| Subject | : | Shared design and common attribution |
| Author | : | Michael Wendorf |
| Date | : | 08-27-2000 on 04:29 p.m. |
| Dear Daniel: First, congratulations on a fantastic salon presentation featuring beautiful examples to discuss. Thank you. You ask the question, when is shared design a basis for common ethnic attribution? Second, are shared designs an expression of the ethnic mix of these areas? The second question seems to be easier to address. I think that shared designs are unquestionably an expression of the ethnic mix of these areas. I think it is also more than that. The greater lower Caucasus, eastern Anatolia and northwestern Persia have seen many comings and goings over the last thousands of years. In a way, the region has been a kind of portal. At the same time it has avoided control and domination by centralized authorities for extended periods. In this way, the area has also served as a buffer between empires as well as a bridge among and a portal or gateway for different peoples. Given this heritage, it seems that the ethnic mix and the cross-fertilization of ideas, culture and must find various expressions, including and not limited to shared designs. The question of when is a shared design a basis for common ethnic attribution is much thornier. Case in point is your beautiful ivory ground long rug with memling guls and secondary white ground guls with eli belinde like motifs. This distinctive rug is one of at least four rugs with the same distinctive design and color scheme that I am familiar with. Of the other three, one is in the collection of Dr. Bob Emry in Arlington, Virginia ("Bob's rug"). I believe that Bob will be posting images of his rug for this Salon. The other third and fourth examples can be seen in Hali as follows: Hali 105 page 157 labelled Kurdish, SW Caucasus; Hali 111 page 119 on the wall of the Pap gallery image. The first of the four rugs I handled was Bob's rug. I initially thought this rug was Kurdish based on the color palette, red wefts and selvedge. This selvedge consisting of 3 units wrapped in two color groups and reenforced with extra selvedge wool yarns seemed definitively Kurdish. However, I then realized that the warps were three ply. The use of three ply warps virtually eliminates the possibility of a Kurdish attribution insofar as Kurdish rugs are virtually always woven with two ply warps. The next possible attribution was Shahsavan, assuming the pile rugs we are currently calling Shahsevan were in fact woven by these people. However, the handle and structural details ruled this out as well. Having known Bob's rug for at least 5 years, I still do not have a comfortable feeling about who wove this rug. The structural analysis you have provided for your rug establishes that however close the shared design is with Bob's rug it is also very different insofar as it is woven with two ply warps. The border is not one I associate with Kurdish weaving although it is known on Shahsevan weavings including a spectacular long rug with cruciform medallions that Peter Pap advertised and exhibited several years ago and a fragment seen at the last ACOR that formerly was in the Rudnick collection. Is your rug Shahsevan, is it Kurdish, is it something else? My recollection of the Pap rug with this design is that it had two ply warps as well, but with an unusually heavy handle. A handle much heavier than Bob's rug, for example, which is light and floppy. I have not personally handled the fourth rug (Hali 105)and do not know its structure. So-called Memling guls have been a favorite design for weavers of many groups and areas over at least the last 500 years. We could go crazy attempting to trace the design to an ethnic group. It simply cannot be done. The real problem is even worse in terms of answering your question. Even when we confine our analysis to one small and distinctive group of 4 nearly identically designed rugs, we cannot, in my opinion use that shared design as a basis for common ethnic attribution. There may be groups of rugs where that is possible, but I think generally we use design as a basis of ethnic attribution at our peril. At a minimum, shared design must be accompanied by shared or related structure, side and end finishes and colors. Thank you again for this most engaging Salon. Michael Wendorf |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Michael Wendorf |
| Date | : | 08-27-2000 on 10:39 p.m. |
| Dear Daniel: Having now gone through sections 3 and 4 of your Salon I would add that the carpet Herrmann published which would be the 5th member of this group with shared design also has three ply warps. Although Herrmann also suggests a Kurdish origin, the three ply warps would seem to rule this out. As I mentioned above, nearly all Kurdish rugs have two ply warps. In what seem to be the oldest rugs this is often one ply of light, undyed wool and another of darker, browner wool. There is another type, like the garden carpets, that may be woven on cotton, but again two plys. The Kurdish attribution used by Herrmann, Bozoglu and others for this group must be questioned insofar as only your rug has been confirmed to have two ply warps and the others seem to mostly have three or a number that is unrecorded. Based on the border and other structural charateristics, it seems unlikely that your rug is anything other than what you have described - Moghan or Savalan area, possibly Shahsavan. |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Bob+Emry |
| Date | : | 08-27-2000 on 10:45 p.m. |
| emry@starpower.net Dear all, First I'll illustrate a Memling gul rug
very similar in design to Daniel's "Moghan" (Memling gul) rug, and second
will be another rug that might be related to the first by virtue of its
similar structure. The "Memling gull" rug: Except that this rug has
three columns of Memling guls, its field design is otherwise virtually
identical to that of Daniel's rug (and to the ones Michael Wendorf
mentioned with references to pictures in HALI)---same memling gul, same
minor gul, and same smaller elements in the interspaces. The border is
different-note (see detail pix) that the central design of the octagonal
border medallions is repeated as the central design of the memling guls. I
also notice that in Daniel's rug, the central design of the memling guls
consists of four triangles, which is similar to the center of the border
rosettes, but rotated 45 degrees. Picture of the rug |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Bob Emry |
| Date | : | 08-27-2000 on 11:12 p.m. |
| emry@starpower.net I have some higher-resolution pictures of the two rugs I discussed above posted on web space. If you wish to see them, the following two URL's will take you to index pages with links to the pictures. http://emry.simplenet.com/mem/ http://emry.simplanet.com/jk/ |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Guido+Imbimbo |
| Date | : | 08-29-2000 on 11:14 a.m. |
Dear Daniel, here is another reference to your beautiful
Transcaucasian/Moghan/Shahsevan rug. It is lot 14, Rippon Boswell 28 March
1992. Unfortunately the original picture is in black and white. The rug is
labelled North West Persia, Meschkin region. Size: 306x111cm, dated second
half 19th century. That's how far goes my German language skills in
reading the catalogue description.
 |
| Subject | : | RE:A shared design |
| Author | : | Daniel Deschuyteneer |
| Date | : | 08-30-2000 on 10:48 a.m. |
daniel.d@infonie.be
Dear all, Until now our discussion permits us to isolate at least one
group of rugs, wherefore a Kurdish or a larger Northwest Persian
attribution has been proposed by their owners. These rugs are related
through shared design and color scheme, as well as through common sizes,
and common structure characteristics except for the use of 2 ply or 3 ply
wool warps. Here are the structural characteristics of this group: Runner
format – mean ratio 1/3 Pile: long and silky symmetrical knots - 56psi to
79psi - Warps: 2 or 3 ply white or mixed wool – no depression. Wefts: dark
red or brown wool – 2 to 6 picks – no information about eventual weft
crossing Reinforced selvage (selvage warps interlaced by the ground wefts
and reinforced with interlacing additional yarns) Ends: no information
Design: gul forms and other old Turkish devices Color: characteristic
apricot and light yellowish green as well as light blue. Notice also the
frequent use of extremely corrosive dark brown color.  The rugs I have included
in this group are from left to right in this constructed photo, 1/The
"Bob-Memling gul rug" 2/ The "Pap-Hali 111" rug 3/ The Herrmann rug 4/ The
"Bozoglu-Hali 105" rug 5/ The "Bob-Grid rug" Details of each of them are
available at the bottom of this message. A NW Persian/Azerbaidjian
attribution seems to me to be the safest one. Michael, noticing that most
of them had 3 ply wool warps rules out an eventual Kurdish attribution and
this make sense at least if we were in a strictly Kurdish area but that’s
not the case. My Moghan rug as well as the rugs I related to this group
through shared design and color scheme in the second part of my Salon are
may be from other origins or woven by other groups. The question is open
Thanks, Daniel GROUP 1 – AZERBAIDJIANI RUGS 1/Bob’s "Memling gul" rug – NW
Persian Kurdish? Yarn: spin Z Size: 39" x 118" (1m x 2.95m) Knot:
symmetrical, H 9pi V 8pi 72 psi; H36/dm V32/dm 1152/dm² Warp: 3 ply wool
mixed white, tan, brown. Most warps have two white and one tan or brown
ply. No warp depression (some areas have slight depression) Wefts: dark
red wool, usually 4 picks (rarely fewer) Selvage: 3 units(2,4,2)
reinforced with extra selvage wool yarns in two groups with the colors
arranged to form checks. 2/Pap rug Hali 111 rug page 119 on the wall of
the Pap gallery NW Persian Kurdish? Probably 2 ply wool warps (Michael
Wendorf) Not available 3/Herrmann Eberhart: Southern Transcaucasian rug
101 x 302 cm Warps: natural 3 ply ivory wool, Z3S Wefts: brown wool, 2
singles, 2 & 3 picks Pile: symmetrical knots, one or two wool singles,
height 8-13 mm, V/H 42/30 – 1260/dm² 79psi 4/Hali 105 page 157- Ziya
Bozoglu NW Persian Kurdish? 5/Bob’s "diamond grid – reverse design" rug NW
Persian Kurdish? Yarn: spin Z Size: 39" x 109" (1m by 2.77m), some is
missing. Knot: symmetrical, H8pi V7pi 56psi, H32/dm V28/dm 896/dm². Warp:
2 ply wool, white or mixed white and brown or white and black. Weft: dark
red wool, except for one section (about 10 inches) where the wefts appear
to be undyed. Principally 4 picks. Selvage: 4 units (2,2,2,2) reinforced
with extra selvage wool yarns in four pairs. Erratically, the extra
selvage wool yarns interlace each of the two warps of the inner unit and
the other units in pairs. Uneven covering showing the ground wefts and
wedge shapes formed by groups of ground wefts Group 2 - Rugs related to
Group 1 through shared pattern which may have been woven in other area or
by other weavers. 1/Jim Burns rug South Caucasian Size:3’4" x 12’3" 1.02m
x 3.73m Size ratio: 1 / 3,7 Pile: wool, symmetrical knot V6 H8 48psi
768/dm² Warp: 3 reply (Z5S) ivory cotton – no warp depression Weft: ivory
cotton – 3-4 picks Original edge and end missing 2/Ulrich Shürmann plate 8
Caucasus – Bordjalou 129 x 244 cm Size ratio: 1/1,9 Structure analysis not
available 3/Peter Willborg Caucasus – Bordjalou Size: 63-66cm x 145 cm
Size ratio: 1/2 Yarn: spin Z Warp: 3 ply dark ivory beige wool – no
depression Weft: 2 ply madder red dyed wool, 4-6 picks, mostly 4 Pile: one
or two singles, pile height 2-10 mm mostly 6-7 Knot: symmetrical H33/dm
V27/dm 891/dm²; 56psi Selvage: flatwoven over 4-5 warps via the wefts (?)
Ends: bottom not extant, Top: red wool weft faced plain weave skirt
Colors: 14, ivory, pale yellow, pale greenish yellow, yellowish apricot,
pinkish red, dark red, brown red, yellowish green, sky blue, greenish
blue, navy blue, violet, grey black heavily corroded. 3/Taher Sabahi -
Ghereh 20-p.52 Northwest Persian/Southern Azerbaidjian/Hashtrud Structure
analysis not available 4/Rippon Boswell lot 14, 28 March 1992 North West
Persia, Meschkin region Structure analysis not available Size: 111 x 306
cm Size ratio: 1/3 5/ Daniel- Moghan rug Dimensions: 7’ x 3’2"; 237cm x
94cm Yarn: spin Z Pile: 2 singles Knot: symmetrical, H6.5 V8 52psi ; H26
V32 832/dm² Warp: fine 2 ply ivory wool; no depression Weft: fine 2 ply
light brown wool , 3 to 6 picks Selvage: not original Ends:
missing |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Michael Wendorf |
| Date | : | 08-30-2000 on 10:16 p.m. |
| Dear Daniel, Guido, Bob and all: Thanks to Guido for another image. It is helpful to see all the pieces together. In thinking about these pieces, I recalled another possibly related piece. It was advertised in Hali Issue 95 on page 124 by John and Suzan Wertime. I hope Steve or someone can post an image of it in this thread. [image] The rug measures 3.66 m x 1.09 m or 12' x 3'7". Like the Rippon Boswell rug posted by Guido, this rug is on a blue ground and has three rows of memling devices. The minor elements are similar but the overall design is more "crowded" (Daniel note!)making the negative space less powerful. But as I recall this carpet it had very beautiful, deeply saturated colors and was more free and playful in its drawing than the other examples. The border was different from all of the other carpets we have identified with an ivory ground major border and a yellow and brown sawtooth minor border. The major border is a border I have seen on some Kurdish rugs. John called his rug Shahsavan, Northwest Persia 19th century. I called John but he did not recall any structural information from this rug. So we do not know if it has two or three warp plys. Regarding the Rippon Boswell rug posted by Guido, I have translated the catalouge notes. They do not contain structural information to help us but do state that the rug is woven on a "night blue" ground and that it has undyed camel hair. The notes also state that it has gloosy wool and a soft handle and specific color tones that indicate a Kurdish origin. I would speculate that the color refered to is the apricot color we have seen in these rugs. However, I do not necessarily associate undyed camel hair with Kurdish weavings in this area and the border is also not one I would expect to find on a Kurdish rug. Given the other rugs that Daniel has identified such as the Burns and Willborg pieces it is possible that all these pieces come from some area associated with Bordjalu Kazak. I do think the 3 ply warps make a Kurdish attribution, even to an area not dominated by Kurds as Daniel offers, unlikely. My experience is that whether in Kurdistan or the Kurdish heartland or not, Kurds wove with 2 ply warps. The only alternative I can come up with if Kurdish hands wove these pieces is that the materials and looms were all supplied independently and Kurds merely tied the knots. These carpets seem too distinctive for that answer to be accepted by default, in my opinion. Thank you, Michael Wendorf |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Wendel+Swan |
| Date | : | 08-30-2000 on 10:52 p.m. |
| Dear all, Having examined the Emry, Bozoglu and Pap Memling gul rugs pictured as well as the Wertime long rug referred to by Michael, I think it would be very difficult to group them structurally or by handle or by tactile qualities. The Wertime rug, with some cotton in the wefts and some genuine camel hair in the pile, may be Shahsavan. But it might not be. In any event, it feels and simply looks completely different than the Emry rug, for example. While they may both come from the Southern Caucasus or NWP and they share a Memling gul design, the relationship probably ends at that point. I once had the chance to spend a couple of hours looking at several examples in one collection of antique Caucasian or NWP rugs, all with a Memling gul design. I was overwhelmed by the variations in the weave, even though it might be possible to write technical descriptions that would make them seem closer to one another than they are. Wendel |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Bob+Emry |
| Date | : | 08-31-2000 on 12:16 a.m. |
| emry@starpower.net Dear Daniel and all: I had just scanned the HALI
picture of Wertime's rug and was ready to post it, when I found that
Michael Wendorf had just mentioned it with a suggestion that someone post
it. Here it is: |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Michael Wendorf |
| Date | : | 08-31-2000 on 08:01 a.m. |
| Thanks to Bob for posting the image of the Wertime rug. We now have two rugs in the same design group that apparently have at least some undyed natural camel hair in them. The Rippon Boswell rug identified by Guido and the Wertime rug. Both of these are also on a blue ground. I do necessarily associate camel hair in Kurdish weavings from this area and its appearance is entirely consistent with Shahsavan weaving. Given Wendel's comments, it may be that the two blue ground rugs are Shahsavan. But maybe not. In either event, I do not believe that this necessarily make the others Shahsavan. I do not remember the Wertime rug well enough to comment on the warps but they do appear to be varied with both brown and ivory areas. This looks to be different than all the others. However, it is not something that is unique to any weaving group. Perhaps the most telling distinction would be end finishes. The Rippon Boswell rug looks to have a very distinctive end finish, but I cannot see it well enough to disuss it. Perhaps Daniel, who is working with Marla Mallett on a project involving end finishes, will hazard a guess from the image we do have. Thanks, Michael |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Guido Imbimbo |
| Date | : | 08-31-2000 on 09:53 a.m. |
Dear Daniel, Michael and all, here another picture (again originally
in black and white) of a related carpet of the Daniel's
Transcaucasian/Moghan/Shahsevan rug.
 |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Deschuyteneer+Daniel |
| Date | : | 08-31-2000 on 03:03 p.m. |
Dear all, I looked carefully with a magnifier at the end finishes of
the “Wertime” rug advertised in Hali95. All the dark brown and white warps
lie parallel and not any specific end finish can be seen. It’s impossible
to guess what the end finishes of the Rippon Boswell rug posted by Guido
look like. But if Guido would agree to post to Marla and me a magnified
photo with high resolution of these we can try. Nevertheless get a look at
this close up from the “Wertime” rug.
 |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Michael Wendorf |
| Date | : | 08-31-2000 on 10:29 p.m. |
| Dear All: All this attention on the so-called "memling gul" and the distinct group or groups of rugs we have isolated can sometimes cause us to lose sight of the forest for the trees. Reading the description Guido has provided of the Lefevre rug reopened my eyes a little. Lefevre describes the rug as consisting not of "memling guls" but of "multi-colored stepped polygons." Of course, at their core, memling guls are exactly that: "stepped polygons" with hooks emanating off of them. Now, if you think memling guls have been around for a while, try stepped polygons. I would be hard pressed to come up with a more basic building block of design. Note also that in Bob's diamond grid rug the border is made up of? If you answer "stepped polygons," you go to the head of the class. There are also few basic designs that are as conducive to the interplay of positive and negative and the sensitive juxtaposition of color than stepped polygons. I do not know what to make of the apricot color that Daniel and others have commented on that appears in most of the ivory group we have identified. It is a color that does appear in some Kurdish rugs. However, I have already stated why I do not believe these are Kurdish. In addition, I have this evening reexamined rugs in my own collection. I have three Kurdish rugs and bags with stepped polygons and notes on a fourth that I no longer have. All are have two ply warps ivory or ivory and brown. None have the same apricot color we see in the ivory long rugs. Two are woven on a brown ground(one corrosive the other non-corrosive, two have a pumpkin color as the ground (a color that is close to the apricot color but still clearly different) and warps that consist of wool and goat hair. None has the dimensions of the long rugs we have been commenting on and they all have very different handles from the ivory group. All four of the rugs I am referring to have some ivory, but very limited and mostly in the borders as opposed to the fields. In short, no discernible similarities to the rugs we are discussing. Finally, it is probably another salon or thread but, with all due respect to Mr. Lefevre, whatever these rugs are, they are not "Yuruk." -Michael |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Michael Wendorf |
| Date | : | 09-01-2000 on 03:57 p.m. |
| Dear All: I have attempted without much success to obtain additional information concerning the Pap rug. I have spoken with Cornelia Montgomery who reports that the rug is not at either of the galleries. The best information they currently have is that the rug has "very thick ivory" warps, "probably 3 ply" but they are uncertain about the number of plys. The handle is thick and floppy. Peter Pap and Cornelia both feel the rug is probably Shahsavan although there does not appear to be any cotton in the wefts. Regarding the apricot color in the group, Cornelia agrees that these rugs are most probably not Kurdish and commented that she has seen this apricot color in other rugs that are more clearly Shahsavan and from the southern Caucasus. Based on this information, limited as it is, we should probably modify the Group 1 listing for the Pap rug and prepared by Daniel in this thread to read "probably three ply ivory warps." -Michael |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Deschuyteneer Daniel |
| Date | : | 09-03-2000 on 04:26 p.m. |
| Dear all, I looked carefully to the closeup of the border of the Rippon Boswell rug posted by Guido. I think its impossible to say if there is any specific end finish. Sorry but I can’t help. Daniel |
| Subject | : | RE:Shared design and common attribution |
| Author | : | Deschuyteneer Daniel |
| Date | : | 09-06-2000 on 12:38 p.m. |
| Subject | : | RE:Similar handle |
| Author | : | Wendel Swan |
| Date | : | 09-06-2000 on 03:21 p.m. |
| Dear Daniel, You have asked WHY rugs having exactly the same structure can have a different handle? I trust that you aren't basing your question on my observations. If you review my post above, you will see that I said of the Emry, Bozoglu, Pap and Wertime rugs: "it would be very difficult to group them structurally or by handle or by tactile qualities." As to the Memling gul design group I once looked at, I said: "I was overwhelmed by the variations in the weave." Except for what may physically happen to a rug after it leaves the loom (i.e., wear, washings, exposure to chemicals and sun, climate, etc.), there shouldn't be a noticeable difference in the handle if the structure is exactly the same. The problem is that structure is never EXACTLY the same, although it can come close in tightly controlled workshops. Handle is dependent upon too many different factors of materials and execution for me to provide an explanation. Sometimes I'm surprised at how similar rugs from a given region look and feel; other times I'm surprised by the dissimilarity. My recollection is that John Howe tried to correlate handle and structure at one time. While I can't speak to his results, I'm sure it was a daunting task. Wendel |