Taking the country on either side of the Caucasus Mountains, for the study of textiles which are strictly speaking Caucasian, we find that the commercialism of the present day has established methods of using the entire output of the looms of the district, and of opening centres for the sale of rugs ancient and modern.
So absolutely, though, have the dwellers in this part of Asia adhered to tribal designs, that the Caucasian group of rug patterns stands to-day more easily recognised and authenticated than any other. Here we may study design not only in its purity, but as it shows evidence of Mongolian and Russian thought: Here we may stand in imagination with those whose speculations in regard to natural phenomena led them in earliest ages to make a visible sign of definite thought, and in the mingling of elemental motifs with those which show the influence of many generations of Mohammedan thought we find it possible to work our way through the centuries, proving, as we go, that which we believe to be true.
This we are particularly well able to do in old Caucasian fabrics, which resemble each other in general appearance but differ materially in detail, the warp and woof sometimes showing the difference between rugs which have close resemblance to each other in pattern.
A rug may be called ” Shirvan ” by its owner, though another specimen quite like it in design is known as ” Cabistan ” because the weave is finer ; and both of these may suggest Daghestan designs, with which we became familiar when first Oriental rugs found their way into this country. Kazak rugs, bearing well-known tribal designs that are sometimes claimed to be Shirvan and sometimes Daghestan, are distinguished from them by their heavier pile, and by evidence of more virile workmanship.
Without experience in handling rugs, the ability to note these differences is more quickly gained by close analysis of objects than in any other way. Given two antiques that look exactly alike, on examination of the warp of one we find pure wool, while in the other a thread of brown wool is twisted with two threads of cotton, and the knots vary sufficiently to number more to the square inch in the first specimen than in the second.
Noticing the weft of the finer rug, we find nothing but wool in the woof threads, while, in the other, cotton is twisted with the wool, and the surface of the rug presents a ruffled appearance. Although the patterns are alike, dealers and Orientals always distinguish between textiles that show these differences in weave ; and, while it is necessary for us finally to make these distinctions, during our preliminary study it is wise to hold to the broad classification and use the name ” Caucasian ” until the sub-divisions force themselves upon us.
Mohammedan prayer-rugs arc now made in all the well-known patterns of the district, but, while they show that the region has been conquered by the sword of Islam, they fail as yet to show any predominance of the art motifs of the Moslem’s faith. In this way the prayer-rugs of the Caucasus differ so materially from Turkish products that one but slightly trained soon distinguishes between them.
There are certain antiques which were made by tribes who have so stoutly resisted the sword of the followers of the Prophet, that they have fled from their mountain homes absolutely impoverished rather than render either allegiance or tribute. Such rugs are valuable to-day as bearing distinctive features of tribes that no longer exist as such. Ethnologists trace the migrations of these tribes of the western Caucasus Mountains, in two directions : southerly into Turkey, where, as nomads, they lead a wandering life, and northward into Europe beyond the Black Sea.
One product of the Caucasus is of so individual a nature that it proves the exception to all general laws, and that is the ” Soumac ” or ” Shemaka ” rug. When this style of carpet was first carried to Europe, it was called by Occidentals, who so often give English and French names defining process to products from afar, ” Cashmere,” because in their method of weaving these rugs resemble the shawls made in the valley of Cashmere in India, of pashm wool of the ” Cashmere goat.”
In designs, these rugs bear evidence of Caucasian origin, but differ absolutely from any other known style of weaving employed by rug-makers, as they have no raised surface ; the long ends left after weaving the warp and woof hang loosely from the back of the fabric, thus making it other in kind from the rugs and tapestries of the Orient.
At the present time the Caucasus district is well known to all interested in Oriental rugs, and under modern names its vast output is grouped. The three lakes Van, Urumiah, and Gokcha, and the cities and towns near them, bear names well known to the trade, but not so well known to the antiquarian, who divides his rugs into classes bearing very different names from those used to-day. Large factories are now established where carpets and rugs of all sorts are copied carefully and well,— so well that all art value is often worked out of them by the precision of present-day labour.
Comparison of the modern map with that of the sixteenth century will give a clear idea of the changes which time and government are sure to bring about. Places that once were reached only by caravan are now accessible by railroad, and every possible facility is provided for weavers to dispose of their rugs, while in a past not very remote the individual weavers were sought out by enterprising agents who secured family heirlooms and treasures that now serve the trade as -copies for reproduction.
In Caucasian fabrics geometric ornament stamps itself in set precision ; we might almost state that unless the pattern be geometric the rug examined cannot be Caucasian. This forces us to try to classify the details of geometric ornament very broadly at first, before endeavouring to detect the slight deviations in primitive pattern brought about by outside influences of late years.
In this way the different types in Caucasian fabrics will force themselves upon our notice, and before we are aware we shall discover that though there are a few features which seem to be used as common property, such as the eight-pointed star, the reciprocal border motifs, etc., among the people of different districts, the methods of handling these patterns, and the preparation of materials in each case, seem to show some individual touch which helps us to distinguish between rugs in which at first we discovered only strong likenesses.
Into eight divisions, and only eight, does it seem wise for our special purposes to group the rugs of the Caucasian district :
Daghestan, Derbend, Kabistan, Shirvan, Shemakha, Karabagh, Kazak, and Tchechen (Chichi or Tzitzi).
These are the names given to these eight styles, and whether the weavers are dwellers in the towns or wanderers in the mountains they are apt to adhere to certain tribal patterns which we soon learn to know as such.
Antique Daghestan rugs were made of wool, and were well made. They were marked off in divisions that seemed very different, when first we saw them, from the markings in other fabrics. Medallions and octagonal forms appealed greatly to our imaginations when we were first told that such designated not only where the host and his guests should sit, but in various ways had tribal significance.
These outline forms were later bereft of all meaning when changing customs multiplied them in numbers and distributed them in broadened and flattened lozenge and octagon shapes, surrounded by angular hooks and serrated edges, upon oblong rugs. In small rugs, of which many were prayer-rugs, a crossing of the entire field with a sort of trellis gave numberless diamond forms in which were generally to be found geometric tree designs.
In the borders an arrangement of angles and squares gave first one and then another set of decorative forms, a few of which have, strictly speaking, become more historic than others. In fact, in antique Daghestan rugs we find one of the fullest manifestations of the triangle that has been preserved for us in the art of Asia, and it is possible to break up any of the old designs of this region into its parts, which will always, even when presented in block, cube, or star form, present the right-angled triangle as base. This of course gives us the eight-pointed star in the ornament of that region, which differs from the zodiacal or twelve-sided figure of the Chaldeans, and from the ensoph of the Hebrews shown in the shield of David.
A strong Mongolian influence is felt in Daghestan patterns, and perhaps helps us more than anything else to recognize the designs of that district. Daghestan, it should be remembered, is a district or province, and one of the most northerly of any known to us as specially connected with rug-production.
Daghestan rugs should be studied architecturally, for the distribution of_ octagon and medallion forms was at first indicative of floor spacing, and in old Russian adaptations of Mongolian ideas, even in the construction of cathedrals, special attention was paid to the relation of one form to another.
The tent-roofs of Turkestan were often divided into eight sections ornamented with set patterns which were later copied upon the ceilings of tombs and the domes of cathedrals. The tomb has always preserved more accurately than anything else the absolute expression of thought in art. An individual who might borrow for the decoration of some meaningless article the patterns of neighbouring tribes would be very careful, in making a pall, to adhere to tribal patterns.
In the grave carpets of the Caucasus are found designs that were once the repositories of belief. In form these rugs are many of them like prayer-rugs, but are very long and narrow, with the niche at the extreme end of the attenuated panel.
Derbend rugs resemble those of Daghestan, but are distinguished from them by minor points unnecessary, for ordinary purposes, for us to investigate. It is only honest to say that of the weavings of this quarter learned experts give varying opinions, so that it is unwise to force the frail bark of the novice into too deep water.
It is universally asserted that in old Daghestans the warp and weft is more apt to be wool than is always found to be the case in Shirvan products, where, particularly in the weft threads, one or two strands of cotton will be twisted with one strand of wool in such a way as to cause the finished rug to pucker most objectionably. The sides of Daghestan rugs are generally overcast with coloured wools.
Kabistan (Kuban) rugs are full of interest as showing the attempt to render geometric ornament in a floral manner, and the weavings are so full of character and interest that we soon become accustomed, in handling them, to classify them correctly. The use of a border design which has gone through many vicissitudes since it started in a northern province of Persia, in the long ago, as a strictly floral rendering of the beautiful single pink, so well known as a Persian motif, finds in the Kabistan weavings a most interesting expression in geometric form ; and the water pattern, in which at first veritable fishes were found naturalistically portrayed, has been converted into a broad band decoration of stripes and dots.
This border often surrounds a field strewn with large effulgent stars which are quite different from other star forms in appearance. At the extreme centre of these forms is generally found a small elongated star arranged upon a diamond shield from which project lines which suggest crossed weapons. From this whole centre a burst of radiance fills out a large figure which, though bounded by broken outlines, suggests in the main a star form. These ” effulgent stars ” are so disposed upon the field of the rug as to bring different colours in such order as to produce diagonal stripings.
In fact it is a Kabistan feature to use diagonal ornamentation in both border and field patterns. In beautiful old specimens of these rugs the warp is of dark-brown wool of natural tones, and the knot yarn is of the softest quality ; the finished pile being so closely clipped as to render the Kabistan weavings the thinnest and most flexible of any of the Caucasus district. In old Kuban (Kabistan) weavings are treasured time-honoured designs, and very occasionally an antique specimen is found, in which the field, worked in very dark blue crossed with a tan lattice, elicits the warm admiration of its owner.
Daghestan and Shirvan rugs resemble each other very closely, and one has to consider and balance very fine distinctions when deciding between them. The types of traditional purity which have been preserved in prayer-rugs have been copied again and again of late years with poor materials and wretched dyes, so that it were better if time-honoured patterns had passed away with the finer weavings of other days in order that antique specimens might remain distinctive.
It is not unusual to see, in very old designs of the Shirvan district, great variety in the narrow border stripes that appear on sometimes one but oftener on both sides of the broad stripe in which tribal pattern is invariably found. In these narrow stripes highly conventionalised flower forms are stiffly arranged, but without attempt at the connecting-line or meander effect noticed in Kabistan patterns. The fringe of Shirvan rugs is usually tied and falls in loose ends.
There are four patterns that more than any others appear in border stripes of Caucasian rugs ; these are the ” tarantula,” the ” hooked swastika,” the ” reciprocal trefoil,” and the ” tree pattern.” Of these the first and second may be traced to Daghestan, where, with the ” link-in-lozenge ” and the large “S” pattern, they find expression in old rugs and lend designs freely to Kabistan and Shirvan weavers, while Karabagh and Kazak weavings claim the other two designs. Like all other motifs of the district, however, these are interchangeable, and we are thrown upon details of method and weaving to settle the disputed origin of fabrics.
All the scattered motifs of ornament with which the Caucasian district abounds seem to have been gathered together by the weavers of the flat-stitch carpets known for a long time as ” Cashmere ” rugs. All the designs found in the knot carpets made to the north and to the south of Shemakha are woven in these fabrics, but particularly are two Mongolian designs perpetuated in old Soumacs,—the ” mountain ” and the ” knot of destiny.” These two patterns are skilfully woven and are almost universally found in antique rugs as tribal patterns in very truth.
Because of the similarity of weave in these rugs to camel’s-hair shawls they were erroneously dubbed ” Cashmere ” ; for, as the shawls came from the Vale of Cashmere, why not the carpets? It is now many years since the mistake was made and corrected, and among the weavings of the Caucasus, both ancient and modern, nothing of more interest or value exists. The long ends which hang from the backs of the rugs give weight and a clinging quality, both of which desirable features make it possible for these tapestries to serve as satisfactory floor-coverings.
Karabagh rugs are at their very best when there may be detected in them an evident intention to copy the gloss and sheen of a leopard’s skin, and in some antique specimens we find the black and tan spots freely disposed over the field of the rug in such a way as clearly to indicate the intent of the weaver. Both this pattern and perfection of weave are, however, rare, and only by way of making it possible for the mind quickly to grasp an idea of the quality of the earliest and best Karabagh weaves is attention called to rare fabrics now almost impossible to obtain.
The output of the district into which we are led by examination of fabrics that bear resemblance to and yet differ from Daghestan and Shirvan weaves consists at the present time, as it has for the past century or more, of a strange mingling of ideas and methods in the rugs made and sold. Loose, careless knotting and hasty work at times show great deterioration, and still there is an undeniable charm about Karabagh weavings.
Both Persian and Turcoman influences are felt in designs ; and these may be easily accounted for, because of the location of the district midway between the north and the south, where, during former Persian rule and later Russian possession, thought and style have been successively dominated.
Heavy beyond all the fabrics of the district, Kazak rugs may be grouped as coarse nomad products ; but their interest and value to the collector and the student is very great, for they show a virility which is found only when methods are crude and ideas unfettered. We knew of the ” Cossacks ” as horsemen in central Asia long before we had become familiar with Oriental rugs, but to wanderers and skilled riders the name ” Kazak ” belongs, and tribes dwelling in the mountains of the Caucasus adhere to old ideas there, as wherever they roam and settle.
It is first, last, and always the intent of those who make rugs for their own use to continue in well-established ways, and to perpetuate tribal patterns. Such are more or less necessary when personal property is indicated by marks put upon it to distinguish it from that which belongs to another. Were it possible for us to do so, we might without doubt find on vast numbers of Oriental rugs tiny devices which look to us like fragments of pattern, but which are in truth marks of ownership.
Broadly speaking, Kazak rugs are heavy, and are tied with fewer stitches to the square inch than other rugs of the Caucasus. A rare quality about the wool or hair used causes it to untwist, when the knot ends are cut, into numberless single hairs, so that one knot-end will separate into upward of 200 parts, making upon the surface almost a smooth effect notwithstanding the great length of the nap and the few knots tied on an inch length of the warp.
In modern Turkish carpets, in which the knots are made of materials prepared by mechanical apparatus, there is never this feature which so distinguishes the heavy Kazak weaves of the Caucasus, making them beautiful, substantial, and lustrous.
Wandering mountaineers north of the Caucasus range have given to the nomad products of that quarter a different element in design from those adopted by the Kazaks of the trans-Caucasus region. A crude rendering of animal and human forms is shown in these rugs, which adhere in the main to Daghestan styles, but which in antique specimens show many Mohammedan devices.
In their prayer-rugs, the comb, and the spot upon which to place the bit of sacred earth brought from Mecca, are almost invariably worked with .other Moslem symbols. In the latticed field of old Tzitzi rugs the tree pattern is found most geometrically drawn, and the fringe extends from a webbing not unlike that known as ” Turcoman,” upon which an outline pattern in knots is sometimes tied.
At the present time the rugs of the Caucasus are made to order in such great quantities that wherever we roam we meet and may study them in hotel corridors, office buildings, and in our homes, where they wear well and are most, useful. They are invaluable in the study of ornament ; for as yet the designs are apt to be the same as they have ever been, or in any event the line is distinct between designs which are native and those which are borrowed, and the day is not far distant when the student panoplied with courage will dare to distinguish between them.