Tree designs of carpets
Tree designs of carpets
Wc encounter trees, or large tree-like plants, with or without flowers, in two principal forms in carpets: as subsidiary elements in pictorial designs of many different kinds; and as the main motif around which a whole design is constructed.
The classification of tree designs into a separate section in this book poses some problems of cross-referencing, since trees occur in so many other forms. Apart from the designs shown in the following pages, the reader should also bear in mind the illustrations from several other sections: (i) pictorial designs, because while some ‘tree’ designs represent no particular tree, others present quite realistic pictures; (ii) animal and bird designs, because the trees often have animals and birds around them; (iii) prayer designs, because trees are just the right shape to fit into the ground of a prayer rug; (iv) vase designs, because the distinction between a single vase of flowers and a single tree is often lost and the two merge into one design; (v) ‘garden’ designs.
237 THE TREE OF LIFE
Frequent references will be found in books on carpets to the ‘tree-of-life’ design. We cannot today attach much importance to the ancient symbolism which is associated with many carpet designs, but there seems little doubt that at one time religious or mythological ideas had a great influence on carpet art. The reader who wishes to delve further into this subject should look at the Series of lectures by S. V. R. Cammann printed in the Textile Museum Journal (Washington, D.C.) in December 1972. Some of the connections suggested seem rather far-fetched, but these essays contain a great wealth of background information on Islamic and pre-Islamic oriental mythology which certainly adds to our appreciation of the culture from which our present-day carpets are derived.
Not only that: what of the creators of religious art? It is too easy to listen to Bach’s Mass in B Minor or Beethoven’s Missa Solemnis — for many people among the have been ardent Christians. They may or may not have been: the question is irrelevant to the greatness of their works, nor does one need to be a Christian to appreciate their greatness. Musicians are guided by musical impulses and obey musical rules, whether they are composing ‘religious’ music or operas. If Beethoven fought the angel for the key to the fugal Credo in the Missa Solemnis (as Wendell
Kretzschmar tells us in Thomas Mann’s Dr Faustus), it was a struggle for the technical means to carry his universal humanism to all who hear the work. The same principles apply in carpet design. However, it must be remembered that in folk art there are no creators – the designs are representative of a cultural background and, as such, reflect the symbolism with which it is imbued. We shall therefore keep coming back to the cultural background in considering the extra layers of symbolic meaning underlying certain motifs.
From the apple tree of the Garden of Eden to the Druids’ groves of Celtic Britain, we can find countless examples of trees with religious associations. Fig. 237 illustrates the sacred tree of Assyria flanked by two
priests in symbolical bird form. This stone carving comes from the palace of Ashurnasirpal at Nineveh (ninth century BC). The ‘tree’ is constructed from the same palmette motifs as are seen in the stone ‘carpet’ illustrated in fig. 57, a further reminder of the particular importance of the palm-tree in western and central Asia
(see also p. 54). These stone panels often include a symbolic representation of the national deity Assur, in a form similar to that frequently found in connection with Ahura Mazda, the sun-god of the Parsee religion, Zoroastrianism, which was the official faith of pre-Islamic Iran. Traces of the symbolism of these ancient religions may be found in carpet design to this day and some scholars, indeed, regard the bird symbols as a primary source of carpet motifs.
Now the second city of the Soviet central Asian republic of Uzbekistan, Samarkand was once the gateway to China. It was one of the principal cities, if not the very pivot, of the Silk Route, the main commercial artery of Asia until the Portuguese discovered the sea route around the Cape of Good Hope in the late fifteenth century.
It thus became the western outlet for the carpets of the oases of the Tarim Basin (now the Chinese province of Sinkiang, but known to carpet men everywhere as East Turkestan). The principal production centres were Kashgar, Khotan and Yarkand; their carpets were known simply as Samarkands. Today they are collector’s items, but they are not so rare as to be priceless, especially in the large long and narrow sizes (e.g. 500 X 200 cm; 16′ 6″ X 6′ 6″) which are typical. All their designs are strongly individual and quite unmistakable. The ‘pomegranate tree’
illustrated here was common in all three centres; it is most often found in red on a dark-blue ground. The weave is as distinctive as the design, a hard, dry-feeling structure, fairly coarse, with a markedly ridged back and a thin, close-cropped pile.
Hans Bidder’s excellent book on the carpets of Eastern Turkestan draws attention to the antiquity of this design, which may well be linked to the symbolic tree of the ancient religion of Assyria (fig. 237). The earliest settlers of the Tarim Basin were Indo-Europeans; marauding Western Turks entered the region in the sixth century, but they were expelled by the T’ang Dynasty Chinese. After the latter’s demise the area was incorporated into the great Tibetan empire of the eighth century; the Uighurs, or Eastern Turks, occupied it by the tenth; it fell to the
Mongol empires of the thirteenth and fifteenth centuries and returned again to Chinese control in 1757/8. But the region was never entirely assimilated by any of these outside powers, so that the essential Indo-European cultural strain remained.
Although East Turkestan is geographically isolated from the main areas of west Asiatic carpet production, its roots are firmly set in the traditions that inspire the designs of Persia and Turkey. Like the tree motif itself, the pomegranate also had important symbolic significance for the ancient world. Its use as a fertility symbol
can be traced back to ancient Sumer; the Muslims carried both the fruit and its symbolism throughout the Old World, from Spain to China. It is possible that the many scaled motif looking like a pine-cone in the Pazyryk carpet (fig. 56) and the Nineveh panel (fig. 57) may be meant to represent a pomegranate. Note also the presence of the motif in the hand of the eagle-headed figure in fig. 237.
Carpets are still made in the Tarim Basin – a low-quality controlled production sold by the Chinese government simply as Sinkiangs. Some of the designs, such as the ‘pomegranate tree’ illustrated here, are modelled quite closely on Samarkand originals. Others (see, for example, fig. 450) interpret the geometric East Turkestan
style rather more loosely, but in neither case can they be compared with either the structure or colouring of the antique originals. The new production is made in different grades, some loose in weave and using very thin warps and wefts, giving a rather floppy construction, while other pieces have more body and a much heavier
handle. The main features in their favour are the remarkably low price and the use of a very appealing soft yarn. A wide range of rug, runner and carpet sizes is made in a variety of ground shades, creams and buffs being most common.
A reminder from West Turkestan of the universal importance of tree forms in carpet motifs is contained in the detail reproduced here from the Chodor enssi shown in fig. 354. The panel design itself is discussed in detail on p. 153, but from a cursory glance at the panels one is apt to miss the fact that the ‘garden’ is full of
trees, in this case trees with botehs. Another detail from the same rug is shown in fig. 144, and from these details the reader will realize how much more there is to a good Turkoman rug than at first meets the eye and that the restraint described on p. 176 hides a wealth of imaginative and expressive detail in no way inferior to the
more showy rugs of Persia.
Tree designs crop up time and again in nomadic weaves. Their form is never standardized. Each one is different as
each tree is different. The Gebbeh rug illustrated here is woven in natural undyed wools, in different shades of white, yellow and brown. This is rather unusual. The ‘natural’ Gebbehs are usually in the bold designs shown in figs. 478-80; tree designs are more often found in the ‘coloured’ Gebbehs (see p. 214), especially on red
grounds. Of the latter, fig. 113 (with or without botehs), is a good example of the style one often finds.
The Lurs, one of the oldest tribes in Iran, are spread over the whole western side of the country, from Lar in the south, near the Persian Gulf, through the Zagros mountains and the province of Luristan right up into the Caucasus, to the river Pambak in the Kazak region just south of Tbilisi (Tiflis), and also into Turkey.
The presence of the ‘latch hook’ in a design (see fig. 475) often indicates Luri influence in a particular area, but where the Lurs have merged with the local population their carpets are generally known by the local name: Lori Pambaks, for example, are thought of as Kazaks rather than Luri rugs. However, even if one limits the name Luri to the products of the Zagros mountains area, where the Lurs predominate, there is still a big range of types, all of which may include tree designs.
The Gebbeh of fig. 113 is probably a Luri; the other main types are Ilam, Khorramabad, Behbchan, Owlad, Yalameh and Nasrabad. The principal features of these origins are detailed in the section devoted to bold geometric designs, beginning on p. 208. Here we have included only one tree design. This is a small carpet from the area north and west of Shiraz which produces the fairly standardized category of goods known simply as Shiraz Luris. They are fairly easy to recognize from several salient features: the size, which never varies much from 250 X 150 cm (8′ 3″ X 5′ 0″), the goat-hair or woollen warps, the firm but coarse ribbed structure, the shaggy pile and the overall light tenor of the colourings. From the wide range of designs made other examples are illustrated in figs. 453, 477 (see also index).
Halfway between Shiraz and Sirjand lies the village of Niriz, which is the production centre for a rather low-grade type of Shiraz rug. Shiraz is really an incorrect description because the Niriz production is influenced by many different factors.
The bulk of the production is in a medallion-plain version of the Qashqai hebatlu design (fig. 595), but unlike these rugs from the Shiraz area, the warps and wefts of Niriz goods are of cotton, which makes the weave very like that of the cheaper Afshar types, such as Sirjand. Some of the more unusual designs also show strong Afshar influence, and the colourings, too, are more characteristic of Sirjand than Shiraz – the red, in particular, is brighter than the Shiraz red. Yet another influence certainly comes from the Arab nomads living in the countryside to the north.
Within all these influences, however, Niriz has a distinct style of its own. The tree design illustrated here is a good example: while certain elements are strongly reminiscent of Afshar work the general style is clearly recognizable as Niriz. Note, too, the wealth of realistic detail within the stylized overall conception.
As with the boteh design (see figs. 151, 152), the Afshars have dozens of versions of the tree motif, some forming just a small element in a larger design (as in fig. 244), others dominating the whole rug (as in figs. 245 and 246). Figs. 244 and 246 both illustrate how close is the connection between tree motifs and the concept of the
vase design, while fig. 245 – with several different trees, birds, animals, botehs, the two medallions, sundry stylized flowers and purely abstract rosettes – shows how difficult it is to classify the richly decorative products of the tribal imagination. This vision of Paradise, this dream of a rich ly watered garden, of superabundant nature, epitomizes the universality of the appeal of the nomad rug. In view of the wealth
of the designs the Afshars use, the layman’s design knowledge will not be enough on its own to enable him to recognize a particular place of origin. T o take an obvious example, what is it that permits us to state categorically that the tree in fig. 241 is a Luri Gebbeh tree, the one in fig. 243 is a Niriz tree, while fig. 246 illustrates a Sirjand Afshar tree? The answer is that this is one of the cases where recourse must be had to the other more subtle clues mentioned in the Introduction. With Afshari rugs, the first clue is an additional design element: their makers love to include birds, animals and little human figures scattered all over the place, sometimes quite realistically drawn, but more often like curious little puppets. The Afshars are not the only weavers to do this but it is certainly a very characteristic element which helps in identifying their goods. The other points are: colour, warp and weft materials, sizes, wool, weave. It takes experience to recognize subtle variations in these features, but the principal differences are shown in the table opposite.
Trees (and flowers) often feature in Kurdish rugs. The piece illustrated in fig. 250 is typical of a design made throughout Kurdistan, as well as Luristan. Its origin goes back hundreds of years, but such pieces arc still made with all the individuality and freshness found in many examples of the design seen in museums and antique
collections. Fig. 249 dates from the eighteenth century and is particularly valuable for the unmistakable connection it reveals with both the Joshaqan design (see figs. 7, 8) and with vase carpets (e.g. fig. 180). Here is a perfect example of the way nomad weavers have borrowed the ideas of the great manufactories, set their own
imprint upon them and developed an independent tradition from them. The rug from the Zenjan area (fig. 248) shows a further development of the idea in a village manufactured style which is typical of much of rural Persia. This is an exceptionally good example of the Kurdish-influenced production described on p. 91. Note here, as there, a certain stiffness in the design which the weaver of the rug shown in fig. 250 has succeeded in avoiding completely.
The evolution of the Bakhtiari panel design from eighteenth-century garden carpets is illustrated in figs. 327-339. Within these ‘gardens’ tree motifs are widely used. Sometimes the trees are quite naturalistic and easily recognized; in other carpets, such as figs. 255 and 256, the motif is reduced to an indeterminate stylization of
tree or bush forms.
Another distinctive tree design, very weird and always just like the present example, is made in the Kharaghan district north of Hamadan, and also by the Shah Savan tribe near Saveh. For the difference between Shah Savan and Kharaghan, see fig. 536. The design illustrated is most often found on a cream ground in zaronim size, but other sizes and colours do exist.
260- 265 ISFAHAN, TABRIZ, KERMAN, KASHAN, NAIN, QUM
All the towns of Iran that use a fine weave produce designs with trees as the principal motif within a prayer-rug layout. They are not prayer rugs in the proper sense but are produced commercially in a range of sizes from pushti up to about 6 m2 (70 ft2) in area, the most common size being the dozar. In cases of doubt the weave and colouring are the only sure guides to the origin, but all the pieces shown in figs. 260—265 are recognizable to experts on the basis of their design style alone.
Indeed, all six origins are instantly identifiable purely on the basis of the pieces illustrated here; this is remarkable proof of the independent identity preserved by all Persian production areas, whatever designs they use.
The Kashan of fig. 263 is a silk carpet of a type made in considerable quantities (in both silk and wool) between the two World Wars, although this piece, with its rigid central axis, may be somewhat older. Perhaps the most important feature of the Kashan style is the perfect balance of its proportions: the Kerman of fig. 262
includes no more detail than the Kashan but looks crowded by comparison. Other Kashans of the same era are illustrated in several sections of this book. The strength of the Kerman carpet design is its exuberance. The tree designs are derived from ancient carpets such as figs. 327 or 696; but here the Kerman designer has gone much
further than his illustrious predecessors in the abundance of the floral elaboration: every inch of both ground and border must be crammed with richly coloured blossoms. This is a characteristic feature which may be observed in other Kerman designs (see fig. 309 and index).
The most striking stylistic feature of the Tabriz tree rug (fig. 261) is its use of perspective. This is a rare element in Islamic art: the restriction of all representation to two-dimensional forms constitutes a widely accepted compromise between the artists’ search for expression and the religious law’s ban on the portrayal of
living things. The use of perspective here is a facet of the Tabriz tradition of universal adaptability of design, linking fig. 261 with fig. 361 and many others like it.
Nain rugs (fig. 262), the finest of all modern Persians, are made in hundreds of villages in the area around Nain. Some of the best pieces come from places as far afield as Biabanak, over 200 miles out into the desert. Isfahan used to have a small production equally fine (as well as a large output in medium grades), but typically Isfahan in style and clearly distinguishable from Nain. Now, however, the style of Isfahan has become so like that of Nain that even experts sometimes cannot tell them apart. Nain prayer designs are not very common but they are occasionally found: for a rug which is so fine and expensive that many people prefer to hang it on the wall, the prayer-with-tree motif is an obvious choice.
Similar ideas as the basis of a prayer design are used in the superfine Isfahan shown in fig. 260. One clue to the difference between Nain and Isfahan is the use of silk. Isfahans mostly have an all-wool pile but use silk for the warps.
Note that figs. 261-3 all employ the same border design, but all three show evident stylistic differences which are as helpful clues to the origin as are the differences in the tree motifs themselves. A similar style (not illustrated) is also found in the northern half of Khorassan province, in rugs of various sizes (dozars are the most common). They are usually village products, inferior in weave to the Meshed and Kashmar carpets which they otherwise resemble. They are much coarser than the other rugs shown in this group and are also easily distinguished by the bluish red and the very heavy dark blue which are typical of the region.
As has already been noted, Qum makes a panel design based on the Bakhtiari
‘garden’ style. In this various tree motifs are used, but the same comments apply as in the case of the boteh motif (see fig. 159). Beyond this there is a huge range of designs made in Qum. Tree designs are common, with or without a prayer mihrab.Fig. 265 shows a typical Qum version of a design also used in Kashan and Isfahan.
There arc several stylistic features which indicate that this piece is from Qum (the shape of the border motifs, for example, and the colours, especially the red), but there is a further clue in that parts of the motifs in the woollen ground are woven in silk. Although this practice is not unknown elsewhere, it is a particularly common
feature of the modern Qum production. Qum also employs a design including groups of trees arranged in a layout similar to that seen in the present example.
Many ground shades are used including several – such as light blue – which are rarely found elsewhere.
266, 267 THE SHAH JAHAN STYLE
The great Moghul emperors ofTndia, Jahangir and Shah Jahan (who built the Taj Mahal), established a very distinctive style in art.
Many examples of this art will be found in carpet collections in museums. In recent times, several attempts have been made to revive this style in India as a means of giving a more recognizably Indian flavour to the huge output which has blossomed in Kashmir, Gwalior, Agra itself and the Benares district. The rug shown in fig. 267 illustrates one such attempt; the original which inspired it is in the Metropolitan Museum of Art, New York. The rug shown, made in Khamariah, near Benares, is not a copy but has many features in common, notably the naturalistic treatment of the chrysanthemum plant (not strictly a tree design, of course!) and the very un- Persian feature of not having outlines to separate the individual colours in the leaves.
The similar treatment of flowers in the wall panel in Lahore, shown in fig. 266, should also be compared with figs. 180 and 249; from this comparison one can see how strongly interdependent some designs are in widely separated parts of the Orient.
The student of carpet design often has cause to marvel at the ability of the peasant weavers of the Orient to create exciting and expressive forms out of the simplest of geometric motifs (see, for example, figs. 101 and 484). The Anatolian village of Do§emealti provides a further example of this: the key to its effectiveness is the
openness and bold simplicity of the layout and the elusive significance of the motifs, which at one moment resemble some form of tree of life, at others shady trees in a flowery park, and at others a purely abstract elaboration of’latch-hook’ and related motifs. The border design and the all-wool structure are sure indicators of the
Do§emcalti origin, but the main design itself is also quite commonly encountered.For general notes on Dosemealti see fig. 495.