Monday 26 March 2012

Slip-up at Christie`s?

Christie`s online catalogue seems to have used the TIEM Seljuk carpet,which is more or less complete,to illustrate the fragment currently on offer.

1-Christies 24.4.12,lot 100 

2-after Sahin

The actual fragment appeared on the Internet:


(where it could be insouciantly admired)


Hopefully the error has not been repeated in the printed catalogue.

Sunday 25 March 2012

It`s a Pinwheel,Carlo

A call for information shall not go unanswered.Carlo Kocman`s interesting rug is a member of a small group of prayer rugs whose design is based on the "Swastika" or "Pinwheel" kazak.The pinwheel group is a large pack of rugs which can be divided into two stacks:those with minor ornaments in the field,and those without.

1-Bausback 1983-41

2-Edelman 25.4.1981-187
The ornamented type is more common.A standard medallion -and-hooks border is shared between both groups,and there are a few border variations.Otherwise they are interchangeable,and standards of excellence rely purely on harmony and proportion.Most classic types of 19th century caucasian rugs are available in a prayer format,although a Star Kazak version has yet to be sighted.There are two types of Pinwheel prayer rugs.In the first,the original form with its ramshorn curling leaf and sloping beams is still clearly derived from the original main carpets;in the second group the beams have disappeared,and the swastika forms have been condensed into isolated free standing figures-this is the "plinth-ing"common in late 19th century Caucasian rugs.

Group A:

3-Nagels 9.10.1976-224

Writing about this in 1981,Ian Bennett could not remember having seen another similar example.Optimistically dated 1122.

4-Bukowski Zurich

A last piece with the original form was published by George Gilmore:

5-Hali 53-186

Group B:
6-Teppich Engelhardt-now Ralph Kaffel

The most enigmatic of all.Kaffel rightly remarks that the ornaments have been borrowed from Borchalou carpets,not Fachralo(the two are often confused)Dated 1829.

7-Kendrick & Tattersall nr.138

The first publication of such a piece(1922)

8-Christies October 1993-441

The third piece with this border.

9-Bausback 1978-177
The Bausback piece is similar to the Kocman.
10-The Kocman rug

The swastikas are in this author`s opinion a faint reminiscence of the original dragon forms seen on 17th century Caucasian carpets.

Saturday 24 March 2012

Some Classic Carpets

Amongst the many fascinating pieces uncovered by rugrabbit during its recent foray into the world`s museums,three items are of particular interest.


The MET`s Ballard 22.100.74

First published in the catalogue to an exhibition of carpets at the Cleveland Museum in 1919,and later by Maurice Dimand in a controversial article(The Art Bulletin 1923)this large medallion carpet from the Ballard Collection has had a chequered career of changing attributions.Starting life as 15th century Persian,it mutated to late 16th-early 17th century Asia Minor,before ending in the Met`s carpet catalogue of 1973(page277-8)as late 17th-early 18th century,declared as Caucasian.It formed part of a spirited rebuff from Arthur Pope in 1924("Research Methods in Muhammadan Art")Pope took Dimand to task for his over-reliance on dating Persian carpets based on Miniature paintings.The Ballard  piece is in fact(according to Dimand and the 1973 catalogue entry)Turkish knotted,although this was no reason for Pope to rule out its North West Persian origins("inasmuch as most Persian rugs and practically every Caucasus rug are woven with a Ghiordes knot")This is clearly false when applied to the Tabriz Medallion carpets,all of which are Persian knotted.However,according to a recent analysis by P.R.J Ford,the carpet IS Persian knotted.

Cleveland 1920

Tuesday 13 March 2012

186 Fair Street

Charles Grant Ellis`s house is up for sale.

This post rounded off with a personal letter and its subject.

Friday 9 March 2012


This piece was offered for sale at Lefevre`s in April 1984,and again in June 1985.It is one of a small group of five with rounded central medallions,the other three having been published by Batari,Douglas,and Ronnie Newman.
 Click for more photos

1-Lefevre June 1985
There are two major types of 19th century Kula rugs,the Demirci(medallion)type,and the Komürcü(prayer type)These two are often confused in the rug literature.A third cluster features large palmettes and is used mainly on long rugs.Within the Demirci there are again three subgroups:Group A,perhaps the oldest,is also closest to its Transylvanian forebears;Group B,perhaps the most common,with an elongated lozenge medallion.Both groups feature vases  as in Transylvanian carpets .Group C  carpets employ a rhomboid medallion flanked above and below by tulip and carnation forms,again betraying their Cairene origins.All three major groups invariably  employ a yellow-ground border with an undulating meander and 3-flower design,which also occurs in East Turkestan carpets.
2-Group A

3-Group B

4-Group C

5-Large Palmette group

From the Manoyan Group,the Douglas carpet best demonstrates the Cairene connection,when placed side by side with a carpet from the Textile Museum.The Douglas piece has rounded corner medallions in the "etcetera" style.The seahorse appendages connected to the central medallion(via Transylvania)are actually the Saz leaves of the Egyptian carpet`s field,transformed and attached.(the leaves are quite distinct on the large palmette group)The colour schemes are not unalike,with much white,pale yellow,and strong primaries.Even the overall shape in this case is similar.



Technically the Kula(Demirci)carpets have a depressed warp,and use a curious wefting technique:rows of single weft alternate with two shoots,producing a loose and floppy weave.This is not unlike the "soft"vase technique of three wefts employed in the Khotan carpets(with a similar handle)One might describe the Kula technique as a soft vase(3 weft) weave with a row of knots inserted between the three wefts.Despite the use of Turkish knotting(as opposed to the Khotan Persian knot)the backs are quite similar.



Prayer rugs from the Kula area have a long history,perhaps due to the proximity of Ghiordes.There are numerous reproductions of the simpler type of 18th-19th century prayer rugs,usually in black and white,and instantly forgettable.Presumably large numbers were faked.The "Komürcu" type of Kula carpet simply refers to any rug with a brown ground,but has come to be associated with the prayer rugs,which are in fact a mixture of the Bellini Style and the highest quality Transylvanian carpets,albeit for the modest purse.The best types are characterised by an upper panel of calligraphic inscriptions,which in later(or simpler)examples either disappears or is perfunctory.A salient example,now in the Keir Collection,retains its re-entry modus and Siebenburger Cartouche borders.

12-Keir Collection

13-Nagels 1992
The early 19th century piece shown here has lost its closed re-entry character but still sports a pidgin calligraphy.The border is typical,and interestingly employs a ragged leaf and vase carpet palmette as seen in a group of Indo-Persian Red-Ground Floral carpets.In the late 19th century this cluster squeezes a vase between the lower ex-mihrab,with a large sprig of flowers in the field.One last Transylvanian reminiscence.

This may also be a symbolic depiction of a grave,adorned with flowers.A famous group of Kula carpets,the Mezzarlik, represents a graveyard with cypress trees.

One last connection to East Turkestan can be found in a small group of prayer rugs,which also employ a calligraphic panel above the mihrab.In terms of this discussion a long shot,but who knows which ideas were transferred back and forth along the Pan-Turkic Highway?
 more East Turkestan prayer rugs

Saturday 3 March 2012


Treasure chests sometimes wash up on the shores of the Internet.Two carpet groups,which appeared in the last few years,were amongst the driftwood.



Many people could mistake this for a Tekke or Arabatchi ensi.True,the insikush look wilted,and there are no bovrek fingers clutching the central spine.There are no camel trains or opulent curled leaf borders.But the panelled inner borders have been taken from the Arabatchi,used on their ensis as well as on a small proportion of chovals.However,the border has been further simplified,when compared to the original.In this group it is usually an hour-glass form with diamond appendages.Sometimes the soynak border encompasses the entire field,as in this example.This is never the case with Tekke or Arabatchi ensis,where the elem play a crucial role.Here the elems are always of the standard Tekke-type.

A subgroup utilises a shield design,which is basically a mirrored ensi or keyhole form.


The chroma is mostly pastel,although some have stronger colours.They employ the simplest of natural dyes,and an attempt has often been made to mimic the madder exhaust pinks which feature on arabatchi ensis.Green and yellow are unknown.Along with a rough Baluch-type weave,this imparts them a dessicated look.

None appear older than 1900,with a taq around 1940.Their first appearance seems to have been at Mangisch`s Auction on 16 September 1989(lot 1127,labelled Kirghiz),although there may be others lurking in older publications.


Is it is possible that we are looking at the remnants of a long-lost tribe?`s more likely that a group of dealers commisioned them to replenish stocks of diminishing Turkmen ensis,which were in demand as export items.They are all quite varied and skillfully made,which would indicate a routine work environment where small rugs could be run up in a short time,with the simplest of materials and labour.They have a scurrilous quality,like pulling the leg of the Turkmen,for whom the Baluchis often worked as shepherds,and who treated them like dogs.


Roughly contemporary,and also from N.E Persia,is a group of bagfaces with an 8 pointed star,or Khatam,as central Logo.Persian dealers call them "Mumluke" 


They have also been described as Afshar,Nishapour,Qashgai,Teleghan...the consensus seems to fall to the Afshar,but those groups living in Khorasan.The earliest appearance is in Engelhardt,1976,pl.560.Again,more examples may come to light in older publications.

5-Engelhardt 1976
The quaint misnomer"Mumluke"is not entirely inaccurate as the design appears on at least two very early Mamluke carpets(both in the V & A).These are the Mounsey carpet,where it features centrally,and the famous "anatolian " style fragment,employed as a corner piece.The khatam,or"seal of the prophet,Seal of Solomon,etc"is an ancient design found frequently in Mamluke marquetry,and even earlier on Nasrid textiles.One carpet maker thought so highly of it that he created a legend around this one design-the so-called "chessboard" rugs were born.


The pieces are uniformly well made with glossy wool,tight weave,and excellent,varied colouration.A faded red has been observed,probably dyed with Bakkam.Knotted middle pieces with an allover pattern are common,as well as Turkish rosettes placed in all four corners.The Khatam is often placed in a larger medallion outlined in white,giving them a typical "Baluchi" look.

It`s unlikely that they are a group of mis-placed Mamlukes.Like the Ensi Group,they are probably more the products of commerce than of custom.Design- poaching has been a constant in carpet production through the ages.In this sense dealers and pickers are more like archaeologists,digging up the remains of what their predecessors once instigated.

The two carpet types described here can hardly be found in the carpet literature.Without the Internet they would not have surfaced as a group,being too insignificant for the major auction houses.It is noteworthy that unknown types of carpet are still emerging.Let us hope this trend continues.