THE HISTORICAL CONTEXT OF TRIBAL FLATWEAVES
Historical evidence suggests that the knotted-pile weaving technique was brought westwards from Central Asia by groups of nomads who became generally known as "Turkmen", and especially by the group of migrating Oghuz tribes. Knotted-pile weaving was a technique which basically developed from plain weave. Like such flat weaving techniques as the cicim, zili and sumak, the basic plain weave was supplemented with extra colored pattern threads. The knotted-pile technique probably evolved later than flatweaves since not only extra threads are introduced to the weave, but also they are knotted between the basic system of horizontal and vertical threads.
The oldest known knotted-pile rug is dated 5th century BC, and was preserved in one of the frozen tombs of Pazyryk in the Siberian Altai. Flatweaves and felts were also found in the same tomb. Small fragments of weft-faced extra-weft float and weft-wrapped weaves were also found in the Bashadar tomb in southern Siberia and Noin Ula in northern Mongolia, dating from the pre-Christian era. It is not certain that these were floor coverings, but, nevertheless, their techniques do bear similarities to today's flatweaves.
If, as is thought, the knotted-pile technique was brought to Anatolia by the nomadic Moslem Oghuz tribes then it follows that flatweaves which undoubtedly were their predeccessors may have arrived in a similar fashion. Flat-woven rugs are not as thick and as long-wearing as the knotted-pile rugs. Being mainly nomadic artifacts they were generally used until totally worn out and then they would invariably be cut up for other uses. Because they disintegrate relatively rapidly, archaeological evidence of them is rare. Apparently they were not used by the patrician classes of sedentary societies and so they have never been preserved by these groups as heirlooms.
Weft-faced, extra-weft wrapping and weft-float woven fragments of wool, goat-hair and linen dated 7th century BC were found in Phrygian Gordion. Traces of woven textile were found on the floor of a room in Gordion's Megaron III, dated 690 BC. It was impossible to preserve the cloth but motifs and colors were identified.
Woven fragments thought to have been used as floor rugs were discovered during the Dorak excavations too, so it appears even without more conclusive evidence, that flat weaves were extant in Anatolia before the arrival of the Oghuz Türkmen. We also know that some form of woolen textile was woven by early Anatolian peoples, as the discovery of spindle-whorls and spindles in excavations throughout Anatolia attests. Textiles in general have a past far beyond that of the woven rugs we are discussing here, but the history of woven fabric is beyond the scope of this study. In the sheepraising areas of Anatolia which were the continuation of the steppe regions of Asia known as the rug-weaving belt, flatweaves were certainly being woven long before the arrival of the Oghuz.
In the Near East in general, tapestries woven in the neighboring ancient Egyptian, Coptic, Arab and Byzantine territories were apparently similar to the kilim as we know it, as were also the so-called Egypto-Arabic weft-faced fragments of pieces dating from the 8th and 9th centuries found in Fostat in Egypt and now in the Metropolitan Museum.
One of the oldest datable Turkish flatwoven rugs is a weft-wrapped sumak rug in the Textile Museum, Washington, D.C which has a composition of a central octagonal medallion and small peripheral octagons. It has been dated 15th or 16th century because the composition is so similar to that of the so-called "Holbein" knotted-pile rugs of that period.
There are other examples that give us historical clues. A small group of tapestry-woven kilims may be dated between the 16th and 18th century because of their similarity in design to Ottoman court styles of that period. One of these, a tapestry-woven kilim in the Mevlana Museum, Konya bears the large, floral palmettes of the imperial style of the 16th and 17th centuries. Six other tapestry fragments from the Ulu Mosque, Divriği near Sivas which are now in the Vakıflar Kilim and Flat-Woven Rug Museum in Istanbul, have designs strikingly close to Ottoman textiles, tiles and tents of the 16th to 18th centuries. A similar kilim in the Vakıflar depot was collected from the Gümüşlü Mosque in Amasya, and still another sample of this distinct group is to be found in the Armee Museum of Ingolstat, Munich, in a kilim which lies in an Ottoman tent dated 17th century.
Two later tapestry kilims can also be found in the Kilim and Flat-Woven Rug depot. These carnation-patterned rugs were collected from the Hisarbey Mosque in Kütahya. Another "Ottoman-style" kilim in the Textile Museum in Washington, D.C. was woven with a rare technique, one which is more reminiscent of Iranian court kilims Other pieces of a later date are a medallion kilim in the Kestner Museum, Hanover and a prayer kilim in a private collection.
Only one type of very late curvilinear floral-patterned tapestry-type kilim appears to exist. It has a large, rather coarse, predominantly red medallion motif and is said to have been woven between Afyon and Uşak. This small group, the designs and technique of which are so markedly in contrast to those of traditional kilims, may have been also woven at the Ottoman court by craftsmen familiar with the tapestry technique who had been brought from Egypt for this purpose. Another possibility is that they could have been woven in Egypt from designs sent from Istanbul. In any case, they must be seen as quite sui generis. But although this group has quite a different origin from the more traditional Anatolian flatweaves, it did, in fact, have an influence on them. The carnation, tulip, hyacinth, rosette and other characteristic court motifs appear on many traditional kilims in stylized and geometric variations. But these may also have been directly inspired by Ottoman fabric, tile or embroidery motifs. It is extremely difficult now to detect how such motifs which usually appear only as isolated floral motifs among more tribal designs were transferred from one medium to another.
The history of the traditional flat-woven rug is even less definite since there is little historical evidence about the weaving of these tribal products. From the 16th century onwards, Ottoman Palace Records, Regional and Township Judiciary Registers (Eyalet ve Vilayet Muhakeme Siciller) and Judiciary Legacy Records (muhallefat Kayıtları) containing lists of personal and household effects refer frequently to kilims together with what are thought to be knotted-pile rugs (kalice and halı). However, since little information is given abouth the kilims mentioned in these sources, we are unable to say whether they were kilims of the tradiitional type or floral tapestry weaves.
The Turkmen tribes began their spas modic migrations across the Asian steppe towards Anatolia in small, isolated groups. They formed encampments along vastly varying routes and settled under a variety of local governments. They periodically be came part of local, clan-based principalities and chiefdoms, or were forced by such climatic factors as drought or political factors such as Mongol pressures to remigrate and scatter. Finally, they began to arrive in Anatolia in or even possibly before the 11th century, where they either joined existing groups or regrouped as isolated, very traditional units.
Tradition among the Turkmen tribal groups has been affected by a great number f factors such as the influence of the groups with whom they shared their central Asian environment and encampments during migration, the legacy of early Anatolian civilizations and other local ethnic groups, the Crusades, Mongol invaders and the movement of cultures from West Africa to Central Europe during the Seljuk and Ottoman periods. All have profoundly affected the tribal groups' life styles. In addition these influences added to the idiosyncrasies of each individual nomadic weaver. On top of these the characteristic dyestuffs and wool of each area help explain the variety of weaving techniques and the amazing multiplicity of design on Anatolian flat-woven rugs.
Another important factor affecting tribal life was the evolution of the religious tradition Oghuz groups while in Turkistan and Khurasan and then later in Anatolia had contact with Moslem dervishes, ones who were wandering or who came for visits. The heterodox beliefs and practices of the mystic orders affected a number ofthe tribal rites and customs. Each isolated group absorbed the teachings and popularized the practices of the dervishes in its own way. The individuality was so great that the teachings of different pirs of the same order might be interpreted quite differently by different tribal groups. This actually resulted in the formation of new Sufic branches and splinter groups. The close relationship between the tribal groups and religious mystic orders and the evolution of belief naturally changed tribal traditions.
Each separate tribal group is the product of a distinct past. However, apart from orally transmitted tribal mythology and tales of lineage there is little evidence to support the conclusions which are drawn about this past. Written sources mainly record tribal conflicts of interests between the groups or with large states. We would suggest that woven rugs may help fill this void and can be a substitute for written tribal historical documments. If the few settlement areas in which authentic tribal designs are still being woven .were to be systematically examined in detail it might be ibal possible to reconstruct both the context of these designs and the true ethnic origins of the weavers.
Although there may be a basic similarity ing between the life style ofvarious tribal groups throughout Anatolia, the simplest artifacts ps and customs tools, utensils, birth and death rituals, diet, and most of all woven artifacts may vary immensely from settlement to settlement. Even at only a distance between In groups of 10 to 20 kilometers the contrast in customs and local artifacts can be remark- able. The nomadic women in a particular village or in a pasturing encampment (oba) of yoruks may behave with great freedom, while those of another a short distance away may be extremely restricted. The cicim or zili may be the most common weave in one village while in the next even the namezili or cicim might be unknown. These examples show the outline of a broad study but, more important, they point out that before we can really comprehend the context of the weaves we need to studt the tribal groups in depht.
Studies of the archives of the Ottoman period (Bas Vekalet Arsivi) have revealed the names of over 7,000 variously described ribal groups in the territories under Ottoman jurisdiction. This indicates the wide distribution and the high level of fragmentation of the original tribes. The Oghuz tribes (originally twenty-two in number) were divided into many segments each bearing its mark or symbol (damga, im, ey, generally some kind of bird or animal crest. In late 7th and 8th century sources the Oghuz were referred to as Turkic tribal groups especially in the orhan inscriptions, the earliest known texts in the two Turkish languages. Byzantine records of two centuries earlier refer to the "Turks". But as tribal groups with distinct entities they first appear in Arab, Iranian and Byzantine sources from the 10th century onwards. Both historical information and mythological tradition concerning the formation and the damgas of the various Oghuz tribes are given in manuscript works dating between the 11th and the 14th centuries: the "Divan-l Lugat-ut Tirk" of Mahmud Kashgari dated 11th century; the 14th century "Cami-ut-Tevarih" of Reshid- ed-dir; the "Secere i Terakime" of Ebul Gazi Bahadir Han; and Yazicizade's Tarih-i Ali Selcuk"
According to Mahmud Kash gari's Divan, the Oghuz were then a
"Turkish section, Turkmen. They were divided into twenty-two tribes each with its own mark or damga which is also the brand they used fortheir herds. It isthis by which they recognize one another."
The author lists the damgas of each tribe.
Reshid-ed-din records that Oghuz Han (the mythical founder of the Oghuz) had six sons, whom he called Giin, Ay, Yildiz, Gok, Dag and Deniz, they in turn each begot four sons making a total of twenty-four grandsons. Each grandson of Oghuz Han took as a totemic sign a distinctive mark chosen from among the animals of the hunt.
Two names and damgas not recorded in earlier works are included in the "Secere i Terakime" by Ebul Gazi Bahadir Han. His work indicates in all probability that some tribes had been absorbed by others while some had segmented. But since the majority of accounts give the number as twenty-two, it is generally accepted that, at the level of tertiary segmentation the Oghuz had become twenty-two tribes.
Throughout Anatolia 444 widely scat- tered villages and districts bear names related to the Oghuz. In the 16th century there were approximately 883. The domi- nant tribal name among such villages is Kayu. 20 Many other tribal names have long been forgotten. A large number of villages and settlements refer to themselves as Turkmen or Yoruk and claim Khurasanian ancestry, but they are unable to link their genealogy to that of a particular tribal group. Those who have been sedentary for a long time have only a vague sense of tribal identity and refer to themselves simply as Turk.
Yoruks will say that true self-respect and self-awareness (in a tribal sense) lie in the truly nomadic, the yoruk, way of life and that the origin and lineage of those who have become sedentary, living among highly populated groups, have become vague and uncertain. Some groups will call themselves yoruk to distinguish themselves from settled Turkmen implying that they are, as they call themselves, yoruk: "We who roam" (from the Turkish word yiriimek meaning to walk, to be in constant movement). The nomadic groups distinguishing themselves as yoruk say that they are the "Original Turks' (possibly Turkish lineage implied).21 According to Mahmud Kashgari, one section of the Oghuz became totally urbanized and "did not venture out of the city." They were called (by other nomadic groups) yatik (passive, submitted) implying that they had been laid aside or abandoned This indicates the disdain with which settled groups were held by the nomads in general. This attitude has, with the passage of time, become reversed. The sedentary people of Transoxiania, the Tajac, were called the Oghuz Turkmen Tik-manend (Turk-like, Turkic); this word in time generally became used to refer to the Oghuz. As they moved westwards towards the end of the 10th century, the Oghuz gradually became Moslem. It was the Moslem Oghuz in particular who came to be identified with the word Turkmen, al- though they did not recognize themselves as such for some time. They were referred to in early Arabic sources as El-Ghuzz, but by the 13th century the name Turkmen had become widespread.
Under the Mongol threat some sections of the Oghuz Turkmen migrated first towards Khurasan, then to Anatolia, thus often both settled Turkmen and nomadic yoruk groups claim a Khurasanian origin and can trace a very perfunctory ancestry to Khurasan. Some of the groups migrating along this route formed local dynasties, such as the Kumuk, which became the Anatolian Seljuks and the Kay, from which sprang the Ottomans. Occasionally other tribal groups established local principalities and formed confederacies strong enough to oppose or form alliances with the Byzan tines, the Seljuks and the Ottomans.
During the Seljuk and Ottoman periods some nomadic groups were forced to settle or be reiocated from time to time in order to build up a strong, effective border front or else to prevent the eruption of inter-tribal conflicts over land usage rights or economic needs. Tribal groups were known to have been widely relocated and settled in Anatolia especially during the 17th and 18th centuries. This continued until and even after the founding of the Turkish Republic.
Groups tended to settle temporarily for in time they either moved back to their former homeland or were sent back. Yoruks sent the Balkans from Anatolia during the Ottoman period, for example, returned hundreds of years later as immigrants to the Republic. It is thought that the so-called Monastir (Monastery) kilims were woven in western and central Anatolia by such groups of returning nomads.
A nomad yoruk of the Bozdoganli, a group of nomads in Bakirdag, describes this process of deliberate fragmentation in tribal terms:
"It was the Ottoman sultans who sepa- rated us Bozdoganlis, scattering us about one by one. There was a time when it was we who ruled over sultans, but then they split us up, and since that time, there's been nothing of the tribe left in us, our kin has dried out."
These words vividly describe the bellig erent attitude of the tribal nomads and their resistance to centralized administra- tion and forced settlement. In 1865 duringa resettlement campaign a serious reform movement was directed towards the nomadic life style by Dervis Pasa, the leader of the campaign. There were serious repercussions. In Cukurova for example spasmodic conflicts broke out between the Cerit, Tecirli and Bozdoganlu which resulted in the Tecirli being moved to Osmaniye and the Cerits to the Ceyhan plain. Meanwhile the Afsars who had previously been sent to Rakka were at the same time returning but when they attempted to settle in the region of Sariz and Zamanti, south of Kayseri where they had previously had their summer pastures, they were confronted by Circassians who had already been located there. This left the Afsars without and or even pasturage. Ali Riza Yalgınn in 1939 did research among these tribes and confirmed the effects of the movement.
Some settled tribal groups such as the Cepnis around Balikesir persist in calling themselves yoruk, while for example the Cepnis near Trabzon call themselves Turkmen. Perhaps the distinction here is related to the different religious sects of the two groups, the Trabzon Cepnis being Sunni-Hanefi, the Balikesir, group Shiite- Alevi The Tahtacus found in the regions of Balikesir, Mugla, Adana, Maras and the Hatay are referred to in Ottoman sources as These terms however are also thought to have been used to describe groups of different ethnic origins.
According to an analysis of Ottoman documents in the archives of the Office of the Prime Minister (Bas Vekalet Arsivij published by Cevdet Turkay:
"The Afsars were yoruk communities (asirets) known (i.e. named in the sources) as Afsar) or Afyaramanlu who the Begdili tribe (again asiret) and had their winter quarters in Cukurova on the plains between Adana and Maras."
The same author reports evidence of their distribution in the regions of Maras, Karaman, Kayseri, Kairli, Sis, Adana Erzurum, Ankara, Balya (Biga), Tokat, Isparta, Teke, Sivas, Tarsus, Diyarbakir, Amasya, Yeni II (Sivas), Halep, Kas, Kirsehir, Bor, Nigde and GUrdes. Although a date for this evidence is not given, it nevertheless indicates the widespread distri- bution of a single tribal group.
The Karakefilis, renowned for their cicims, zilis, knotted-pile rugs and especially their kiims, also appear in Ottoman sources distributed throughout the regions of Adana, Diyarbakir, Siverek, Eskisehir, Kastamonu, Siirt, Bilecik, Ankara, Kutahya, Urfa, Aydin, Kirsehir, Balikesir, Gordes, Mardin, Esme (Usak), Kula and Bursa.
Various groups linked to the original Oghuz gradually splintered into an ever- growing number of small tribal groups. These were either named after the original group or given other synonyms. In such cases a patrilineal name may equally have been superceded by a geographical place name, such as that of the pasturage or wintering places frequented by the group Even a synonym relating to a particular physical attribute of their chief was sometimes used.
Today there are very few groups bearing genuine anthroponyms (i.e the name of the parent tribe to which they were originally attached). However those bear- ing identifiable tribal names also tend to be well-known for their flat-woven rugs. These include the Afsar, Sacikarali, Karakecili, Aydinli, Honamli, Herki and Yuncus There are also mahy tribal groups, pastural communities or villages where the more generalized pseudo-ethnonyms Turkmen, Yoruk, Kurt or Cerkes are encountered.
Curiously very few of the tribal synonyms mentioned in Ottoman or earlier sources are related to a form of weaving. Notable exceptions are the Kilimli or Kilimlu in Mentese, the Qullus of the Aydin, Erzurum and Maras regions; the Akal of the Corum, Samsun, Kirsehir, Sivas and Ergani (Diyarbakir) regions, and the Sumakli of Biga. 28 Although some groups acquired synonyms related to their herds, or even their tents, they did not seem to adopt names from their flatweaving designs or techniques. This may have been because in the archetypal tribe such artifacts were considered a normal everyday necessity, an intimate part of the home of a nomad. They were not a commercial commodity so had no economic value between groups and had only an implicit function in establishing or carrying tribal identity.
As we have suggested, the history of the flat-woven rugs by tribal groups is inexorably linked to the individual past of each traditional community. Until we know more about the distribution and relocation of these groups throughout Anatolia and the factors related to change in their customs and traditions, the past of flat-woven rugs will remain a mystery.