<---previous article

"Prayer Rugs of the Timuri and Their Neighbors"

by Robert Pittenger



FOREWORD
This is an article reproduced from OCTS 7, a publication documenting the presentations and poster sessions at the 7th ICOC in Hamburg, Germany in 1993. Robert Pittenger is a relentless researcher who has documented much in terms of hard facts and figures that few other writers and researchers on the subject are willing to do or have done, myself included. This paper offers some interesting information regarding a group of rugs that are still available in the marketplace of today, of which there is little real understanding or a grounded appreciation.



Figure 1. Mrs. Benn describes buying a carpet from a Baluch weaver beside the Helmund in terms which can refer only to a pile rug and not a flat-weave in addition, she includes three photos of Seistan residents with rugs on the ground (but unfortunately without specific attributions


The blue-ground rugs made by Timuri, Baluch, or Aimaq tribes in the Iran-Afghan border area between and around Heart and Meshed have been collected for a century (1). But have only recently been studied. Dr. Alfred Janata, in his talk about the 4th ICOC (2), gave a good kick to previously held Baluch attributions by suggesting that many of these rugs were not made by Baluch sub-tribes, but mainly by Timuris. He illustrated a small number of mid-20th century rugs he had obtained in villages, often from the weaver's family. His ICOC talk and subsequent article in ORIENTAL CARPET AND TEXTILE STUDIES have been so influential that there is now a counter-trend which denies that Baluch any pile weaving at all. To correct this temporary misjudgement, one should consult material published by Mrs. Edith Benn, in her 1909 book AN OVERLAND TREK FROM INDIA. Mrs. Benn lived in Seisten with her husband, Major R.E. Benn, who was consul at Nusratabad (now called Zabol), on the banks of the Helmand river, from 1901-03. Mrs. Benn describes buying a carpet from a Baluch weaver beside the Helmund in terms which can refer only to a pile rug and not a flat-weave.(3) in addition, she includes three photos of Seistan residents with rugs on the ground (but unfortunately without specific attributions) (Fig.1). Mrs. Benn's account seems to give distinct proof that Baluches DID weave pile rugs, at least in Seistan. But the focus of this paper will be on rugs produced farther north.



Figure 31 from Bogolyubov/s Carpets of Central Asia. " a blue-ground main rug which he bought at Takhta-Bazar on the Russian-Afghan border near Murghab. Bongolubov says it is called “…tapis afghan de Guerat ((Herat)) de Teimour”;


For the past 150 years the Timuri tribes have lived at well-documented locations, primarily in eastern Iran, centered around Khaf, including the valleys of Khaf, Baharz, and Jam (4). Additional groups lived south of Nishapur (5), in the Turshiz valley(6), and also in Afghanistan south of Herat (7). There was a mass exodus of Timuris out of the Herat area in 1856, when they found themselves on the losing side of the battle(8), they then moved to the Khaf valleys, but since 1900, some have returned to the mountains north of Herat (9) and to Ghor, east of Herat. (10) In 1884-85, the Afghan Border Commission was sent to the area north of Herat to delineate the Afghan-Russian border. The staff included a number of British officers, especially surveyors and intelligence officers, who compiled and pblished a list of the Timuri sub-tribes (11), re-published in Adamec's Iranian Gazeteer (12), and later by Azadi. (Note that material in Adamec should be used with caution, since entries from 50 years apart are found side-by-side, without author or date!). To confuse the matter, another similar list of tribes was compiled and published by Alexander Finn (13).



Figure 30 from Bogolyubov's, Carpets of Central Asia. “…tapis de Bloudje )de Beloudjistan)"


Although members of the Border Commission later published copius geographical information and mentioned purchasing rugs in the Herat area (14), they published no specific information on the Timuri or Baluch rugs of 1883. Other 19th century travelers through ther area also published very little of use to us, with the exception of General A.A. Bongolubov. Plate 31 of his book TAPESTRIES DE L'ASIE CENTRALE shows a blue-ground main rug which he bought at Takhta-Bazar on the Russian-Afghan border near Murghab. Bongolubov says it is called “…tapis afghan de Guerat ((Herat)) de Teimour”; and “Quant au nom de Teimour, c'est le nom de tribu, qui les fabrique.” - Probably the dealer's attribution, but Bogolubov does distinguish between teimour in Plate 31 and “…tapis de Bloudje )de Beloudjistan)” in Plate 30. Thus the so-called Blue-Baluch rugs have in recent decades been assigned to Timuri tribes by dealers and collectors. The most reliable information on this group is still contained in the articles of Alfred Janata (15), who was in NW Afghanistan in the 1960s, and of Dr. Dietrich Wegner, who was in NE Iran in the 1950s. Wegner's long article from TRIBUS (16), translated and published in Oriental Rug Review, is a gold mine of information on early 20th century production, but one must use caution in applying his information to 19th century rugs. Work has now begun (17) to isolate groups of so-called Baluch rugs, based on rug structure, and to a lesser extent, on designs. This paper will divide the group of blue-ground prayer rugs into four smaller groups, which remain unnamed, in the hope that tribal names can be assigned later, once there is more proof.



Figure 2. A "stepped inscription band over the entrance to the main Iwan of the Buq-a of Shaikh al-Din at Taybad, just at the Afghan bourder on the road between Herat and Mashad, in the center of the area peopled by the Timuris"


From the approximately 170 blue-ground prayer rugs of the 19th and 20th centuries known to me, about 90 stand out as the oldest and best. Of these, about half have been published. All of them have similar basic structure, using ivory warps, wefts of various browns, and usually two-cord selvages. Their handle is loose to floppy, generally without significant warp depression. They all use the asymmetrical knot open to the left can be no separation by type of knot in this group.

Their spandrel hand panels are at the very outside corners of the rugs, with the borders bending in around them. In some of them, the hand panels are connected at the top by a patterned bar. In later rugs, the hand panels move closer to the center, with fewer of the borders bending around them.

The square mihrab shape, typical of these and camel-ground prayer rugs (and also kazak prayer rugs), is also seen in a
similarly-shaped stepped inscription band over the entrance to the main Iwan of the Buq-a of Shaikh al-Din at Taybad, just at the Afghan bourder on the road between Herat and Mashad, in the center of the area peopled by the Timuris (Fig. 2)(18).

The oldest of these prayer rugs probably date to the 3rd quarter of the 19th century, or slightly earlier; middle period examples date to the last quarter of the 19th century, and later ones to the 20th century.

Of the ninety earliest or best rugs, about half are rugs in small groups of three to five similar examples, which we can only mention briefly and lump together here, saving them for future detailed study. Some have features such as borders and selvedges similar to the large group of camel-grounded tree-of-life prayer rugs, while others may be later versions from the three early groups which will be described now.


Figure 3
Group A
Group B
Group C
Doktor-i-Qazi
24%
33%
----
----
17%
47%
----
----
19%
60%
----
39%
53%
----
76%
----
21%
----
77%
----
SMALL GUARD BORDERS
75%
33%
60%
45%


I will begin then with Group B, which has the highest know count; this group accounts for 15 of the ninety best rugs (Fig. 4). The knot counts range from 70 to 140 ksi (1100 to 2150 k/dm squared); the average for the group is 110 ksi (1700 k/dm squared), which is rather high for Baluch-type rugs. The ones I have examined are also more tightly woven than other groups.

One of these, formerly in the Ernest Roberts collection (19), is dated to the equivalent of 1869. The field patterns are usually a more complicated version of the tree-of-life pattern used in the camel ground prayer rugs. There is frequent use of small white borders. Main border patterns vary considerably, but a kotchak border and a simple triangle-vine border are used in more than half of the rugs (Fig.3). In general, the border patterns used in the blue-ground prayer rugs are quite different group from those used in the camel-ground prayer rugs, and also quite a different group from those used in the camel-ground prayer rugs, and also quite different from most of the borders used in the rugs in Azadi's book of Baluch-type rugs (20).

The selvedges of this group are usually two-cord, with wrapping in brown wool or goat hair, which is seen less frequently in the
two groups which follow. End finishes vary, the most common one alternating narrow and wide strips of weft-floated repeat designs. This group tends to use much more brown in the pile than other blue-ground types. If any of the blue-ground prayer rugs are made by Baluchi weavers, this may be the group, but I can offer no proof. Some may be made by Hazaras (21), Jamshidis or other Aimaq tribes.

The next group (12 of the 90 rugs) is a familiar one; the so-called Dokhtar-I Qazi prayer rugs; they are included here since they are structurally vary close to the other groups, and were probably woven in the same areas (22). Their knot counts range from 63 to 126 ksi (970 to 1950 k/dm squared), averaging 90 ksi (1400 k/dm squared).

The earliest examples of the type nearly always use the same four borders in the same order. Selvedges are two-cords, usually covered with red and blue wool in a checkerboard pattern. End finishes are usually plain flatweave in narrow stripes of brown, blue and red; some have a thin row of weft-float (nearly always the same pattern).



Figure 4.
An exmaple of Pittinger's Group B Baluch prayer rug


At this point, it is not clear who wove these rugs. Dr. Dietrich Wegner was the first to use the name Dokhtar-i Ghazi in print (23) for a Baluch sub-tribe in the region north of Herat; more recently these rugs have been attributed to Timuri weavers. The translation of the name of Dokhtar-i Ghazi or Qazi to “daughter of the war hero”, or “daughter of the judge” as the name of the Baluch or Timuri sub-tribe seems unlikely, since according to . Dr..Janata, no sub-tribe in this patriarchial area of the world would name themselves after a woman (24). It seems most likely that a term Dokhtar-i-Qazi is merely a trade neame for the pattern of allover little blossoms or shrubs. Similar patterns were extremely common in 19th century Persian fabrics for clothing (25), and used in pile rugs from other areas of Persia (26).




Figure 5. Pittinger's Group A type. "They are attributed to the Timuris by most dealers and collectors, although other than the reports of Gen. Bogolubov and Dr. Wegner, there is no specific proff so far. These prayer rugs tend to be nearly square in format, and most likely date to circa 1850 to 1880"


Dr. J.E.T. Aichison, in the same report, also included a list of 28 natural dyes used in the Herat and border areas (29). Twelve of these dyes have been discussed by Bruggemann & Bohmer as in use in Anatolia (30), and twelve by Konieczny as in use in Pakistna (31). Some of the others may be local to the Herat-Mashad area, including three red dyes other than madder and cochineal (32). Here is good material for further research: once these red dyes are obtained and wool is dyed, pile colors formerly considered to be obtained from synthetic dyes may prove to be of natural origin, giving impetus for dating rugs which use these colors to pre-synthetic decades.

The Dokhtar-i-Qazi pattern was used in the late 19th century by other tribes in the area, for rugs with other borders, colors, and weaves (33)
The most noteworthy are the so-called Barawi-Baluch (34) rugs, which were most likely woven by Brahui-speaking people in the area between Herat and Farah (35).

Leaving the best for last, the final group to be discussed (20 of the 90 rugs) contains some of the oldest and best rugs; these I will provisionally call Group A (Fig. 5). They are attributed to the Timuris by most dealers and collectors, although other than the reports of Gen. Bogolubov and Dr. Wegner, there is no specific proff so far. These prayer rugs tend to be nearly square in format, and most likely date to circa 1850 to 1880. Knot counts on this group vary from 60 to 90 ksi (920 to 1525 k/dm squared), average 76 ksi (1180 k/dm squared), and the handle is loose.



Figure 6. A type of so-called Timuri group prayer rug with field ornaments similar to those found in the Doktor-i-Qazi types, but with a larger, bolder articulation of these classic shrub forms associaed with that group of rugs. (Plate 13, Belouch Prayer Rugs, Adraskand)



The field pattern of this group of rugs have repeating motifs, usually floating on the blue ground, without trellis division. There are six types of patterns:
1. Little trees of life with four leaves on each trunk (36). In later examples, these small trees become longer, turning into poles with the leaves often unattached.
2. Khaf guls (38) or similar guls (Fig. 5)(39).
3. Shrubs or blossoms like those used on Dokhtar-i-Qazi prayer rugs, but larger (40) Fig. 6-.
4. A short stalk with three angled branches on each side, and a hooked pot at the bottom (41).
5. Overlapping octagons; the best one of this type was acquired
by Vistoria & Albert Museum, London, in 1883 (42).
6. Mixtures of the above (43).
These motifs are also seen on some of the oldest blue-ground bag faces.

Border pattern vary, with 25 patterns being used in the group. The leaf and vine border is used in over half of this group, making it the most common, and one of the identifying indicators of the group (44). Fig. 3 compares frequencies of the most commonly-used borders. Border patterns tend to be an indication of age in all of these rugs, since rugs woven after about 1890 tend to use a different group of borders.



Figure 7. Another variation of a so-called Timuri blue ground prayer rug with a pallete and border system that may place in in the Chakhansur region of w. Afghanistan. (Plate 9, Belouch Prayer Rugs, Adraskand)



The selvedges of this group are two cords of wool. Overcast with red and blue wool in a checkerboard pattern, with later rugs of the group using orange and blue checkerboard, and more recent ones orange and brown. Goat hair wrapping is not used. Although finished of many of these old rugs are gone, those which remain include twelve examples with very wide weft-float panels, done in elaborate patterns; others use stripes of plain flatweave alternating with thin weft-floated strips.

The weft colours of this group are alightly differently from the other groups, where the range of weft colours is rather evenly
spread from tan to dark brown in various types. In these earl rugs, however, lighter brown wefts are used twice as often, and very dark browns are used half as often, as in all the other types. Only two prayer rugs of this group have silk details.

To summarize, this group with the squarish format differs from other groups in having a slightly lower knot count, different borders and end finishes, different field patterns, no uses of goat hair for selvedges wrapping, wets of slightly lighter colour, and more blues and fewer browns in the pile.



Figure 8. A classic Doktor-i-Qazi prayer rug with the distinctive mihrab and shrub elements in the field. (Plate 14 Belouch Prayer Rugs, Adraskand)



After about 1890, characteristics of these separate groups of rugs begin to merge, resulting in new combinations, with many new border patterns not seen in the older rugs.

Of course the question remains, just which tribes made these various types of rugs: Timuri, Baluch, or various Chahar Aimaq or other Aimaq tribes in the area (45)? With the exception of Drs. Janata and Wegner, no on-the-spot observer has equated in print specific rugs to specific villages or sub-tribes (46). Thus we must conclude that all other published attributions are either those of dealers, to which we may pay some cautious attention, or those based on some studied guesswork. Assigning a work to a specific place or date is a valid art-historical technique, but it
should be done with some care when publishing rugs, with reasons for the attribution included; unfortunately, this has seldom been done with these rugs.

Persian dealers will say that rugs with the finest weaves are Persian. But when the Timuris moved en masse from Herat area to the valleys centered on Khaf in 1856 (47), did they suddenly begin to weave rugs with higher knot-counts? Not very likely. Dr. Wegner's article is the best information we are likely to obtain, but must still be used cautiously in attributing 19th century rugs. I suggest we continue to use the trade-name Timuri for these blue-ground prayer rugs, but with the constant reminder that we cannot be sure that they ARE Timuri.




Figure 9. An Afghan Baluch prayer rug, probably a Taimani weaving of Group C as defined by Pittenger. (Plate 11, Belouch Prayer Rugs, Adraskand)



NOTES

BSOAS=BULLETIN SCHOOL OF ORIENTAL AND ASIAN STUDIES, London
JRASB=JOURNAL OF THE ROYAL ASIATIC SOCIETY OF BENGAL, Calcutta
JRCAS=JOURNAL OF THE ROYAL CCENTRAL ASIAN SOCIETY, London
JRGS=JOURNAL OF THE ROYAL GEOGRAPHICAL SOCIETY, London
PRGS=PROCEEDINGS OF THE ROYAL GEOGRAPHICAL SOCIETY, London
ORR=ORIENTAL RUG REVIEW, Meredith, NH, USA



Original text and photos appeared in Oriental Carpets & Textile Studies. All text by Robert Pittenger © 1994
Additional colour photos from Belouch Prayer Rugs © 1983 Michael Craycraft / Anne Halley
No parts of this text or any photo may be re-produced, transmitted or copied by electronic means or otherwise without permission from the Robert Pittenger, Michael Craycraft / Anne Halley or myself.