THE ORIGINS OF THE KILIM:

The following manuscript will appear as Chapter VIII in Peter Davies' forthcoming book, The Tribal Eye: Antique Kilims of Anatolia, to be published by Rizzoli, New York, in June 1993.

Undoubtedly the most traveled road in exploring the meaning of the motifs and designs of Anatolian kilims is the ongoing investigation into the kilim's origins and history. As etymology has contributed to an increased understanding of language, so a study of the origins of the kilim, it is reasoned, should lead to an expanded understanding of its visual language. The focus of this approach is on understanding how, when, and where the kilim originated, by which culture or tribe it was created, where its imagery and design came from, and how the kilim tradition and other weaving and design traditions interacted and evolved over time. For purposes of discussion most of the conflicting points of view on the subject of origins may conveniently be grouped under two general hypotheses: the Türkmen and the Goddess theories.1 In the Türkmen hypothesis the Anatolian weaving tradition, including its technology, techniques, and design, is believed to be a legacy of the Türkmen tribes which started their major migrations to Anatolia from Central Asia in the eleventh century A.D. Such discussions emphasize the pan-Asian origins of the kilim, tracing its beginnings to Khurasan, historically the Central Asian homeland of the Türkmen, which lies to the east of Iran between the Amu Darya (Oxus) River and the Paropamisus, a region still occupied by the Türkmen of Iran, Afghanistan, and Turkmenistan. Some kilim motifs are traced further back to the original Far Eastern Türkmen homeland in Mongolia on the fringes of the Chinese Empire. The influences of Islam, the Türkmen having been largely converted by the ninth century, are also emphasized, as well as other religious allegiances that predated Islam, in particular, Central Asian shamanism.

The Goddess hypothesis emphasizes the indigenous origins of the Anatolian kilim, arguing that Anatolia had developed its own weaving tradition long before the Türkmen migrations. Rather than being the product of outside influences the Anatolian kilim is seen as descending in a direct line from Neolithic prototypes, presumed to have existed in the form of woven hangings displayed in cult rooms devoted to the worship of a fertility goddess. One of the attractions of this theory, and one that certainly explains its great appeal, is that, if its early Neolithic origins are proved, the Anatolian kilim gains increased prestige as one of the oldest, if not the oldest, continuous design traditions in this world. The two schools of thought are not as rigidly or neatly divided as suggested here: Türkmen proponents are willing to acknowledge indigenous Anatolian influences and Goddess proponents certain Türkmen influences. Nevertheless, underlying the present discussion on the kilim is a great divide on the subject of origins. The point in history at which these two schools diverge is A.D. 1071, when at the battle of Malazgirt in the Van region of eastern Anatolia invading Orguz tribesmen decisively defeated the armies of the Emperor of Byzantium. As a result of this momentous battle (occurring only five years after the Norman Conquest of England) the former Byzantine province of Anatolia was opened up to successive waves of Central Asian nomads. It is estimated that between 350,000 and 600,000 people along with seven to eight million sheep came in four large waves between 1071 and the late thirteenth century.2 The Orguz, the predominant tribe in the migrations, evolved in Anatolia into the tribal groups identified as Türkmen, as well as a large component of the groups known as Yörük. It is the various subtribes of the Türkmen and Yörük that have woven the bulk of the kilims we now identify as Anatolian.

Although there is no evidence for the use of wool at this early period, it is generally accepted that sheep were becoming domesticated around 7,000 B.C. and that southwestern Asia was probably their place of origin. By the fourth and third millennia B.C. there is a great deal of evidence that flocks were being managed for wool production.4 As to loom technology, it is established that central and west Anatolia were part of the warp- weighted loom tradition from at least the fourth millennium B.C.5 The discovery at Çatal Hüyük of what may be ceramic warp weights and possibly a heading band seems to prove the existence of the warp-weighted loom in early Neolithic Anatolia.6 Since the horizontal ground loom is composed entirely of perishable materials, its introduction into the Anatolian region is difficult to date. However, in nearby Egypt, where conditions for preservation are ideal, evidence suggests that at least by the fourth millennium B.C. a horizontal loom with shed, heddle, and beater bar was in use.7 When Anatolian weavers adopted the horizontal or two beam loom is a crucial factor when considering the possibility of tapestry production since the maintenance of even warp tension and stability are essential to this weaving technique. The instability of the warps of a warp-weighted loom makes it an unlikely vehicle for the production of tapestry weave.8 While specific evidence for the existence of the shedding device in Anatolia at this early date has yet to emerge, it is difficult to believe that some form of it had not evolved over the eastern Mediterranean by the late Neolithic.9 While it is possible to identify the existence of the technology necessary to produce slit-weave tapestry at these early dates, the evidence for the existence of the technique itself is much later. By the early to middle second millennium B.C. slit-weave tapestry seems to have entered Egypt from Syria, Anatolia's immediate neighbor.10 At least one expert has associated the invention of the slit-weave technique with Anatolia.11

Polychromy, another prerequisite for the emergence of tapestry weave, is difficult to establish because of the carbonized state of extant textile fragments, which makes it impossible to determine whether dyes were used. However, the picture that is emerging suggests that polychromy developed gradually in the second and third millennium B.C.12 There is evidence to suggest the stripe was in use in the fourth millennium B.C. in Susa, Mesopotamia, another region bordering on Anatolia.13 In the third millennium B.C. there are three "hazy" indications of tapestry weave in the Near East, one of them for Anatolia;14 however, the evidence for Anatolia is unfortunately undocumented and very controversial.15

The first really comprehensive view of Anatolian weaving emerges from the ruins of Gordion, an ancient Phrygian city in west-central Anatolia remembered best as the place at which Alexander the Great cut the "Gordian knot." Preserved in several tombs and a layer of the city destroyed by fire in 690 B.C., a significant cache of textile fragments has survived, including various kinds of plain weave, felts, weft wrapping (soumak), and, significantly, scraps of slit-weave tapestry. The preserved patterns include simple stripes, quadruple lozenges, meanders, and double-barrel stripes.16 These fragments establish that 1,800 years before the arrival of the Türkmen the weaving technology, the slit-weave technique, and simple geometric designs were already in the indigenous tradition.

Another cache of woven fabrics, dating from the late 5th and/or early 4th century B.C., also casts light on Anatolian links with slit-weave tapestry. These are the Pazyryk textiles, which were preserved frozen in the ice of a chieftain's tomb in the Altai, a region in southern Siberia. Two strips of woollen cloth in slit-weave tapestry design found in this tomb are believed to have originated in Anatolia or neighboring Syria.17 This evidence not only points to Anatolia as being in the slit-weave producing region but also suggests that at this early date Anatolia or its near neighbor was already an exporter of weavings in this technique. Therefore, fifteen hundred years before the Türkmen settled in Anatolia, examples of the technique were being exported to the East from the Anatolian region or nearby Syria. However, it is one thing to recognize that Anatolia had a highly developed weaving tradition long before the eleventh century A.D. and that this tradition included the technology, technique, and the beginnings of a design tradition out of which a kilim tradition could have emerged. It is another to establish, as proponents of the Goddess theory have attempted to do, that Anatolia had already produced a fully developed kilim tradition in the sixth millennium B.C.

The Goddess Theory

The murals, Mellaart conjecture, were used in some cult rooms where woven hangings were not feasible, probably because of expense. The imagery of these murals, together with surviving plaster reliefs and terra cotta figures, has been interpreted as a codification of a number of ideas expressed in symbolic form which have been central to mankind's intellectual growth since Paleolithic times. As descendants of this tradition Anatolian kilims are by this reasoning the bearers of a repertoire of religious and symbolic motifs which date from Neolithic and even Paleolithic times.20 The Goddess theory when it was introduced quickly intrigued many connoisseurs and captured the imaginations of quite a few kilim gallery owners worldwide, but even its chief proponent, Mellaart, has admitted that to believe this hypothesis "...requires an act of faith almost as great as, and not entirely unconnected to, belief in God."21

A wide range of experts, including kilim and carpet scholars,22 a professional weaver,23 and an archaeologist-anthropologist,24 have seriously questioned this hypothesis and shown an unwillingness to make this leap of faith. Instead they have quite reasonably asked that it be proved by the usual methodologies of science and scholarship. Their serious challenges to the validity of the Goddess hypothesis have yet to be satisfactorily answered.

To begin with, the documented evidence for the existence of the requisite wool yarns, the loom technology, the polychrome textiles, and the slit-weave tapestry technique in the early Neolithic period has yet to be produced. The historical record of the Near East established to date suggests that weavings of the sort attributed to Çatal Hüyük are not likely to have appeared before the third or fourth millennium B.C.25

The fragments of burial cloth found at Çatal Hüyük have been proven to be linen, not wool as Mellaart once thought.26 Linen, because it does not take dyes readily, is not suited to polychromatic weavings, certainly not of the sophisticated type attributed to Çatal Hüyük. Further, there are additional problems with linen. While it is suitable as a tapestry wrap it is not a suitable weft being deficient in those qualities that make wool so ideal for tapestry: loft, elasticity, flexibility, and compactibility.27 Even the evidence for the use of the most basic dye, ochre, at Çatal Hüyük is very ambiguous.28 While the domestication of sheep was underway in the early Neolithic, sheep at this date were still in their primitive, kempy form rather than the white woolly domesticated animals we know.

The kempy coats of primitive sheep were coarse, hairy, and hence unsuitable for spinning or weaving. At the most there may have been a short, skimpy, woolly undercoating that would evolve over the millennia through selective breeding into full woolly coats.

The kempy coats of primitive sheep were coarse, hairy, and hence unsuitable for spinning or weaving. At the most there may have been a short, skimpy, woolly undercoating that would evolve over the millennia through selective breeding into full woolly coats. Further, these primitive coats ranged in color from black and dark brown through reddish and buff or gray colors, with very little white, colors typical of wild goats. These naturally pigmented fibers would have been unsuitable for dyeing and thus for polychromatic weavings -- other than for such weavings using animal fibers in their natural colors.29 Not until the fourth or third millennium B.C. does evidence begin to appear for managed flocks of sheep suitable for wool production.30

While there is evidence suggesting the existence of the warp-weighted loom at Çatal Hüyük,31 the loom at this early period would have been at a very primitive level of technology lacking the necessary warp tension and stability and hence not equipped to weave fabrics consisting of discontinuous wefts or eccentric wefting. While fragments of burial cloth survive from Çatal Hüyük, these seem to be the product of extremely elementary constructions. The cloths are either the result of weft-twining or transverse wrapping, simple techniques more related to basketry, or in the case of tabby cloths, to darning, a laborious method of weaving in which alternative warps were picked by hand. In all cases the cloths are the product of slow, tedious work. From a technical point of view it is an enormous leap from these simple burial fabrics to the sophisticated, large scale tapestries claimed by Mellaart.32 Further, it seems very unlikely that a slit-weave kilim tradition would have developed on a primitive warp-weighted loom and then been transferred to the ground loom tradition that has come down to us.

Not only is 6,000 B.C. impossibly early for the existence of the kind of weavings Mellaart attributed to Çatal Hüyük, but what strains credulity even further is the impossibility of the existence of the much earlier weaving tradition out of which such sophisticated weavings would have evolved. Furthermore, if a full-blown kilim tradition existed in the early Neolithic period, it is impossible to explain how a kilim tradition which still showed considerable vitality as late as the nineteenth century had remained virtually static, seemingly frozen in time for approximately 8,000 years.33

In addition to these serious questions about whether Çatal Hüyük had the necessary level of weaving technology to produce kilims, there are questions regarding the primary evidence which Mellaart presents to support his theory. The wall paintings as recorded in photographs, scale drawings, and what are described as "sketched reconstructions" have come under serious scrutiny. The bulk of the evidence chosen to demonstrate striking resemblances between the Neolithic murals and Anatolian kilims is drawn not from the photographed paintings and scale drawings included in the official archaeological reports published in the 1960s, but from the 44 examples of sketched reconstructions which did not appear in print until 25 years later. These are not, as would be expected, supported by photographs taken at the site or by scale copy drawings executed by trained artists, nor has their authority been corroborated by anyone associated with the original excavations. But what is most damning and calls the credibility of the sketched reconstructions (and hence Mellaart's entire argument) into question is that in many cases the information in the official archaeological record contradicts the very existence of some of the murals depicted in the sketched reconstructions, their locations, and their content.34 The suspicion aroused regarding Mellaart's methods and motives is understandable given the lack of any documentary evidence for materials which have since been destroyed or covered up. Leaving apart the sketched reconstructions Mellaart's only credible evidence is reduced to a few photographs of wall paintings that have simple designs of triangles and rhombs that are not especially striking in their resemblances to Anatolian kilims. Not only is the authenticity of these undocumented sketches very doubtful, but their plausibility as painted copies of weavings has been carefully examined and found wanting. From a weaver's point of view the sketched reconstructions of the painted murals are not convincing depictions of weavings. Given the technical restraints of the slit-weave technique, most of the designs could not have been produced using this method. Some of the designs (those shown in the photographs) are more akin to those of basketry or feltmaking. In slit-weave tapestry some designs would be impossibly weak in structure. Nor is the way that some designs are organized convincing, consisting as they do of non-objective geometric forms scattered randomly rather than arranged in the aligned, regular fashion of woven patterns. Some designs could not have been executed in weavings, depicted as they are with the designs turned the wrong way in relation to the warps.35

There are even more problems with the sketches. The ideational content of the authenticated and the reconstructed evidence is markedly different. Mainly these differences have to do with the presence of the Mother Goddess. There are figures in the authenticated plaster wall reliefs and small terra cotta figurines from the excavation that could be interpreted as Mother Goddesses, but such images appear only twice in the photographed murals. In marked contrast, "Goddess figures" appear in no less than 25 of Mellaart's 44 sketched reconstructions.36 In light of the very serious and legitimate objections to this undocumented evidence, it seems very lame indeed to argue that sketches had to be used since photography proved impossible because of the poor condition of the murals.37

Whether there actually was a Goddess cult at Çatal Hüyük has been cast in doubt. Since the civilization in question had not produced a written language, the evidence for its religious practices rests entirely on archaeological artifacts. Anyone, therefore, who attempts a reconstruction of the religious practices of this unknown people who inhabited a still-unnamed city is working in the realm of the speculative, since the archaeological evidence is open to a number of interpretations. An archaeologist-anthropologist, who left aside the suspect sketched reconstructions and focussed on the documented evidence, has concluded that Çatal Hüyük most likely had a religious pantheon of deities and spirits, not a fertility cult built around a Mother Goddess. There is, according to this scholar, no evidence for a Goddess who served as a preeminent figure, to say nothing of one who was dominant. While it is true that a Mother Goddess figure has been present throughout much of Anatolian history, the first convincing evidence for such a cult is no earlier than the fourth millennium B.C.38

In light of this evidence belief in the Goddess theory requires a considerable leap of faith. Certainly it is the romantic nature of the hypothesis rather than the convincing nature of the case made for it that explains why many in the kilim world have embraced it so readily. These converts argue that the resemblances between the extant Anatolian kilims and the Çatal Hüyük murals, as recorded in the sketched reconstructions, are too great to be merely coincidental. But what if the influences have actually flowed in the opposite direction, as sketched reconstructions of murals so badly damaged they could not be photographed were carried out years later by someone familiar with Anatolian kilims? Or, as has slyly been suggested, by someone working with an Anatolian kilim book at his elbow?39 For those who have nevertheless made the leap and believe in the Goddess theory there is yet another major difficulty: There is virtually no evidence of the kilim tradition between its supposed Neolithic origins and the kilims surviving from the last few centuries. Proponents of the Goddess theory when faced with this hiatus of almost 8,000 years argue that possessing the head, that is, the supposed Neolithic weavings as recorded in the sketched reconstructions of the Çatal Hüyük murals, and the tail, the surviving body of Anatolian kilims, we should be able to infer the existence and character of the missing body. Or using an analogy drawn from paleontology, if one discovers an ancient fossil prototype of a still-living fish, then the stages in between are not important; they must have existed even if they have not been found and thus are bound to turn up sooner or later.40 However, if the head is proved to be false or the wrong one, then we are left with only the tail. We thus seem to find ourselves right back where we started, with the very end of a tradition and little sense of its origins or evolution. But not quite.

Even those who question the validity of the Goddess theory do not find it unreasonable to expect some elements of Neolithic imagery to survive in Anatolian kilims. However, to accept the possibility of some Neolithic survivals in the Anatolian kilim tradition is quite different from arguing that the Anatolian kilim is an artifact that originated in the early Neolithic and maintained its essential character over 8,000 years. While the original hypothesis of the Goddess theory may not hold up under scrutiny, the research it engendered has stimulated interest in the indigenous origins of the kilim. This new emphasis strikes a responsive chord in kilim studies since it accords with the growing perception that the kilim may have origins different from those of other flatweavings and the knotted carpet. The old assumption that the kilim was brought to Anatolia wholesale by Türkmen nomads is no longer entirely convincing.

Peter Davies bought his first Anatolian kilim in 1962 while living in Turkey, where he was on the faculty of the American College in Izmir. He has served on college and university faculties in this country and in Canada for more than 20 years and holds a Ph.D. from the graduate school of Yale University. He founded Turkana Gallery, a New York City gallery that specializes in antique kilims from Anatolia and other weaving regions. He has mounted numerous exhibitions, lectured widely, and built a number of private and corporate collections.

Notes