a book on the Dafineh Museum in Tehran, with text in Farsi and (bad) English.
The book shows 110 Persian rugs.
Only three of them are dated: all pre-1925 and with “lunar” dates.
The detail of the “1315 Bakhtiari” above is one of them.

The sample is small (3 out of 110) but significant and the source is irreprehensible.



Posted by Gene Williams on 04-04-2006 11:33 AM:

Post 1925 Rug dates - document dates


It occures that if "hay" is used to signify the use of the solar calendar after 1925 in Persian rugs...ought not the same letter be used in dates in documents, pictures, miniatures or other uses as well? Has anyone looked to see how the date is represented say in marriage or birth records before and after 1925 or even copyrights or some such? It would be interesting to see how widespread the use of the "hay" (and/or "sin") it stop afterawhile as use of the solar calendar set in, etc.


Posted by R. John Howe on 04-10-2006 08:20 AM:

Dear folks -

We recently ran into an instance of a Caucasian rug with an inscribed date that demonstrated to us anew the dangers that western readers are exposed to as they attempt to read such things.

The date began with what seemed to me the older islamic numerals for "1" and "2".

The chart below provides not terribly legible instances of these two numerals in its fourth and fifth lines at the left.

So far so good.

The third numeral on the rug was "9." and in fact that is the same in the older numerals used for such inscriptions. (Look at the chart again.)

But the last "numeral" looked for all the world like a sloppily drawn "2."

So a couple of us were suckered to read this inscribed date as 1292 and to muse that it was perhaps the first instance we had seen in which an inscribed date on a rug had "mixed" old and new numerals.

I asked a learned Persian friend about it. He smiled and say "You're misreading it. The fourth numeral in this date is represented by a dot. That's how a "zero" was indicated in the old numerical system. (Look again at the chart above.) So the date should be read as 1290 (about 1874). " We looked more closely and sure enough there was a dot before the "2."

The part of the inscription at the end that looks like a "2" to you is the Farsi word "senneh" (like the town and knot). It means in this context "in the year of."

The friend with whom I was examining this inscribed date said that it was the first instance he had seen in which the word "senneh" occurred after rather than before the numerals.

But this experience suggests that it is easy for westerners to read things into inscribed dates things, like twos, that are not there.


R. John Howe

Posted by Jack Williams on 04-13-2006 09:37 PM:

Turkey also switched to solar calander

From this web site,

apparently Turkey under Attaturk also switched to the solar calander about the same time as Persia. Perhaps other turkic related tribal groups followed as well...especially in the Caucasus. In any case, other well known problems with dated caprets are touched on in the cited reference above.

Re-reading the famous Hali interview with Jerry Anderson, I found that he commented on dated carpets. Here is his comment taken from that interview as featured on Tom Coles site...

..."What is this about a woven date here? I really doubt it – for a start most Baluch have no concern for dates and when they do, the inscribed dates in what are normally workshop rugs are usually placed in or near a corner, not floating freely in the field. I used to buy fragments of rugs which had woven dates, just to get some idea of how to date rugs in general. I had a whole collection of Turkoman and some Baluch fragmented prayer rugs with dates. But they’re all gone now."

What a shame we cannot review his collection or know his other thoughts about dated carpets. But, his statement does cause one to ask the question of how frequently carpets from even a semi-tribal environment would be woven by a person knowledgable about dates and calanders.

The date on my carpet (which began this thread) is apparently sound (after blacklighting it) and I guess I will accept the "hay" symbol...unless I try to sell it of course! The details of the carpet do appear older than 1935 though. True, I don't have a lot of other carpets circa 1935 to compare it with...but dyes and weave and composition seem older.

Whatever, the carpet is being cleaned and is very interesting to me, as has been this intelligent and informative discussion, thanks to all.

Posted by Jack Williams on 04-13-2006 10:08 PM:

On a belated note, I located the following on JBOConnell's web pages - Guide to Caucasian Rugs -

(note: his web site discussion section on Kuba-Afshan rugs is under "Guide to Kuba Rugs.")

After discussing the research that led to the opinion by some academics that all "patterns came from Russian Government (first Czarist then Communist) Kustar pattern books," he notes his position, then states...

"To confuse the matter even more I suggest that many of the best rugs attributed to 4th quarter 19th century are actually 1st or even second quarter 20th century. A constant problem is that Rug Collecting is more about money and ego then it is a science. My hope is that by opening my notes, thoughts, and theories I may help to stimulate others to take this subject further than I can."

I found that statement interesting in light of the seeming contradiction between date and "look" of my rug. Regards.

Posted by Steve Price on 04-14-2006 01:45 PM:

Hi Jack

There's no question that the Russian kustar patterns existed, and a number of them are reproduced in Wright and Wertime's book on Caucasian rugs (exact title escapes me at the moment), and many can be seen on line in at least one of the superb essays on Richard Wright's website.

I think the conventional wisdom is that they were used from about 1875 on, and that the designs imposed upon the village weavers were more or less true to the designs traditional to their villages. Wright, who usually checks his facts, believes that the kustar field designs were true to the local traditions, but that the borders weren't.


Steve Price

Posted by Jack Williams on 04-14-2006 10:17 PM:

So much for romance...

Hummm...kind of takes some of the romance away when what you imagined was...

...a family-made long carpet, designed and lovingly woven in an eons old pattern passed down by the 101 year old great aunt who constantly kibitzed about the construction from the vantage of the pillow mound in the corner...the design modified by the whims and moods of a mountain tribal woman who then used it in the smoky fire-warmed caucasian village hut to cover the dirt floor while her numerous urchins rolled around on it playing with the family wolf-hound and her giagantic full bearded husband and his friends roared songs, drank vodka and danced...

...and suddenly like a cold polluted stream of Hurricane Katrina flood water (I'm from New Orleans), you discover an alternative universe for your rug with the image of soviet commisars with leather coats, boots, bad hats, and whips forcing cowering peasants to slave-labor produce communist officially "approved" designs for the glory of the fatherland...the finished product, damp with tears of the starving weaver (thank goodness for fast dyes) torn from her grasp and bundled off to the glorious people's market after all 428 forms have been filled out in triplicate...

Naaahhhh....I'll stick to the first image, reality isn't nearly as great as it is cracked up to be.

Still, this is quite interesting. Steve, I wonder how detailed these designs were and whether they were just individual elements or full patterns. Guess I will have to search out that book. It is hard to imagine that this particular rug, with it's asymetric elements, was the prouduct of a full rug design. If it was, then the designer deserves for me to buy him another glass of vodka...

Thanks for opening this door for I will have to start learning something other then the names of every baluch tribe, including some who never existed. Regards...

oh, one other thing, this discussion line led me to hours on the computer, turning up this web site with about 4,000 pictures of Caucasian Azerbaijan rugsi rugs, links to the state musuem, etc. And to think all this baluch rug lover wanted was a little more color in his life (forgive me tom cole!)...

Posted by Steve Price on 04-15-2006 06:07 AM:

Hi Jack

I understand your reaction; I lead a rich fantasy life myself.

The Wright/Wertme book is worth owning, but you can see a number of the kustar rug drawings even before you get your copy. The page on Wright's website that I linked in my previous message includes many of them (maybe all the ones that are in the book), but you have to scroll down quite a way before you get to them. Here's the link again:


Steve Price

Posted by Filiberto Boncompagni on 05-02-2006 01:32 PM:

Hi Don,

Check this link for Armenian dates on rugs: