Transoxiana Homepage

Transoxiana 14
Agosto 2009
Index 14
ISSN 1666-7050

Click here to lend your support to: Transoxiana Laptop and make a donation at !

Early Turks: Male Costume in the Chinese Art

Second half of the 6th – first half of the 8th cc. (Images of ‘Others’)

Sergey A. Yatsenko


Two main groups of Early Turk depictions in Chinese art come under analysis (practically all of them showing nobles, 9 complexes with 14 compositions on the whole). The earliest images (the earliest known ones belonging to the epoch of the First Turkic Kaghanate) are presented on mortuary beds and sarcophaguses of Chinese Sogdians of the 2nd half of the 6th c. We can see later ones in funeral terracotta figurines of the Early Tang Dynasty of the 7th – the first half of the 8th cc. (the epoch of the Eastern Turkic Kaghanates). These depictions trustworthily represent the costume: its completeness, its silhouette, coloring and the manner of wearing. There are no traces of “Sinicization” in both groups; some influence of Sogdians and, partly, of peoples of Kutcha and Khotan Oases should be mentioned. Some strong differences in the costume of early and late compositions are evident (within the first ones there are four clothes complexes with tribal specificity). In early depictions of nobility we can see black woolen belts without metal details which are wrapped two times around the waist and red trousers and others. The later costume differs in the predominance of two triangular kaftan lapels, the simplification of its decorative scheme, clothes being either very long or very short, and changing of color preferences. The higher status for personages in white clothing than for those in red (if they are grouped in couples) is documented in Chinese depictions. In terracotta there are noble people in two quite different styles: in the official festive costume and as horsemen in the military-hunting garment. The opposition of “fastening – overlapping” for earlier depictions or “fastening – throwing over” for later ones, probably depended on the situation the owners of the costume were in the activities they were doing.

The political might of the Early Turks and the vast territory in the period of the Great Turkic Kaghanate (551-603), and later – Western (up to 657) and Eastern (603-630 and 682-744) Turkic Kaghanates, their effective arms, horse trappings and other elements of their military subculture borrowed by many neighbors are widely known. The Early Turks costume was comfortable for warriors and hunters; the symbolism of its details (especially of the belt) was very attractive for other peoples. Unfortunately, authentic remains of Early Turks clothing are rare and fragmentary; in their motherland they may be dated to the time of the Second Eastern Kaghanate (from the end of the 7th c.) and they come from rather modest burials (Kubarev 2005, p. 27-40). Authentic early written information on the costume belonging to the time of the united Turkic Kaghanate and both Eastern and Western Kaghanates is even more scarce. The earliest fragmentary scripts of the Turkic epos were made later and do not contain “costume’ details. In many cases there is little hope for a special value of later “ethnographic” witnesses on the costume of Turkic peoples in the plan of clearing out the early system of views on clothes (though ethnologists and archaeologists have often made efforts to do it) even if we refer to the materials on the most conservative groups in the region of the Turks origin1. So, detailed depictions of them in early medieval art are very important.

For a long time the costume of the Early Turks has been known mainly from the earliest carved stone images on barrows and funeral sanctuaries (fig. 1)2. Such statues (with a rare exception are dated to the general period of the 6th-8th cc.) have been well-studied to date (see, first of all: Evtyukhova 1952; Sher 1966; Kubarev 1984; 1997; Tcharikov 1986; Bayar 1991; Khudyakov 1998, p.11-25; Bayar, Erdenbaatar 1999; Eleukenova 1999; Kubarev 2005, fig. 5-10; Ermolenko 2004, p. 25-29; 2008, p. 56-57; Dosymbayeva 2008). In a number of cases Sogdian influence is reflected in the manner of depicting poses and some artifacts in these statues (Hayashi 2006, p. 245-259). These carved stone images more often depict male warriors, definitely not of the lower classes; the personages’ looks being stereotyped (of not numerous iconographic types of this or that period for a certain region).

As it is evident to any specialist, clothes and costume accessories (usually adjusted to the belt) are depicted on funeral statues rather fragmentarily (only single elements which are important for the customer are shown in detail and they may often contain significant distortion of proportions), they are often rather schematic and many parts of their surface have become obliterated. It is highly probable that important costume details on carved stone images were painted (Ermolenko 2003, p. 238), which is typical for many developed art traditions in the ancient world beginning with Egypt and Greece. Among others, belt plaques could be painted if they were not engraved and where there were other important belt elements and items suspended to it (compare: fig. 26, 1-2)3. However, the initial painting has not been preserved (unfortunately, Turkic statues have never been investigated by specialists in chemistry to trace microscopic items of paint). We cannot completely exclude statues being clad in real garments and plaits made of natural hair added to them (Ermolenko 2007, p. 126-128). It should be mentioned that sometimes statues (first of all the face or some part of it) were painted even not long ago by the Turkic population nowadays during ceremonial rites (fig. 2), an ancient statue was tied with a modern belt, a head cover and other items necessary for a successful treatment of some diseasd (Kubarev 2004, p. 33).

Upper body garments of various proportions are depicted on statues completely enough (though schematically) (garment length; sleeve width; collar form, its cutting and the method of adjusting to the garment; wearing manner; decorative scheme and probably, cutting – are significantly variable). The most ancient Turkic term for such kind of clothing, unfortunately, not documented, but in a later period the Persian term chaftan gained popularity (evidently for some serious reason) and was applied to name a garment that is open down the front, with a bodice and skirt widening downward, and often with a fur collar (the latter can be seen in some carved stone images in the Altai Region: Kubarev 2005, fig. 9, 1, 12). I conditionally use the term caftan in this article. It was made of three layers (the inner felt layer, the upper layer of ornamented silk, and the lining of cheaper silk), probably even before Chinese silk reached the height of its use in Turkic territories beginning in the 7th c. (judging from finds of silk even in poor burials in the Altai in the turn of the 7th-8th cc.: Kubarev 2005, p. 27, 40). As a rule, only the costume worn in a definite season was depicted in the art of traditional societies. It was with a rare exception a warm season, in order to make the image of the portrayed person more informative or recognizable (Yatsenko 2006, p. 21-22). There is an opinion of single cases when personages are clad in winter fur coats (Kubarev 1984, p. 27) and there is also a suggestion that the number of fur coats depictions on statues are equal to the number of caftans (Kubarev 2005, p. 46), but the latter suggestion is not proven. Besides that, on carved stone images the hair style of nobility is often presented in full detail (plaits); there are different (for different periods of time, different regions and different status of depicted characters) details of plaque and other plate decorated belts, more seldom it may be the silhouette ( and also the decorative scheme) of headdresses. Unfortunately, we cannot say the same (with a rare exception) about undershirts, trousers and footwear. Sketchiness and extreme electiveness in the interpretation of details in such images automatically mean that any conclusions made on their basis should be analyzed properly (Yatsenko 2006, p. 23).

Two series of detailed colored depictions are also used by researchers to characterize the costume of the Early Turks. One of them which dates according to the opinion of the majority of scholars of the period, to just after the collapse of the Western Kaghanate (after 657) is found on the western wall of so-called «Hall of Ambassadors» - one of rich dwelling complexes in the site of the ancient town of SamarkandAfrasiab. Unfortunately, the vast majority of depicted personages (the upper – third line of painting) were ruined by a bulldozer in the course of illegal earthwork in the reservation territory in 1965; the other part of the composition was being destroyed much earlier – as toilets were being excavated on the site in the 10th–11th cc. (the middle line of painting) and much more earlier (at the beginning of the 8th c.) – by the action of Arabs who conquered the town (eyes and bodies in pagan images damaged with the help of different blades).

Wall paintings of “Hall of Ambassadors” are known in colored copies made by artists from L.I. Al’baum’s detachment (1965) (Al’baum 1975) and black and white copies of G.V. Shishkina’s detachment in 1978 (the western wall only)(see first of all the entries of articles from the collection of papers of special conference on these wall paintings in Venice “Royal Naurūz”: Mode 2006, p. 110-122; de la Vassière 2006, p. 149-159; Arzhantseva, Inevatkina 2006, pp. 190-199 and the Russian version of the latter ones: Arzhantseva, Inevatkina 2005, p. 123-137). However, there is a serious divergence between the two series of copies of the same objects; the discovered depictions are copied in drawings only (which can naturally be subjective and not precise in details) but there are no photographs of good quality. The photographs made in field in 1965 have not been preserved, neither have the originals of the initial drawings of 1978 (I would rather think that they could have been more exact). Partly preserved in the Samarkand Museum, the fragments of the actual wall paintings have changed in their original coloring and surface structure as a result of a very unsuccessful “experimental restoration” by A.A. Abdurasakov; many details have been “restored” with analogous paints (Royal Naurūz 2006, pl. 4-6, 12). So, there is no opportunity to work with them productively (though Markus Mode, Francois Ory and a number of other western colleagues have tried to do it, but conclusions made on the basis of such material are rather doubtful). Some variants of reconstructions of the whole original composition with Turks are suggested today (see some of them: fig. 3); however, due to the majority of the figures having been damaged, there is no opportunity to prove their authenticity.

Notwithstanding all above mentioned circumstances and serious divergences in the existing copies, the value of the first most ancient images of Turks created not by they themselves but to the west of the Tien-Shan and Pamirs Mountains is rather great. The costume of the group of Turkic officials on the wall-paintings in this hall was systematically studied for the first time in a special chapter written by me in 1985 where later copies were taken into consideration, but published only ten years after and in English (Yatsenko 2004); later on the text was supplied with additions (see the publication in Russian which takes new details into account: Yatsenko 2007).

There is one more group of colored detailed images which is often used to reconstruct the costume of Early Turks – depictions from different materials connected with the Second Uighur Khaganate in Turfan Oasis (from 866 till the Mongol invasion and also in the period of Mongol rule). It is a far more numerous and various series presenting different social, age and gender groups. However, it should be taken into account that the Uighurs were enemies of Early Turks and before that, for a short period, they paid tribute to them; besides, they abandoned their forefathers’ shamanistic religion very soon and adopted Manicheism, and then, also rather soon – Buddhism and came under the strong cultural influence of China. A special section on the Uighur costume was written by me in1987 but published only 13 years later (!), being kept in the editorial office of Institute of the Oriental Studies, Russian Academy of Sciences (Yatsenko 2000, p. 367-382, fig. 64-69). Later, there were sections devoted to the analysis of the Uighur materials prepared for the collective publication on China (on the costume of Western Regions (Xinjiang) (Study on the Costume 1995, pp. 166-191, 210-213, 220-222, 234-237).

A series of Early Turkic costume depictions belonging to the 2nd half of the 6th – the 1st half of the 8th cc. in art monuments of neighboring China may be and should become much more important for the study of their dress. Two small series of such depictions attracted the author’s attention in 2007 – in the beginning of 2008. Chinese depictions of Early Turks have some peculiarities which on the whole makes then exclusively valuable for our research. First, all the depictions analyzed below have been completely preserved (which, unfortunately, is explained in many cases by the fact of their being selected by robbers for selling at the antiquity market). Second,they are characterized by a rather high degree of realism and detail. Third, they are polychromic and the coloring of some of items and scene depictions has been preserved practically in the original completeness (to my mind, the coloring made by a Chinese master fully corresponds to Turkic realities: see below). Fourth, the items from the first chronological group are the most ancient among dated ones detailed color depictions of early Turks in the world art. Fifth, in the first (the earliest one) of the analyzed groups (mortuary beds of Sogdians in China) multi-figured compositions predominate; analyzing them we can better understand the status of every person depicted. Sixth, these scenes (excluding terracotta 8) are documented by high quality colored photographs (but not graphic copies as in Afrasiab with inevitably subjective interpretations).

Naturally, images of ‘Others’, in this or that developed fine arts tradition always reflecting (not in the past only) local, artistic and ethnical stereotypes, are, in certain respect, standard and not complete. So, it is worth trying to compare the perception of a contemporary Chinese (or Chinese – Sogdian) master and his client (the degree of significance of different costume elements and small costume accessories, the symbolic of coloring and others) with the evidence (even fragmentary ones) about the early Turks own interpretation of their dress, documented by foreign reports, Turkic sculpture, petroglyphs and single finds of authentic clothes parts.

The work at the article was in many respects initiated due to my contacts with my American colleagues. I was given the opportunity to use comments on a number of details of marble and granite mortuary beds and sarcophaguses of the second half of the 6th c. from North-Western China presented by Dr. Judith Lerner (New York) in 2007. Here Sogdian officials – in the rank of “sabao” – from kingdoms of Northern Zhou and, probably, Northern Qi present accounts of their trips to the territories of nomadic Eastern Turks (see the first most general results: Yatsenko 2007). In the spring of 2008 Prof. Michael Sanders (California) addressed me a number of questions on the ethnical identification of clay funeral figurines belonging to the period of the early T’ang Dynasty in the exhibition catalogue of the Ezekiel Schloss Collection in China Institute in America, New York, March-May 1969 (the largest collection of such kind at the time in the West)4, and also some analogous figurines from private collections from catalogues of Sotheby’s auctions and a number of other catalogues. Unfortunately, up to the summer of 2008 images of early Turks were not properly identified among numerous synchronous depictions of “Western and Northern barbarians”; on personages attributed as Turkic ones the costume was not analyzed in detail and not compared with a large series of depictions created by nomads themselves. As a result, these images could be not used actively by specialists on costume history in their research so far. I am thankful for the technical assistance of my colleagues; J.A. Lerner (New York), P.P. Azbelev (St. Petersburg), A.V. Yevglevsky and T.M. Potemkina (Donetsk), O.N. Inevatkina (Moscow) и A.V. Varenov (Novosibirsk).

Let us proceed to characterizing of the costumes in the above mentioned images. There are 14 compositions of such kind known to me (they are numbered further on) on 9 different objects, 5 of them originating from one and the same grave (relief panels of the mortuary bed of An Qie). It should be mentioned that Chinese chroniclers, Buddhist pilgrims, as well as craftsmen and their clients were absolutely not interested in the costume of Early Turkic women5. All the terracottas analyzed below (chronological group 2) were identified as Turkic ones by the author in 2008.

Chronological group 1: the second half of the 6th c. AD (the period of the Northern Dynasties and the beginning of the Sui Dynasty).

These are the earliest recognizable images of Turks. They refer to the period of their greatest power. They are present in the marble and granite mortuary beds of Sogdian officials in the rank of “sabao” from the towns of X’ian, Taiyuan and, probably, some others (figs. 4 and 13). The painted panels of their backs and sides are pictorial biographies of the dead, such as, their trips to the Turkic steppes and also the reception of rather honorable Turkic ambassadors in one of Chinese towns (Short Report 2001, pp. 27-52; Northern Zhou tomb 2001, p. 4-26, fig. 12; Marshak 2001, pp. 228-264; 2004; Lerner 2005). Due to the accompanying inscriptions in the Chinese language in the mortuary beds we know the main events on the timeline of those Chinese Sogdians. The costume of Turkic personages in some panel details is usually painted against a light brown granite background or a white one of marble with a few colors only – red (two tones) and pink, black, white (on granite), brown and yellow. Comparison with later series of images depicting Turks in other regions allows us to be certain that such use of coloring of some parts (alongside with leaving white marble background unpainted) was not accidental. It was quite a sufficient set of colors as in the costume of Early Turks mostly prevailed one-tone and bright-colored elements. A variety of poses of the same figures (more often half-turned or en face, more seldom in profile or from the back) proves very valuable as it allows us to see the costume completely.

A specific hairstyle consisting of several plaits variously joint together serves an ethnical marker for all Turkic men, differentiating them from other “western and northern barbarians”. Undoubtedly, we mean representatives of nobility (later on, in the vast suite of western Ton-jabgu kaghan, due o the witness of Xuanzang (630 CE, plaits were only worn by 200 aristocrats from his nearest escort at that time: Chavannes 1903, p. 194). Unfortunately, neither client nor craftsman are completely interested in such costume accessories as bracelets, earrings and rings (even taking into account that for Early Turks themselves earrings for men were important and prestigious accessories and one of the main accessories of rulers6). This, in fact, concerns depictions on funeral constructions and accessories of every type analyzed in the article (mortuary beds, sarcophagus, terracotta in both chronological groups). Another most important accessory –the belt– is shown in these depictions with great detail7 (despite the fact, that the accuracy of the construction and decoration of the details of the metal belt accessories is not sufficient for us to create typological and chronological schema); textile patterns are also often represented, and in scenes 2 and 4 a special attention is paid to details of footwear.

Researchers characterizing Turks images in these depictions present them as a single cultural monolith with standard features. In my point of view, in the different panels of An Qie’s mortuary bed, we see three groups of Turks each different from the others in some important elements of clothing and hairstyle. Probably, they may be referred to different nomadic tribes of Turks included into the Great Kaghanate. An Qie had to deal with their representatives (as ambassadors) both in town (groups 2-3), and outside it (group 1).

Granite mortuary bed of the Sogdian official An Qie and his wife (at first incorrectly translated as An Jia).

This bad is of special interest for the early Turkic costume study due to the abundance of figures, the high degree of their detailing and the variety of coloring. An Qie died in 579 in the town nowadays named X’ian (40 years after his death it was called Chang’an’ and became the largest city in the world). This official was at service in China and headed Sogdian community in the Northern Zhou Kingdom and contacted nomadic Turks during the rule of their most powerful Mugan kaghan (553-572), probably one of the most dramatic periods of in the relations of Turks with the Northern Zhou (557-581) and its rival – the Northern Qi (550-577) from 561 to 672. Beside other duties, he was evidently responsible for receiving Turkic ambassadors arriving at the capital.

Turks are presented only on four panels of the mortuary bed which is painted in the richest and most various way. In my opinion, in two neighboring panels (Ns. 1-2) we see one and the same man - a Turk from highest ranks of aristocracy – “beq” – an important person from one of the Kaghanate provinces neighboring China (Tobgach) titled “irkin” or “chor”, and his young servant as well (or a personal secretary?), who met An Qie and his escort somewhere in the steppe. A similar couple is depicted in the panel to the left next to these two mentioned above (No. 3). B.I. Marshak’s attribution of ‘the main character’ from one of the panels as a Turkic kaghan (!) (Marshak 2004, p. 25) is evidently incorrect: it contradicts the information of Xuanzang about the costume and other attributes of a Turkic kaghan, and first of all about the diadem (see for example: Chavannes 1903, p. 194), and the ruler’s attributes on Turkic coins of Transoxiana of the 7th c. – the same diadem and special large ear-rings (see the most realistic images of Tun kaghan with his wife in Chach: Shagalov, Kusnetsov 2006, photo p. 189)8. It is clear that the Great khagan (or even an important regional ruler) would not specially come to the residence of a Sogdian official in the center of some of the Chinese Kingdoms; instead, we are dealing here with one of his deputies.

In scenes where Turks and Sogdians are depicted together the Sogdians are always placed to the right (receiving a guest in a nomadic yurt is not an exception), while the Turks are shown to be on the left or less honorable side. I may point to a number of scenes when both Turks raise their left hands but not right ones as in case of Sogdians (see for example the panel from An Qie’s funeral coach showing An Qie’s meeting a noble Turk, both of them mounted); in the banquet scene in the pavilion we observe the contrary situation. Often figure coloring does not follow the details carved in stone before, but obviously changes the initial contour or makes the outline more exact. In scenes 2 and 4 it was very important for the Sogdian client to accentuate the cutting of boots on the main Turkic personage (marking the seem stitching the sole to the boot with carved lines), as well as “the servant’s” boots как и сапог у «слуги» (the seam adjusting the vamp to the boot top); however, the same can be said about the figure of An Qie himself. Let us first consider three panels placed next to one another (we may suppose plots to be closely connected to each other). Unfortunately, researchers often use only black and white drawings of these scenes (containing some mistakes in details) or blurred photos in small size (the exception: Rong Xinjiang 2003).

No. 1. In this, most important for us, panel in the upper tier we see An Qie accompanied by two of his servants meeting two Turks during a hare hunting (fig. 9). The artist schematically reproduced some details of two belts worn by the main Turk (in the foreground) below waistline (compare fig. 26, 3)9. There are no plaques on the upper “main” belt. We see a hanging strap with a half-oval gold (gilded) end and a gold (gilded) rectangular buckle. On the second belt (it is a quiver belt) there are depictions of only two plaques (a rectangular one and, probably, a heart-shaped one). It is not possible to reconstruct belt details on the base of these schematic depictions (for example, to classify them as an early belt type according to G.V. Kubarev [Kubarev 2005, p. 48-50, 54]). But in any case the shapes of metal belt accessories (the buckle, the supposed plaques on additional straps) do not coincide with those ones on the real belt of late An Qie himself (Anjia [An Qie] Tomb 2003, photo on p. 212). It should be noticed that in the other two scenes where the craftsman depicted small costume details the belt of the Turk also lacks the plaques (however, we see them on the “servants” and the Sogdian An Qie) (compare belts with many details but without plated plaques on early statues of Turkic aristocrats of the 6th-7th cc. – fig. 26, 1, 3). The second, according to his rank, personage (“a servant”) there are no any visible belt accessories made of metal; there is a small round bag – kaptarga – hanging from the right.

Both Turks wore caftans on, approximately knee-length (we can clearly see the length of the main personage’s garment though the skirt of the horseman caftan rode up), without lapels, with an unfastened collar. The “servant’s” shirt has a horizontal collar. Whereas the latter, less important personage (in the background and depicted as a half-figure) has the caftan and shirt of one and the same color, the horseman is wearing the white upper garment with red edging. Such décor is typical for Sogdians of that time (Yatsenko 2006a, fig.181, 40, 184; 2006b, pl. 2, 40, 184)10, but it is very seldom fоr Turks (see the only costume of a Turk in hall 3 in Samarkand next to the famous ‘Hall of Ambassadors’ (Al’baum 1975, table IV). The both Turks are wearing red shirts with wide (?) horizontal collars under their caftans. The horse rider’s trousers are white with oblique black lines and black edging (the trouser legs were not tucked into the footwear): there are trousers of such pattern in one of depictions of Sogd (Penjikent) (Yatsenko 2006a, fig.181, 80) and there is also the tradition to sew on the trouser leg edging (the trouser legs were also not tucked into boots) (Yatsenko 2006a, fig. 181, 78, 82 and 84). The footwear of the main figure (personage is awkward), obviously presents black half boots or shoes. The hair of both of them is plaited, the plaits are as long as 20cm, a very small beard can be traced on the horseman’s face; both of them have no moustache.

No. 2. In the lower tier a game of dice is played on the carpet, probably outdoors, out of the settlement (fig. 6). The two Turks are wearing the same clothes and have the same hairstyle, but due to their different positions we can discern a number of details. In another situation (not hunting but being guests) their caftans are not thrown half-open but completely closed. On the “servant’s” caftan we clearly see a very important detail – large triangle lapels (the more significant personage does not have this feature). We also see that the “servant” is wearing black high boots without any projection above the knee indication that the trousers are tucked into the boots), and the person of a higher rank has black half boots on (his trousers are not painted there, but they were definitely light as a natural brownish shade of granite was left unpainted for them the same as for the personages’ skin). It is possible to see a small broad and thick beard for the “main” Turk. In the both scenes in this panel the more important Turkic personage (he is in the foreground) is wearing a white caftan and the less important one is clad in red; the “servant’ in both cases is depicted standing behind his master and is partially pushed background and hidden. The main Turk is wearing a belt which is tied around his waist two times and lacks any metal belt accessories; the “servant” has a belt buckle and an additional strap hanging from the right.

No. 3. In the second panel we see only one Turkic personage – an aristocrat – in the banquet scene at foothills in a company of Sogdians. He is sitting in the yurt house (fig. 7), however, not in his own one, but a travel yurt made up for him by Sogdians deliberately (in the other panel, in the same yourt depicted in detail and near it there are seven personages in the Sogdian costume and hairstyle). It is the only episode when a Turk but not a Sogdian is shown there in the right side but not to the left from the viewers (see above), probably, as a special guest. The host is treating him to some drink in a rhyton, there is a dish of fruit close to them.

Some of An Qie’s high-ranking companions (in Sogdian clothes) sit on the carpet outside the dwelling; the servants have just taken drinks, flat cakes and fruits (?) out of load packs on the donkey’s and camel’s backs and they are adding them to the refreshment. The Turk is dressed in the same garments as in the first scene. The only difference is that his caftan is completely closed during the banquet (evidently, according to the code of etiquette). Here the oblique black stripes on his trousers and his black footwear (probably, half boots) are seen even better. There is a dagger suspended to his black belt, but there are no plated plaques on it.

No. 4. The third panel shows the reception of two Turks in An Qie’s house (fig. 8). Many details make us suppose that they are a different pair. On the stove-couch “takht” – a comfortably seated host (in the center) is treating the Turk sitting at the edge with tucked up knees giving him a gold cup with a stem. In this, probably, most important banquet scene the clothes of both Turks are completely done up, and (to differ from their Sogdian hosts) they have folded their arms across their chest each hand tucked into the cuff of the opposite sleeve. If compared with previous scenes the main Turkic personage has no beard but his very long plaits reach his pelvis and he is wearing quite different upper clothes; to differ from An Qie there are no ear rings on him. He is wearing a longer caftan made of not white but brown (silk) cloth with red edging. The collar (as the one on the host’s caftan) is turned down and broad (we see similar ones on Turkic statues from Mongolia and Saryg-Bulun in Tuva: Yevtukhova 1952, No. 79, fig. 48, 1; Mogilnikov 1981, fig.17, 1); the reverse of the collar cloth is not distinguished by color against the general background (the sculptor just marked the contour). Such collars, possibly of fur, can be seen on a statue from the Altai (Kubarev 2005, fig. 6, 13) where it is represented as standing upright. The “servant” has short clothes without a border and they are completely red, as with the previous pair.

In this scene (where plated gilded plaques are shown even on the Sogdian servants’ belts and on the belt of the Turkic “servant”) the black belt of the more important Turk lacks the buckle, plated plaques and pendant straps (i.e. his belt is evidently not made of leather plated one but made of fabric); his belt is tied around his waist two times (the same can be said about An Qie’s belt in this scene), whereas the servants of both sides have usual plated belts. The “servant’s” belt is decorated with rectangular gilded plaques at the interval of about 8 cm; we cannot discern pendant straps on them.

No. 5. The fourth panel (a feast scene in town, in a pavilion) is placed separately from the first three exclusively “Sogdian” ones and is obviously not related directly to the two Turkic personages depicted in the other compositions. In the upper tier we see two Turks sitting in a small covered pavilion on the carpet, with one leg dangling from the coach edge: an aristocrat and (to the left) a musician with a lute (and also a Sogdian with a harp; besides, there is a Sogdian servant coming with a jug of wine from the side). Behind them there are Sogdians and Turks in groups of three (only upper part of the body depicted) raising their hands in the greeting (Sogdians – their left hands, Turks – their right ones). The contours of the costume details engraved by the carver often do not coincide here with the contours of paint laid later (paint was clearly used to correct the inaccuracies made by the engraver and then noticed by the customer). Every Turkic banqueter has (where we can see it) one belt without its details. In this scene (to differ from the above considered) the master painted moustaches black.

A Turk seated on the carpet with some small item in his hand (a glass? fruit?), is the central figure there (fig. 9); accompanied by two musicians he is a little taller than the others and is the only one who has a short beard. This man differs from the other Turks in a very light shade of his clothes coloring (light pink, almost white), with yellow edging. From under the bottom hem we see red trousers tucked into black boots. The cuff of the left sleeve is undone and it is seen that it is longer than the arm itself (such unfastening of long sleeves leaving the hand completely covered is documented in the art off many peoples in Eurasia and in situations having a sacral meaning (important banquets, most solemn prey, ritual dance)11. Thin moustache is twisted up (undoubtedly, pomaded at that). Both the main personage and the Turk with the lute sitting next to him (i.e. the two personages in the honorable place), have caftans a little longer their knees and completely done up (to differ from those standing behind); their plaits are rather short (similarly to the couple in compositions 1-2). The musician is wearing a brown caftan and, which is rare, with double coat-breast edging (brown from outside and red from inside), his caftan is as long as his master’s one (these additional details differ him from the others behind).

The three Turks standing behind are depicted in short caftans (fig. 10) (the caftan of the left one is knee long, and the bottom hem is not decorated), the caftans are undone on the breast, which is typical for a non-official manner). The Turks standing to the right and to the left are clad in brown caftans with red cuffs, the central one has a red one on with black cuffs. Their caftans are without lapels but have wide coat breast edging (the two Turks to the left have the edging of the same color as he caftan, the caftan of the Turk who is standing to the right has white edging). Later on the upper edge of the collar of the two side personages (in brown caftans) painted by the engraver was additionally painted across with two horizontal redlines (“shoulder straps”). They may denote the edge of a small cape. The upper edge of shirts with a triangle neck is seen from under the caftans and is left unpainted (which means of light color the same as the body of the personages in the brown caftans and the caftan of the personage in the red upper garment). They are all beardless and besides plaits are wearing only small thin moustaches. But while the moustache of the right one is hanging, the two left ones are wearing theirs twisted up (and , probably pomaded).The left Turk is wearing light trousers (not colored the same as his body) which are rather narrow with oblique black stripes, and which are probably tucked into his footwear (half boots or shoes).

Marble mortuary bed from the Miho Museum in the Japanese city of Shigaraki

(See, for example, a series of articles in the collected articles of the museum: Juliano, Lerner 1997, No. 125; Marshak 2004; Raspopova 2004; Rong Xinjiang 2004). The exact placing where the robbed grave was found is unknown. The panels of the mortuary bed depict the life of Chinese Sogdians among their servants and their Sogdian women (these women according to Zhoushu, were dressed in Chinesewhile their husbands are shown in Cenral Asian garb: see Grener, Riboud 2007). There are Turks depicted on two of the panels.

One of the panels (fig. 11) depicts a mounted procession of noble foreign ambassadors (?) whose costume is neither Sogdian, Chinese, or Turkic; it is unknown from the Xinjiang oases and apparently does not belong to any Iranian-speaking peoples from Khotan in the west (it is also documented in the other panel where we also see dismounted servants and loaded camels).

At first sight an ethnic identification of this ethnos is not difficult since we know from the Chinese sources those countries which countries China had diplomatic and trade contacts as well as those which supplied China with population for “barbarian” quarters in Chinese towns at the beginning of the Middle Ages. However, the problem is that the costume worn by the nobility of many of these peoples in the 5th – 6th cc. is practically unknown (Kashgar, Ferghana, Chorasmia and others). However that may be, the influence of Late Sasanian Iran can be observed in the dress and outfit of these foreigners (upper garments put on over the head with high cuffs and the seam line below armpits, special straps embracing the upper part of the body, ribbons decorating showy umbrellas and horse harness (Yatsenko 2006a, fig. 158, 27, 32; 161, upper; 162-163). The narrow headbands of personages of this panel, with a large medallion are original. They are neither Sogdians, nor Tokharistanians. Instead, it may be proposed that the embassy depicted here and accompanied by Turks occurred during the period of long international stability of the situation which followed peace between the Kaghanate and Iran in 571. The embassy arrived in China before the united Sui Dynasty came to power in 582, which restricted political influence of Sogdians. It makes the dating of the funeral coach from the Miho Museum more exact, n the bounds of 10 years. It was during the period of Tapsar Kaghan’s rule (572-582), when the dependence of both Chinese kingdoms from the Kaghanate was, probably, the most strong.

No. 6 (fig. 11). An evidently Turkic personage in the lower part of this panel is depicted alone. It is a horseman accompanying the embassy. His costume is practically identical in the smallest details (including costume items coloring) to the costumes of those personages from the next panel (No 7), who are giving presents (or serving drinks?) to the aristocrat in the yurt, but his clothes are much shorter. We can suppose that similar clothes are depicted in the same panel on the mounted hunter, but he is partially hidden by the figure of his neighbor. To differ from the above mentioned personages his footwear (black half boots) is better seen.

No. 7 (fig. 12). The second panel of this mortuary bed (Miho Museum 1997, p. 250, fig. C; Juliano, Lerner 2001, p. 309, fig. K) is of even greater interest. It shows the life of a Turkic aristocrat and his attendants visited by an embassy of Chinese Sogdians came to. In the composition of both banquet scenes with the Turk sitting in the yurt in Chinese-Sogdian engraving of the funeral couch from the second half of the 6th c. (described above in the second panel of An Qie’s funeral coach and the panel under consideration from the Miho Museum) we see the same iconographic scheme. In both cases there is a yurt with the main personage in the background (the yurt is either completely white or, as in the latter case, mainly white with red and black decor; those two yurts have some differences in the construction of the top part and the size of the entrance). The sitting Turk is being served with vessel with a drink (?). To the left from the yurt there are 3-4 people from the escort sitting on the red round carpet, there are vessels with drinks in front of them (jugs, goblets or wineskins) and food; a large tree is growing to the left from the yurt; horses (for Turks) or camels and donkeys (for Sogdians) loaded with delivered provision are situated closer to the viewers; in the center of the composition we see the standing servants who has prepared drinks and refreshment and are guarding pack animals (Sogdians have one more man who is preparing food sitting to the right). In the foreground (along the lower edge of the picture) we see schematically presented mountain peaks (in An Qie’s mortuary bed’ mountains are in the background forming some mountain valley round the campsite). It is interesting to mark an interesting distinguishing feature: whereas in the first episode all the personages except for the guest in the yurt are Sogdian (in An Qie’s funeral bed Sogdians generally outnumber Turks in all the scenes), in the second one, on the contrary, all the personages are Turks. The latter regularity is also marked in the other panels of the funeral bed from the Miho Museum (the presence of only foreigners - non-Sogdians, non-Chinese in the two panels depicts western embassies).

Let us consider this scene from the life of nomadic Turks. The noble Turk in the yurt (his figure is approximately one and a half times as big as the others) makes the compositional center of this panel. Unfortunately, most of the Turks (five of nine) are shown from back (only the main personage is sitting en face, and the one standing in front of him is shown in profile). In the lower tier we see mounted Turks hunting in the mountains. The caftan of the yurt host has a wide decorative edge on the coat breasts; it is white with lines of red arch-shaped patterns (the only depiction of cloth ornamentation for Turks in the monuments under consideration to differ from Sogdian ones!), the shirt and the trousers are of the same white color. The caftan is undone in the upper part (it is less clearly seen in the image of the mounted hunter in the foreground).

The clothes in this group are preserved only in three colors: red, white and, partly, black (the lack of paints being not the reason). There were also, probably, blue and gilding [according to J. Lerner] before the panels were inattentively cleaned. Close examination of the stones themselves shows that there were several other colors). We see here in the scene of hunting short caftans (knee-length) on the horsemen, and people standing at the yurt are wearing ankle-length oriental robes (long sleeved coats)(i.e. a “horse riding’ and a festive complete sets. As far as the color of clothes for the upper part of the body is concerned we, very likely, deal with a social convention. The clothing is white for more significant personages, such as in group1 (the aristocrat in the yurt, the horseman in the foreground, the main attendant serving the ruler, the central one of the three guests sitting on the carpet) and red for less important characters (some guests of the aristocrat, the servant with horses, the horseman in the background). The decorative scheme of caftans is unified here – borders edging the coat-closures, cuffs and hem (red for white clothes and vice versa). As it has already been mentioned above, it was on the whole typical for Sogdians.

We see white trousers on the most significant Turks (the noble in the yurt, the horseman in the foreground). The belts on the Turks which are possible to be seen are white without pendant details and plaques (made of fabrics, most probably)12.The footwear is very specific in its details: high boots with triangular projections under knees on the yurt host (the only case in the Turkic series of that period) and red half boots on the horseman in the red caftan. All the Turks are without beards; their moustaches are the same - long horizontal ones (evidently, pomaded). The plaits are rather originally styled. They are as long as the hips (the same long plaits are observed in group 2) but in this case the engraver considered it necessary to show each plait separately. Notwithstanding poor foreshortening of most figures it is evident (judging from the depictions of the two horsemen whose heads are shown in profile and we clearly see the upper parts of their plaits) that they had seven plaits (which is often observed in Turkic statues), the plaits are tied with a white strap at the back part of the head. They were somehow joined together at the lower part as the general contour is shown to be a little narrower there.

No. 8. Sarcophagus of Yu Hong, who died in 592, was discovered in the northern city of Tayuan (fig. 13). That Sogdian was an official in the rank of “sabao”, and also an ambassador of the Sui Dynasty (which united China in one state) who held diplomatic positions in several countries (for example in Sasanian Iran). The most scenes of the sarcophagus are “Zoroastrian” (to be more exact – Mazdeist Sogdian) ones depicting kings, gods, saints and not numerous servants though the influence of the Buddhist art of the former Northen Qi Kingdom (Juliano 2006) is evident. The only Turkic personage is presented in eastern panel No. 4. Unfortunately, only insignificant traces of paint have been preserved (fig. 14), thus it is better to consider the images in the detailed representation in the original Chinese publication. The man is shown here turning back (fig. 15). It is a camel rider and he is the only one among the personages in different panels who is not a lucky hunter (compare: Short Report 2001, p. 38, fig. 20; p. 36, fig. 17). The “negative” representation of the only Turkic personage can be interpreted as the result of the long-lasting recent war (582-589) between the Kaghanate and the Sui Dynasty.

The costume of this Turk is very unusual. Locks of his long hair are joined together at the ends, there is a belt with additional hanging down straps: they are the only authentic Turkic costume elements of the personage. He is wearing shoes and trousers with narrowing down legs which are not tucked into his footwear (some red coloring on the trouser legs has been preserved). The short upper garment has short sleeves flare out at their end, the undergarment is longer and has the usual narrow sleeves. The edges of the cuff edges are scalloped. Such specific combination of two garments worn on the upper part of the body was extremely rare for Sogdians themselves at that period (Yatsenko 2006a, fig.181, 49-50; 182, 19) but was typical first of all for Tokharistanian-speaking population of Kutcha oasis in Xinjiang (Yatsenko 2000, fig. 58, 9; 60, 5 ; 62, 3; fig. 75, 2). The inhabitants of Kutcha for famous for being talented musicians, dancers and composers in China and went there to earn their living. Another original costume element here is a bunch of five short ribbons hanging from each shoulder and typical for noble men exclusively in Kutcha (Yatsenko 2000, fig. 59, 2-3) (officers usually worn then only on one shoulder: Yatsenko 2000, fig. 60, 1-2). The decorative breast ribbon (Ibid., fig. 59, 1, 3-4) the trousers and shoes types are also characteristic features of the Kutcha costume. Plaits of the personage are joined together in an unusual way: their lower parts are tied with a string forming a pointed end, the analogy which is nowadays known only in the art of the same Kutcha (the singe image of the Turk here is presented in the painting “peoples mourning Buddha” in Kizil) (fig. 16)13.

It is a very interesting example of strong costume influence of a small and not powerful at all but highly cultural Kutcha Oasis on Turks. What can the reason be for such en effect (the influence on a small group of Turks or even one certain person only) be? Political immigration? Military alliance? Gifts of clothing made for a specific occasion? We do not know and the image remains a puzzle.

Chronological group 2: the 7th – the first half of the 8th cc. (the beginning of the Tang Dynasty).

Clay figurines of supposed early Turks from burials correspond to the early Tang dating. Unfortunately, the dating of terracotta figurines considered below (except No. 14) could only be approximate as they originate from looted burials. We often see different “northern and western barbarians” in funeral terracotta figurines of that time. However, there are very little images of supposed Turks from the period of the Tang Dynasty if compared with representatives of neighboring sedentary “western barbarians”. This can be explained by the fact that small groups of Turks had just begun to settle in frontier regions and they were not numerous in large cities. Their participation in the different spheres of Chinese life could be rather restricted then. Yet, there was a decision to settle 10,000 Turkic families in the capital of united China as long ago as the 720’s, before the half century-long submittance (630-679) of eastern Turks to the Tang Empire.

As has already been remarked, images of early Turks have not been recognized elicited among numerous images of “barbarians” of that time until 2008. Such a long uncertainty about the question was being explained by at least five factors:

First, the problem of ethnic identification of images of foreigners in highly developed artistic traditions of classical antiquity and early medieval countries of Eurasia (first of all ancient Greece, the Roman Empire, the Persian Empire, Sogd, Northwest India and China) is very complicated. The situation is aggravated by our fragmentary knowledge of modern knowledge about the costume of even large and well-known peoples of the past. To be more exact ethnical attribution of foreigners in literature is given very often but it is not usually correct at the same time. The reason of it is very simple: the authors are not specialists on the ancient costume of several regions, they are not acquainted with the most part of materials on each definite ancient ethnical costume complex (they are mostly unpublished) and, as a result, they cannot competently apply analogies and costume classification and the adequate comparative analysis is impossible.

Second, local artistic stereotypes and stereotypes in the perception of foreigners on the whole (the social rank of migrants should be taken into account) were reflected in the degree of stylization of the costume (see, for example, Yatsenko 2006a, p. 31-33, 47, 55-57, 61, 65-67, 113-115, 129-130, 134, 166-169, 229-230, 245-247, 261-262, 273-274). They have not been properly studied yet. Many funeral figurines belonging to the beginning of the Tang Dynasty are schematic: they lack a number of details so the information on the costume is indefinite14. The images of “barbarians’ are very often stylized and have a standard coloring (artists used more often, besides in addition to glazes of several colors, red or black pigment on unglazed surfaces).

Third, if to speak generally about peoples whose representatives turned out to be in China at that time (about “western and northern barbarians”), we usually see representatives of nobility in the art of their native territories. However, people who usually came China belonged to (with the rare exception of political émigrés, hostages and diplomats) to other social groups (e.g., traders, dancers, actors, artists and others). We encounter their images in their own art only on rare occasions. That is why it is possible to identify only some of costume elements that are, probably, common to different social groups. For this reason ethnical identification is possible only for a part of funeral figurines of the early Tang Dynasty period, and it is to a certain degree approximate for a larger number of “barbarian” figures. Our attribution is based on the correlation between several types of clothing, hairstyle, footwear, and belt elements specific for certain Iranian-, Turkic- and Tokharian-speaking peoples of Central Asia. Many personages represented by the early Tang clay figurines can not be identified as their simple hairstyle and clothes are typical for several regions in Central Asia. Only statuettes in the “civil” costume are analyzed here (personages in armory need special consideration).

Fourth, many ‘barbarians” who lived and worked in China, borrowed some Chinese costume elements rather quickly and we should accept this “Sinicization”. Judging from depictions of the 7th-8th cc. it may be seen in the following: 1. Short open down the front clothes often wrapped over from the left to the right (but the overlap was not deep) the same as in Chinese clothes but not from the right to the left (as it was traditional for early Eurasian nomads15). 2. Different variants of Chinese men’s headdress ‘putou’. They are all date back as long ago as the 7th-8th cc. It is necessary to note that later, from the turn of the 9th-10th cc., they were used by Uygur commoners of Xinjiang (Yatsenko 2000, p. 367). People wearing ‘putou’, were not young. 3. In some situations open down the front clothes have very wide sleeves. 4. The kerchief is adjusted to the belt in a specific Chinese manner. Chinese costume elements appeared, first of all, as a result of those commoners’ long living in China. But the depictions analyzed below there is not a single authentic Chinese element in the clothes and hairstyle and this is a very significant peculiarity.

Fifth, at the beginning of the Tang Dynasty some elements of early Turkic costume (a specific oriental robe, a belt with suspended items and high boots) came into fashion among Chinese city dwellers (including household servants of both sexes; court ladies as a horse riders). Thus, Prince Li Cheng adored exotic Turkic clothes and other attributes of nomadic life from his childhood.16. However, important details of his costume (the headdress, the hairstyle, the overlap from the left to the right) preserved Chinese look. It usually helps to distinguish sinicized “barbarians” from Chinese (Han?) household servants in pieces of art.

Funeral clay fugurines belonging to the epoch of the Tang from Ezekeil Schloss exhibition catalogue of 1969.

In my opinion images of “barbarians” from this catalogue usually present Sogdians and Tokharistians, less often people from Khotan and Kutcha in Xinjiang (their images will be considered in detail in a special publication).This conclusion can be confirmed by Chinese written sources on foreigners in China at the Sui and Early Tang period. There is only one recognizably Turkic personage among them (Schloss 1969, No 42). From my point of view some still mysterious personages may also be identified as Turks by their costume in the course of time (Nos. 7, 10а, 56, in the catalogue).

No. 9 (fig. 17). The costume of that “Sogdian horseman” in a half-seated posture with bridles in his hands is typical for a nomad from Central Asia of that time. We see a short and narrow (knee-length) caftan with two large triangle lapels; not high leggings fastened with straps to the belt from sides (there may be shoes of half boots underneath). The caftan is fastened practically without overlapping of lapels up to the upper part of the breast, there is no decorative border on it.

The horseman has a very peculiar headdress on with a rather high rounded top and a wide and long back piece. The front piece décor presents a large stylized figure of an eagle spreading its wings. There is a headdress analogous in its form to this one known at the time in Eastern Turkic regions of Mongolia: it is a relief from the eastern stone edge (of a “sarcophagus”) from Hul-Ashete I (Voitov 1996, fig. 67); an analogous image of an eagle spreading its wings is also known on a hat of another design (without the back piece) belonging to the east Turkic Kul-tegin kaghan (the beg. of the 8th c.) on the statue from his palace (Novgorodova 1980, fig.194) (fig. 18). It is difficult to suggest that such headdress could be worn by a common (not noble) personage.

Unfortunately, the photo is published only in black and white. The author of the catalogue says that the head of the personage is not glazed, the face is painted black and red; the body is painted green (the caftan?), brown (trousers?) and a lighter color – creamy (leggings?). The green color of the caftan is very interesting here. The point is, it is not typical at all for Early Turkic costume according to Xuanzang who watched in 630 the Western Turkic Kaghan Ton-yabgu with his escort in an everyday situation during hunting specially marked the exclusiveness of a green caftan for him only (Chavannes E., 1903, p.194). Judging from depictions caftans of just green and brown color became popular in neighboring China in the epoch of the Tang Dynasty. It is interesting to know that later at the time of the Song such clothes exactly of green and also bright red colors were perceived by the apologist of national traditions, a famous scholar of the 11th c. Cheng Kou as the most unbearable manifestation of the barbarian trend in the Chinese costume at the time which had spread since the epoch of the Northern Qi (550-577) (Kryukov, Malyavin, Sofronov 1984, p. 150). But the special popularity of green for men’s upper garments in the 7th c. we observe in the above mentioned Kutcha Oasis which bordered on the territories where nomadic Turks roamed and camped (fig. 19).

Figurines of the Tang Dynasty epoch from private collections in catalogues of Sotheby’s and other auctions and exhibitions.

No. 10 (fig. 20) This personage is wearing the same type of the headdress as found on the previous terracotta. The caftan is also knee-length, with three triangular lapels and without any decorative border, it is fastened in the lower part and has some traces of white coloring. His footwear is the black high boots (without a triangle projection below the knee).

No. 11 (fig. 21). The type of headdresses presented here is similar to a silk (?) bashlyk of an ochre color with a long and wide back piece. This type was trustworthily used by early Turks in Mongolia (see, for example, Evtyukhova 1952, p. 99, No. 79), but there is no documentation for it among the sedentary peoples of Central Asia at this period of time it is not documented. The origin of this headdress connected with the powerful Xianbei people, which controlled also the north part of China and was popular there in the 5th-6th cc. (fig. 27). We also know similar, but rare, very long ankle-length caftans (see, for example, on carved stone images in the Tien-Shan Mountains: Eleukenova 1999, fig. 70); in this case they are worn thrown over the shoulders. Their blue-yellow color range (the right side and the lining) is known at that period in Tuva (Kubarev 2005, p. 40), but unusually long sleeves (much longer than actual arm length) are still unique for early Turks. It is clear that they are not “servants” (as they are attributed by the author of the catalogue) but noble people in a very spectacular aristocratic costume (with a very long body and worn over shoulders caftans with very long sleeves, and shirts not convenient in everyday life). However, the most interesting thing is that white is the color of the majority of the costume items: underwear, trousers and even shoes and a belt17; we will speak about it later. The depicted personages are standing with arms folded on the breast (we can see similar things, for example, on the carved stone image from Kemen-Ketchu in the Altai: Kubarev 1997, p. 44).

No. 12 (fig. 22). Some important details of the upper garment of this elderly man are known to us in the ‘Turkic complex’ from Tokharistan (see: Yatsenko 2006a, fig. 189, 3, 37, 41, 85). His silk (?) headdress (the same as in fig. 21) is typically nomadic.

Such clothing for the upper part of the body was used at that time by early Turks. The figure’s caftan with two lapels also worn thrown over shoulders, falls below his knees, with is a wide border along the hem and a narrow edging along the vertical opening (including the lining); along the edge of the collar we see a thick twisted strip with two buttons (made of fabric?) at the end the lapels. Other costume details do not contradict such interpretation. Only his comparatively long beard is rare for this ethnic identification; his exaggeratedly bushy eyebrows and lashes make him look strange18. It is without doubt not a commoner (not a servant and so on). Similarly to the previous picture his trousers are evidently tucked into his shoes. We see bracelets accentuated on his wrists (the only accessory, besides the belt, common for all the depictions under consideration in this article).

No. 13 (fig. 23). The upper garment of the figure a fur coat without a collar, with the upper part of the coat slanted downwards, is made of rough lambskin (such lambskin coat is known to have been partly preserved in barrow 9 of necropolis Djolin I: Kubarev 2005, p. 29). Here we see the summer variant of wearing it fur with the out (which was usual in everyday life, but unique in pictorial representations!) .The upper part of the coat is slanted down but does not form lapels. Between collar breasts a low neck of the shirt can be observed. The hair is rather short. The footwear is presented by leggings of dark color fastened at tights, but not high boots. The clothing lacks any decoration as it was without accessories. It is a rare example of a commoner (a shepherd?), similar to those who settled at that time along the Chinese frontier.

Museum collections of Xinjiang Uighur Autonomous Regon of China.

No. 14 (fig. 24). The figurine was found in Astana Borrow of the 7th-8th cc. in Turfan Oasis. It depicts a bearded man in a typical Xianbei people’ hood (bashlyk) of the same type as in figs. 17 and 20, and which is black in color. A very long caftan without lapels is fastened along to the neck. A stripe of cloth, probably, of another color, is stitched to the caftan along the hem (there are traces of white color: see below the analogous caftan of the personage brining some refreshments to the aristocrat in the yurt in fig. 12). The upper part of the shirt is marked outlining a triangle neck.


Images of the second half of the 6th c. AD

The materials of the earliest detailed painted depictions (belonging to the period of the united Great Kaghanate existence) present a special value for the history of the Turkic costume. With the only exception, the caftans of depicted Turks are either fastened up to the top (on the most solemn occasions – for main personages or at the feast) or unfastened in the upper part only. Obviously, they are sometimes slightly overlapped from the right to the left (which is traditional for early nomads and corresponds to widely spread witnesses of ‘Zhoushu’ chronicles)19. To differ from Sogdians (whose important persons obligatory wear headdresses), all Turks are bareheaded (which we not always see on Turkic images namely).

As has already been mentioned 4 complexes of the Turkic costume have been singled out by me in early depictions. In different panels of An Qie’s funeral coach, 579 we see as I resume three groups of Turks whose clothes and hairstyle are different and they, probably, may be referred to three different tribal groups being part of the Great Kaghanate and with whom An Qie had negotiations in the steppe and in town. The first three neighboring panels (containing episodes connected with two Turkic personages, conventionally called the master and the servant) can be referred to group 1. The third panel (the scene where An Qie offers a Turk some wine) presents another couple of Turks where the main personage can also be singled out (group 2). The panel with five Turks at the feast (the main one is sitting in the centre of the carpet, some musicians are entertaining him) present group 3. And at last a very specific group (group 4) in respect of the costume is presented in the “Turkic” panel of the funeral coach from the Miho Museum and in the only personage in the panel with a horse procession (“the scene from the life of Turks”). Let us analyze costume differences in these groups.

Group 1

Both personages of different ranks are depicted in three scenes supplementing each other. It is significant that the action takes away from any settlement and is connected with hunting and rest in the yurt. They are both clad in a short caftan (knee length) which is completely closed on official occasions but in informal situations hunting) it is undone in the upper part (the “servant” is the only one among other early images who has two triangle lapels).

The master’s silk (?) caftan is on all occasions white with a red border along the coat breasts, hem and sleeves; the “servant” is clad in a red caftan and there is a black border only along the hem. They both are wearing a red undershirt with a wide horizontal collar. The master’s trousers are of medium width, white with oblique black stripes or flesh-colored (not covered with paint the same as the surface of human’s skin). The trousers are worn not tucked into low footwear and may have a border along the trouser leg edge. The “servant’s” black footwear presents high boots without a projection above the knee, the master has half boots, (it is clearly seen in the scene of dice game – fig. 6). The master’s black belt without metal details is made of fabrics and is wrapped round his waist two times. The strap of his plated belt is evidently decorated with gold (or bronze gilded, as on the authentic belt of An Qie) rectangular plaques, it has additional hanging down straps (with a dagger suspended from the left side). However, the master‘s belt of such kind is seen only (together with a quiver belt) but only in the scene of hunting; in other situations (being An Qie’s guest) it was not evidently worn. The “servant’s” leather belt with additional straps has a metal (gilded) buckle only, decorative plates are absent; there is a small round bag-kaptraga fastened to the belt.

It is important that a whole number of details is specific for the Sogdian costume (though belonging to the later period of time: the 7th – fist half of the 8th cc. – the epoch of Sogdiana’s partial Turkization.): the décor with a standard red border along the coat breasts, hem and sleeve edge, trousers with a pattern of oblique black stripes also having a red border along trouser leg edge and worn not tucked into footwear. What is more important is that all theses Sogdian analogies are observed only in the clothes of the master. It makes it possible to come to the conclusion that the aristocracy of the group of Turks in question borrowed these decorative elements from the costume of Sogdian colonists.

Group 2.

We see depictions of representatives of this group in the banquet episode in town. Here, in a scene of festivity. The caftans or coats of Turks are completely closed, the footwear is the same (high boots for the “servant’ and some shorter (?) footwear for the master); one sleeve is tucked into the other one. The upper body garment of the both people has a wide turn down collar. They both have no beard or moustache. The master’s upper body garment is longer and is of brown color (silk?), with a red border; the ”servant’s” clothes, the same as it was in the depictions mentioned above, are red. The plaits are much longer if compared with group 1, they reach pelvis. Similarly to group 1 the master is wearing a black belt without metal details or pendant straps; it is wrapped round his waist two times (i.e. it is a festive one made of fabric). At the same time the “servant” is wearing a short plate-decorated belt with a rather big number of gilded rectangular plaques.

Group 3

It is presented in a single panel not connected with other panels showing Turks. We see here not two but five Turkic personages. The two people sitting on the carpet (the master and the musician) are wearing caftans longer than knees and (according to their role at the banquet) completely done up; the personages standing behind are clad in short knee length caftans undone on the breast (for two of them we see red pelerine edges completed in paint(?). The Turks staying behind have shirts with a small triangle neck. In this panel the artist gave his special attention to the decor of coat breasts in upper garments which is quite various (the musician has double coat breasts, the master has yellow ones, the two personages above to the left have coat breasts of the same color as their caftans are, the personage to the right has white ones). To differ from other groups the Turks here (all of them) are wearing moustache each of them having a different type: pomaded (horizontal or turned up) or hanging down. All the personages are wearing black (leather) belts without metal details. The main Turk, similarly to the one in group 1, is differentiated by a beard, very light (light pink) upper clothes and the color of trousers (red in this case) which are seen from under the caftan. The personage standing to the left is clad in light trousers with oblique black stripes (the same as in group 1). The both people sitting on the carpet have high boots (not very high) without a projection above the knee; the standing personages are evidently wearing shorter footwear. The most remarkable distinctive features of this group are the form of the shirt neck, wearing moustache, the absence of metal details of belts, pelerines (?), variety of coat breasts and, in one case, a double line of coat breast décor.

Group 4

We see the costume of this group in two panels on the funerary bead in the Miho Museum. The dress in this group is more homogenous for people of different social ranks than in other groups (I do not think it to be the result of stylization ordered by the customer or produced by the artist). The clothes here are painted in three main colors: red, white and, partly black. There is a complete set of the “horse rider’s” costume (with short knee length caftans on the mounted hunters) and a “festive’ one (ankle length caftans on the people standing and sitting round the yurt). All the clothes are undone at the neck. The colors of upper body garments are: white for more significant personages (such as in group 1) and red – for less important people (some guests of the aristocrat, the servant with horses, the horse rider in the background). Significant personages also wear white trousers. The decorative scheme of caftans of people belonging to different social ranks is unified (bordered coat breasts, hems and sleeves); this tradition is definitely typical for Sogdians. The caftan of the master in the yurt is unique (it is all decorated with pattern of vertical lines formed by red arches. All the men wear white (made of fabrics) belts (absent in other groups). Methods of plaiting hair differ in hairstyle of other groups. They are, analogously to group two, as long as pelvis, but here in a number of cases separate plaits are singled out; there were seven plaits (evidently, four ones from the left side of the head and three from the right); in the upper part their locks are tied with a white strap. All the Turks are without beards; their moustaches are similar – long horizontal ones (obviously, pomaded). Their footwear is also very specific in details (the master’s black high boots with a triangle projection below the knee and red ankle boots on a less important person).

The costume analysis of the four costume groups of Turks in depictions of the second half of the 6th c. (the period of the united Great Kaghanate), firstly, makes us reject the today’s approach to consider them as a whole and admit a number of important local (tribal) peculiarities; secondly, we have to pay a special attention to comparing the costume with depictions belonging to the next - 7th c. (when the Turks lost their political unity and, for some time, got under the rule of china).

Images of the 7th – first half of the 8th cc. AD

The described period includes both the time of the Eastern Turks submission to China (630-682), and the periods of independent existence of the Eastern Kaghanate (603-630 and (682-744). The nature of political relations between China and the Turks at that time was changing, but the increase of reciprocal influence in daily culture including the costume is doubtless. The described terracotta can be subdivided into three iconographic types. Whatever period of political relations between the Turks and Chine they reflect, the terracotta demonstrate the cultural influence of half century long submittance of the Eastern Kaghanate to China and the changing status of Turks (their getting to large Chinese cities, their settling in frontiers). It is important that, to differ from representatives of other “barbarian’ peoples in similar terracotta, the personages in question (also as in earlier complexes) have no elements of the Chinese costume.

In addition to that we dispose of witnesses of opposite character in Chinese official texts. Thus, immediately after the disintegration of the united Kaghanate the first ruler of its eastern part (dependant on both Sui emperors) Janghar Kemin (604-608) made an official attempt to introduce the Chinese costume elements for nomadic Turkic subjects. Kaghan Mozhou put on a Chinese headdress ‘putou’ and a purple Chinese caftan (Kuner 1961, p. 190). We observe some personages with the caftan coat breasts overlap from the left to the right among the figures of commemoration complex of the Eastern Turkic ruler Köl Tigin Kaghan, who died in 731. The contradictory character of the “costume” information on Turks from the Eastern Kaghanate and China can be easily explained. The matter is that Turks lived in different cultural background both in their motherland and in China itself (in condition of a foreign country immigrants are at times inclined to attach great importance to preserving elements of their ethnical costume).

Type 1.

In terracotta Ns. 9-10 we see a representative of nobility in the image of a horseman holding the reigns. Here a high headdress with a figure of an eagle is combined with a short caftan having two lapels and high boots or leggings.

Type 2.

A standing aristocrat in a long caftan with lapels or without them and in a hood (bashlyk) of a special type (subjects 11, 12, 14). The personages of terracotta 11-12 are very interesting. They present an original type of an elderly (wearing a beard - 12) aristocrat from the period of the Eastern Kaghanate in a solemn pose and in a standard costume: a long thrown over shoulders caftan with long sleeves, a bashlyk with a long back piece, narrow trousers and shoes. Judging from couple 11, white color dominates in the four-colour range of their costume. Personage 14 is rather similar to them in appearance.

Probably, both groups of personages of approximately the same rank but in different life situations and in different social roles (compare complexes of the “courtier’s” and “horse rider’s”/ military-hunting costumes of noble Persians in the Achaemenids time).

Type 3.

The personage of subject 13 for the first time demonstrates a common representative of nomadic Turkic society in his simple mutton fir coat and high black leggings fastened to the belt, clothes without any décor, a short haircut (evidently without plaits) and without any headdress. The more so important because the art of Turks themselves at that time never condescends to depicting commoners.

Let us try to clear it up. What is the difference between costume depictions belonging to the epoch of the united Great Kaghanate (the second half of the 6th c.) and much later compositions of the 7thcc.?

1. In earlier scenes a large number of personages are shown en face or half pro file (12 figures of An Qie’s funeral coach and 3 figures of the funeral coach from the Miho Museum). However, only in one, evidently secondary personage in subject 2 (fig. 6), we see two triangular lapels, “initially typical” for Turks, of the caftan undone at the neck. This detail of style came to dominate in the next century only, judging from the reviewed terracotta and wall paintings in Afrasiab / Samarkand. They are traditional for Turkic nobility in later terracotta.

2. Both very short (above the knee) (9-10) and very long caftans (full length)(11, 14) are known only in the late period. The manner of wearing them is quite various, only in earlier time they are worn completely done, and in later complexes they are done up exclusively in their lower part. (12), having two lapels they are thrown over shoulders (11-12). The evidently not deep overlapping to the left (with small strings) is marked only in early depictions (1, 5/standing).

3. In depictions of about the middle of the 7th c. in Afrasiab the decorative border of caftans comes to very high cuffs only (they, the same as lapels, are made of polychrome silks with medallions) but in earlier times we see another more complicated decorative border and also along hems and coat breasts. The earlier complicated décor is still met in the 7th-8th cc. in long caftans of eastern Turkic nobility (Ns.12 and 14).

4. In the mid. 7th–mid. 8th cc. (judging from Chinese terracotta figurines, wall paintings in Afrasiab and remains of authentic caftans and shirts in the Altai) color preferences change. The color range of the Turkic male costume includes white and red (and not depicted with simple colors – violet/purple?), sometimes yellow, for most significant persons also green and dark/light blue20 for upper body garments, black 9for some part of footwear and headdresses) – on the whole it all coincides both in Sogd and China, being one more proof of authenticity in color rendering21. Thus, alongside with popular before white and red and rare pink, there appear green, yellow (golden silk?), dark blue, light blue but brown disappears from nobility depictions22. We see the same in their detailed Sogdian portraits in the ‘Hall of Ambassadors’ in Samarkand. Turks in these depictions are clad in monochrome clothes but not in motley ones. (The latter does not mean they completely lack ornamentation as in Turkic burials at the turn of the 7th-8th cc. there are a lot of damask silks, the ornament of which differed from the background not in color but in the play of light: Kubarev 2005, p. 27).

5. In early depictions (both in the Steppes, and in festive town banqueting scenes) we see long plaits and no headdresses on all the personages (the same can be said about all personages depicted in the west – in Samarkand). To differ from them in later terracotta of people, probably settled in China none of noble personages has plaits clearly presented, and they all wear headdresses. How can we explain such difference? I resume Turks connected with China itself according to the norms and life style in the Empire could not always demonstrate their hairstyle23. In terracotta of noble people where there are hoods (bashlyks) with long back pieces plaits could be hidden under them.(personages 11-12, 14); however, headdresses with an eagle figure (9-10), judging from the significantly thickened back piece on figure 9, could also hid such plaits.

6.Only in early depictions (5 – the main personage, 8) do we see trousers red color. This is an ancient tradition among Central Asian peoples, dating back to Scythian times (Zaghunluq in Xinjiang, Pazyryk culture of thee Altai Mountains, Issyk Barrow in Semirechye / Jetysu), and is also known later (Hsiung-nu in Noin Ula, Kenkol culture peoples in Kyrgyzstan and others: Yatsenko 2006a, p. 94). Specifically in group 1 (subjects 1, 3, 5) we observe trousers of white cloth with vertical/oblique black stripes (having analogies in Sogd). Exactly in the early period an original belt of black color is documented for more important Turkic in there meetings with An Qie in scenes 2 and 4. The belt is very long and is wrapped round the waist twice, and, differentiating it from the belt of the “servant” standing near him and the belts of Sogdians present at the meeting, it lacks all any metal details, such as buckles, plaques and other attachments. Undoubtedly, what is at issue is the wearing of a festive belt plaited from black wool, probably, with a fringe at the ends. Analogous woolen belts that is documented for many “ethnographically” Turkic peoples (it is not to be surprising that such belts have not been preserved in more or less complete form in early burials).

7. Leggins fastened with straps to the belt are documented in late terracotta (by the way, their not festive character was not attractive for customers of Turkic funeral statues).

The presence of mainly white or light pink upper body garments for personages pairs of a higher social rank, who are placed in the foreground, and of red for people of lower status, who appear in the background and already remarked upon on An Qie’s funeral bed seems to be important. The same can be said about the color of a garments of the personages of a higher social rank in the panel from the Miho Museum (No. 7). One should think it is not a coincidence: we observe the same in the pairs of standing officials or masters of ceremonies in the ‘Hall of Ambassadors’ in Samarkand (personages18-19 and 26-27 of the western wall: Yatsenko 2007). In numerous polychrome group depictions Sogdians (both in China and in their homeland) such regularity is not observed, so we are dealing with some specially accentuated Turkic specificity. Unpractical for daily nomadic life, white, even in a later times, was associated by the peoples of Sayan-Altai Mountains with purity, nobility and wealth; such clothes were usually worn by rich people on solemn occasions (see, for example, Dongak 2004, p. 226).

Two small series of figures with the detailed costume presented in nine examples demonstrate the great variety of the upper garments or caftan-like apparel (the length, the styling of the collar, the manner of wearing and the décor) typical for nobility (but not of the highest rank) of the Early Turkic society. It does not correspond to the concept of only two standard types of caftan, differing supposedly only in the collar form (Kubarev 2005, p. 32-33). Thus, the caftan can be short, above the knee or up to the lowest part of the torso (Ns. 9-10), knee length (1-3, 6, 8), below the knee (4, 5, 12), ankle length (7) and full length (11, 14). The manner of its wearing is quite various: completely done up (early samples); fastening at the lower part only (12), and thrown over shoulders (11-12), not deep overlap to the left with the help of stripes (early ones - 1, 5/standing)24, two flaps close to each other fastened with a belt and, probably, stripes (7 – the personage in the yurt, 14). In early scenes complete fastening was evidently used on the most solemn occasions, in other circumstances the caftan was undone at the neck. No less interesting is the variety of footwear: ankle boots or, probably, shoes (1-3, 5-6, 8, 11-12), black high boots with a triangle projection under the knee (7) and without it (2, 4?, 5, 10), and also leggings (9, 13). The popularity of ankle boots or, sometimes, shoes (connected, probably, with the status of the depicted) and the only detailed image of high boots with a triangle projection (supposedly considered initially more typical for the ethnos is very significant. The presence of white and black belts without metal details (in both chronological groups), and long black woolen belts in addition to leather plate decorated belts and belts without plaques but with a buckle and metal ends, is very important. It should be noted that there is no evidences of high boots or ankle boots that fasten at the ankles with a strap with a buckled strap (Kubarev 2005, p. 47) even in the most detailed depictions from our collection. Different styles of beards are noticeable on some images, probably associated with the individuals’ age as well as their high status (1-2, 5, 12); in reliefs on the Sogdian An Qie’s mortuary bed his Turkic partner in scenes 1-2 also has a small beard, although in the yurt scene (No. 3) he does not have it. The other figures wear only a moustache, either pomaded (both strictly horizontal or sticking upwards) or dropping, or are clean-shaven.

On the whole, in these early Chinese depictions of Turks we have a reliable source for reconstructing some aspects of their culture as we have reason to think that the figures are shown with great accuracy. In addition, the earliest Turkic images give us exclusively valuable information concerning the early stage of Turkic costume history – the epoch of the united Great Kaghanate. It enables us to conclude that the decorative scheme of the caftan in the 6th c. (or earlier in their historical native land) can under the influence of Sogdians whose colonies reached as far to the east as Manchuria; in single cases we observe undoubted borrowings from the costume of Kutcha (subject 8) and Khotan (two triangle lapels –the “servant” in subject 2) – oases of Xinjiang, bordering on the ancient homeland of the Turks. In the 7th c. the Turkic dress evidently underwent many changes connected with the wide migrations of the Turks and the early stages of their settlement, complicated by their social structure, the deepening the influence of China, Sogdiana and Iran, the growth of the textile trade, and the wish of the elite to borrow prestigious elements of dress from the neighboring great states, but also sometimes by their resistance to foreign influence.


Al’baum L.I., 1975. Wall Painting of Afrasiab (Nastennaya zhivopis’ Afrasiaba). Tashkent (In Russian).

Anjia [An Qie] Tomb 2003. - Anjia [An Qie] Tomb of Northern Zhou at Xi’an'. Shaanxi Provincial Institute of Archaeology. Beijing (In Chinese).

Arzhantseva I.A., Inevatkina O.N., 2005. Once More on Afrasiab Painting: New Discoveries Quarter of Century Years Old (Eshche raz o rospisyakh Afrasiaba: novye otkrytiya, kotorym chetvert’ veka) – Central Asia: Sources, History, Culture (Tsentral’naya Aziya: istocniki, istoriya, kul’tura) (Ed. by E.A. Antonova, T.K. Mkrtychev). Moscow.

Arzhantseva I., Inevatkina O., 2006. Afrasiab wall-painting revisited: new discoveries twenty-five years old - Royal Naurūz in Samarkand. Proceedings of the Conference Held in Venice on the Pre-Islamic Paintings at Afrasiab (Ed. by M. Compareti, E. de la Vaissère). Pisa, Roma, pp. 185-212.

Bayar D., 1991. Stone Statues of Mongolia (Kamannye izvayaniya Mongolii). Abstract of the PhD Dissertation. Ulanbaatar (In Russian).

Bayar D., Erdenebaatar D., 1999. Stone Statues of Mongolian Altai Mountains. Part 1. Ulanbaatar (In Mongolian).

Charikov A.A., 1986. Pictoral Specificity of the Stone Statues of Kazakhstan (Izobrazitel’nye osobennosti kamennykh izvayanii Kazakhstana) – Soviet Archaeology (Sovetskaya arkheologiya), 1986/1.

Chavannes E., 1903. Documents sur les Tou-kiue (Turcs) Occidentaux – Collection of Works of Orkhon Expedition (Sbornik trudov Prkhonskoi ekspedotsii). Vol. VI. St-Petersburg.

Chichurov I.S., 1980. Byzantine Historical Works: “Chronography” of Pheophanes, “Breviarium” of Niketporos. Texts, Translation and Commentaries (Vizantiiskie istoricheskie sochineniya: “Hkronografiya” Feofana, “Breviarii” Nikifora. Teksty, perevod, kommentarii). Moscow.

Dobzhansky V.N., 1990. Plate Decorated Belts of Nomads of Eurasia (Nabornye poyasa kochevnikov Azii). Novosibirsk.

Dongak S.Ch., 2004. Color and Its Symbolics in Tuva Nomadic Culture (Tsvet i ego simvolika v tuvinskoi kochevoi kul’ture) – Memoirs of Tuva Institute of Humanity Studies (Uchenye zapiski Tuvinskogo institute gumanitarnykh issledovanii). Vol. ХХ. Kyzyl.

Dosymbaeva A., 2008. Turkic Heritage of Eurasia of the 6th-8th cc. (Tyurkskoe nasledie Evrazii VI-VIII vv.). Astana.

Eleukenova G.Sh., 1999. Essay on History of the Medieval Sculpture of Kazakhstan (Ocherk istorii srednevekovoy skul’ptury Kazakhstana). Almaty.

Ermolenko L.N., 2003. Could Ancient Turkic Statues Be Colored? (Mogli li raskrashivat’sya drevnetyurkskie izvayaniya?) – Steppes of Eurasia in Antiquity and in the Middle Ages (Stepi Evrazii v drevnosti i srednevekov’e). Part II. (Ed. by Yu.Yu. Piotrovky). St-Petersburg.

Ermolenko L.N., 2004. The Medieval Stone Statues of Kazakhstan Steppes (Srednevekovye kamennye izvayaniya Kazakhstanskih stepei). Novosibirsk.

Ermolenko L.N., 2007. Once More on the Problem of Original View on the Stone Nomadic Statues (Eshche raz k voprosu o pervonachal’nom vide kamennykh kochevnicheskikh izvayanii). – Stone Sculpture and Small Art Plastic Wares of Ancient and Medieval Peoples of Eurasia (Kamennaya skul’ptura i melkaya plastika drevnikh i sresnevekovykh norodov Evrazii) (Proceedings of Siberian Association of Primitive Art. Vol. 3). Barnaul.

Ermolenko L.N., 2008. Pictoral Monuments and Epic Tradition on the Materials of Ancient and Medieval Nomadic Cultures of Eurasia (Izobrazitel’nye materially i epicheskaya traditsiya po materialam kul’tury drevnikh i srednevekovykh kochevnikov Evrazii). Tomsk.

Yesin E., 1979. Turk-i mah chihrah (Turkish Norm of Beauty in Iran) - Akten des VII. Internationalen Kongresses für Iranishche kunst und archäologie. München. 7-10 september 1976. Berlin.

Evtyukhova L.A., 1952. Stone Statues of South Siberia and Mongolia (Kamannye izvayaniya Yuzhnoi Sibiri i Mongolii) – Materials and Studies on Archeology of the USSR (Materilay i issledovaniya po arkheologii SSSR). Vol. 24. Moscow, Leningrad.

Gorelik M.V., 2003. History of the Altai People’ Cap (Istoriya altaiskoi shapki) – Altaica. Vol. VIII. Moscow.

Grenet F., Riboud P., 2007. A Reflection of the Heptalite Empire: The Biographical Narrative in the Reliefs of the Tomb of the Sabao Wirkak – Bulletin of the Asian Institute (New Series), 17, 2003 [2007], pp. 133-143.

Grünwedel A., 1912. Altbuddhistische Kultstätten in Chiněsich Turkestan. Berlin.

Hayashi Т., 2006. Sogdian Influence Seen in Turkic Stone Statues Focusing on the Fingers Representation - Ēran ud Anērān. Studies Presented to Boris I. Marshak in Occasion of His 70th Birthday (Ed. by M.Compareti, P. Raffetta, G. Scarcia). Venice.

Huo Xuchu, Qi Xiaoshan, 2006. Buddhist Art in Xinjiang along the Silk Road. Urumqi.

Juliano A.L., 2006. Converging Traditions in the Imagery of Yu Hong’s Sarcophagus: Possible Buddhist Sources - Journal of Inner Asian Art and Archaeology. Vol. 1. Turnhout.

Juliano A.L., Lerner J.A., 1997. Eleven panels and two gate towers with relief carvings fron a funerary couch – Catalogue of the Miho Museum (South Wing). Shigaraki, pp. 247-257.

Juliano A.L., Lerner J.A., 2001. Monks and Merchants. Silk Road Treasures from Northern China. Gansu and Ninxia, 4th-7th Century. New York.

Khudyakov Yu.S., 1998. Early Turkic Statues of Eastern Turkestan (Drevnetyurkskie izvayaniya Vostochnogo Turkestana) – Ancient Cultures of Central Asia and Saint Petersburg. The Materials of the Russianl Conference Devoted to the 70th Anniversary of A.D. Grach (Drevnie kul’tury Tsentral’noi Azii i Sankt-Petresburg. Materialy Vserossiiskoi nauchnoi konferentsii, posvyashchennoi 70-letiyu A.D. Gracha). St-Petersburg.

Kubarev G.V., 2005. Culture of Early Turks of the Altai (on Burial Monuments Materials) (Kul’tura drevnikh tyurkov Alataya [po materialam pogrebal’nykh pamyatnikov]). Novosibirsk.

Kubarev G.V., 2008. “Bending knee” - type Daggers of the Early Turkic Epoch (Kolenchatye kinzhaly drevnetyurkskoi epokhi) – Cultures of the Steppes of Eurasia of the second half of the I millennium AD. Abstracts of the IV International Conference (Kul’tury stepei Evrazii vtoroi poloviny I tysyacheletiya n.e. Tezisy dokladov IV Mezhdunarodnoi konferentsii) (Ed. by D.A. Stashenkov). Samara.

Kubarev V.D., 1984. Early Turkic Statues of the Altai (Drevnetyurkskie izvayaniya Alataya). Novosibirsk.

Kubarev V.D., 1997. Stone Statues of the Altai (A Short Catalogue) (Kamannye izvayaniya Alataya [Kratkii katalog]). Novosibirsk. Gorno-Altaisk.

Kubarev V.D., 2004. Ancient Steles and Statues in Rituals and Superstitions of Peoples in Central Asia (Drevnie steely i izvayaniya v obryadakh i sueveriyakh narodov Tsentral’noi Azii) – Archaeology, Anthropology and Ethnology of Eurasia (Arkheologiya, antrpologiya i etnografiya Evrazii), 2004/1, Novosibirsk.

Kryukov M.V., Malyavin V.V., Sofronov V.V., 1979. Chinese Nation on the Border of the Middle Ages (Kitaiskii etnos na poroge srednikh vekov). Moscow.

Kryukov M.V., Malyavin V.V., Sofronov V.V., 1984. Chinese Nation in the Middle Ages (the 7th- 12th cc.) (Kitaiskii etnos v srednie veka [VII-XIII vv.]). Moscow.

Kyuner N.V., 1961. Chinese Information on Peoples of South Siberia, Central Asia and the Far East (Kitaiskie izvestiya o narodakh Yuzhnoi Sibiri, Tsentral’noi Azii i Dal’nego Vostoka). Moscow.

Lerner J.A., 2005. Aspects of Assimilation. The Funerary Practices and Furnishings of Central Asians in China (Sino-Platonic Papers. Vol.168. December 2005). Philadelphia.

Lu Enguo, 1999. Ancient Corpses of Subeshi - The Ancient Corpses of Xinjiang (Ed. by V.H. Mair). Urumqi, p. 96-112.

L’vova E.L., Oktyabr’skaya I.V., Sagalaev A.M., Usmanova M.S., 1988. Traditional World Outlook of Turks in South Siberia. Apace and Time. World of Artifacts. (Traditsionnoe mirovozzrenie tyurkov Yuzhnoi Sibiri. Prostranstvo i vremya. Veshchnyi mir). Novosibirsk.

Maadai-Kara. Altai People’ Heroic Epic Poem (Maadai-Kara. Alatiskii geroicheskii epos), 1973. Moscow.

Mahler J., 1959. The Westerners among the Figurines of the T’ang Dinasty of China. Rome.

Marshak B.I., 2001. La thematique sogdienne dans l’art de la Chine de la seconde moitie du VIe siecle - Academie des Inscriptions et Belles-Lettres. Compur rendus des séances de l’annee 2001 janvier-mars. Paris.

Marshak B.I., 2004. The Miho Couch and the Sino-Sogdian Works of Art of the Second Half of the 6th Century - Bulletin of Miho Museum. Vol. 4. Shigaraki.

Miho Museum. South Wing, 1997. Shigaraki.

Mode M., 2006. Reading the Afrasiab Murals: Some Comments on Reconstructions and Details - Royal Naurūz in Samarkand. Proceedings of the Conference Held in Venice on the Pre-Islamic Paintings at Afrasiab (Ed. by M. Compareti, E. de la Vaissère). Pisa, Roma, pp. 107-128.

Mogil’nikov V.A., 1981. Turks (Tyurki) – Steppes of Eurasia in the Medieval Epoch (Stepi Evrazii v epokhu srednevekovya) (Ed. by S.A. Pletneva). Moscow.

Northern Zhou tomb of An Qie excavated in X’ian (In Chinese) - Wenwu. 2001. No1. Beijing (In Chinese).

Nowgorodowa E., 1980. Alte Kunst der Mongolei. Leipzig.

Raspopova V.I., 2004. Life and Artistic Conventions in the Reliefs of the Miho Couch - Bulletin of Miho Museum. Vol. 4. Shigaraki.

Road of Silk, 2007. – Road of Silk. 5000 Years of the Art of Silk. Cataloque of the exhibition. St.-Petersburg.

Rong Xinjiang 2003. The Illustrative Sequence on An Jia’ Screen: A Depiction of the Daily Life of a SabaoOrientations, 34/2, pp. 32-35.

Rong Xinjiang, 2004. Daily Life in Sogdian Colonies along the Silk Road - Bulletin of Miho Museum. Vol. 4. Shigaraki.

Tasmagambetov I., 2002. Kulpytas [Funeral Monuments]. Astana.

Royal Naurūz 2006. – Royal Naurūz in Samarkand. Proceedings of the Conference Held in Venice on the Pre-Islamic Paintings at Afrasiab (Ed. by M. Compareti, E. de la Vaissère). Pisa, Roma.

Schloss E., 1969. Foreigners in Ancient Chinese Art from Private and Museum Collections [Cataloque]. March 27 through May 25, 1969. New York.

Shagalov V.D., Kuznetsov A.V., 2006. Catalogue of the Coins of Chach. The 3th cc. Tashkent.

Sher A.Ya., 1966. Stone Statues of Semirechye (Kamennye izvayaniya Semirech’ya). Moscow, Leningrad.

Short Report on the Tomb of Yu Hong, the Sui Dynasty, excavated in Taiyuan, Shaanxi - Wenwu. 2001. No1 (In Chinese).

Study on the Costume and Adornment of the Nationalities in Western Regions of China (Ed. by Li Xiaobin). Urumqi, 1995 (In Chinese).

Vainshtein S.I., 1991. World of Central Asian Nomads (Mir kochevnikov Tsentral’noi Azii). Moscow.

de la Vassière E., 2006. Les Turks, rois du monde a Samarcande - Royal Naurūz in Samarkand. Proceedings of the Conference Held in Venice on the Pre-Islamic Paintings at Afrasiab (Ed. by M. Compareti, E. de la Vaissère). Pisa, Roma, pp. 147-162.

Voitov V.E., 1996. Early Turkic Pantheon and the Model of Universe in Cultic Funeral Monuments of Mongolia of the 6th-8th cc. (Drevnetyurkskii panteon i model’ mirozdaniya v kul’tovo-pominal’nykh pamyatnikakh Mongolii VI-VIII vv.). Moscow.

Wang Linshan, Wang Bo, 1996. China’s Cultural Relics of Grassland in Altay. Urumqi.

Yatsenko S.A., 2000. Chapter 3, “Costume” (Kostyum) – Eastern Turkestan in Antiquity and Early Middle Ages. /Vol. 4/. Architecture. Fine Art. Costume (Vostochnii Turkestan v drevnosti i rannem srednevekov'e. (T. 4). Arkhitektura. Iskusstvo. Kostyum (Ed. by B.A. Litvinskii). Moscow.

Yatsenko S.A., 2004. Costume of Foreign Embassies and Inhabitants of Samarkand on Wall Painting of the 7th c. in the Hall of Ambassadors from Afrasiab as a Historical Source - Transoxiana 8. Buenos Aires.

Yatsenko S.A., 2006a. Costume of the Ancient Eurasia (the Iranian-Speaking Peoples) (Kostyum drevney Evrazii [iranoyazychniye narody]).Moscow.

Yatsenko S.A., 2006b. Late Sogdian Costume (5th –8th cc. AD) - Ēran ud Anērān. Studies Presented to Boris I. Marshak in Occasion of His 70th Birthday (Ed. by M. Compareti, P. Raffetta, G. Scarcia). Venice. (Internet-version, 2003)-

Yatsenko S.A., 2007. Ancient Turks: Costume in Color Images (Drevnie Tyurki: kostyum na raznotsvetnykh izobrazheniyakh)Museum, November 2007 Internet-Conference “Antiquities in Museums Collections” (Internet-konferentsiya “Drevnosti v muzeinykh kollektsiyakh”) (ed. by T.N. Krupa), Kharkov.

Zelenin D.K., 1991. Eastern Slavic Ethnology (Vostochnoslavyanskaya etnografiya). Moscow.


Fig. 1. Stone statue from El’te-Kezhig site, Tuva Republic, the back side (a cap and belt pendants). Kyzyl, lapidarium of Republic Museum of 60 heroes (photo of S.A. Yatsenko, 2008).

Fig. 2. Early Turkic statue “Gengis-khan” – object of modern rituals (Russia, Tuva Republic, Burun-Hemchik county, near Bizhikig-Khaya village) (photo of M.E. Kilunovskaya, 2007).

Fig. 3. Some variants of reconstruction of the scene with Turks, the Western Wall of the “Hall of Ambassadors” in Samarkand / Afrasiab. The middle of the 7th c.: 1 – Étienne de la Vaissière version (Vaissière 2006, fig. 3); 2 – Markus Mode version (Mode 2006, ill. 1); 3 – reconstruction of the lower level (Barbet 2006, fig. 12). Drawings of François Ory.

Fig. 4. Granite mortuary bed of An Qie (579 AD) from X’ian, the Northern Zhou Dynasty (Anjia [An Qie] Tomb 2003, p. 140).

Fig. 5. An Qie meets a noble Turk during hunting (the upper level of the panel) (Anjia [An Qie] Tomb 2003, p. 181).

Fig. 6. An Qie plays dice (?) with a noble Turk (the lower level of the panel) (Anjia [An Qie] Tomb 2003, p. 182).

Fig. 7. An Qie meets a noble Turk in the yurt (Anjia [An Qie] Tomb 2003, p. 184).

Fig. 8. An Qie treats a noble Turk to wine (the upper level of the panel) (Anjia [An Qie] Tomb 2003, p. 188).

Fig. 9. Reception of a noble Turk in An Qie’s residence (the upper level of the panel) (Anjia [An Qie] Tomb 2003, p. 168)

Fig. 10. Reception of the noble Turk in An Qie’s residence. Detail (the upper row of figures) (Anjia [An Qie] Tomb 2003, p. 170).

Fig. 11. Panel of the marble mortuary bed from the Miho Museum (Shigaraki, Japan) with horse riders images (Miho Museum, 1997).

Fig. 12. Panel of the mortuary bed from the Miho Museum with the Turk hunting scene (Miho Museum, 1997, p. 250, fig.C).

Fig. 13. Sarcophagus of Yu Hong and his wife (592) from Taiyuan, the Sui Dynasty (Northern Zhou tomb 2001, No 1, p.34, fig.14).

Fig. 14. Sarcophagus of Yu Hong. Photo of eastern panel 4 from the northern wall (courtesy J.A. Lerner, 2007).

Fig. 15. Eastern panel 4 from the northern wall of Yu Hong sarcophagus (592). with the Turk hunting scene (Northern Zhou tomb 2001, No.1, p.36, fig.19).

Fig. 16. Mourning Turk. Detail of the wall painting “Peoples of the World Mourning Buddha”. Kyzyl, Mingoi, Maya cave (Grünwedel, 1912, S. 180, Fig. 415).

Fig. 17. Clay figurine from E. Schloss collection (Schloss E., 1969, No.42).

Fig. 18. Analogies to the headdress from terracotta 8-9 in the images of the 7th-8th cc. from Eastern Mongolia: 1 – statue of Köl Tegin Kaghan (Nowgorodowa 1980, fig.194); 2 – relief on the stone fence (“sarcophagus”) from Hul-Askhete I (Voitov 1996, fig. 67).

Fig. 19. Male upper body garment of green color in Kutcha oasis wall painting of the 7th c.: 1 – Kyzyl, cave 192, donator; 2 – Kyzyl, new cave 1, donator; 3 – Subashi, reliquary (reconstruction) (Huo Xuchu, Qi Xiaoshan 2006, pp. 94, 97, 171).

Figs. 20, 23. Clay figurines from M. Sanders’ collection (courtesy M. Sanders, 2008).

Figs. 21-22. Clay figurines from private collections (by Sotheby’s catalogues) (courtesy M. Sanders, 2008).

Fig. 24. Clay figurine from the tomb in Turfan oasis (Study on the Costume 1995, p. 200, fig. 371).

Fig. 25. Costume of Turks on wall paintings in the “Hall of Ambassadors” in Afrasiab (Samarkand), the middle of the 7th c. (Yatsenko 2004, fig. 1).

Fig. 26. Turkic belts of the 6th-8th cc.: 1-2 – statues with belts without metal plates and with “bending knee” - type daggers (1 – Kazakhstan; 2 – Xinjiang) (Kubarev 2008, fig. 1); 3 – plate decorated and quiver belts, the drawing of D.V. Pozdnyakov (Kubarev 2005, fig. 29).

Fig. 27. Headdress of Xianbei people’ origin from the Northern China. Silk, the 5th-6th cc. AD. Hangzhou, China National Silk Museum. No. 2592 (Road of Silk 2007, p. 82, No. 15).


1 In recent scripts of Turkic eposes we see evident traces of latest costume preferences (a great number of buttons on the fur-coat and trousers of the hero can serve an example) and accentuating of hypertrophical richness of characters (fur-coats completely embroidered with gold or made of ermine etc.) (see, for example, Maadai-Kara 1973, p. 256-257, 302, 306, 311). Darker shades of the main color range in men’s clothing in the late period should be taken into consideration. Dark shades (instead of red) in clothes of Turfan Uigurs judging from written and pictorial sources started to prevail from the 12th c. (Yatsenko 2000, p. 369). As to ‘ethnologically’ Turkic peoples of the Sayan-Altai Mountains, their origin was rather complicated and not quite clear including a number of originally Turkic and “Turkisized” groups. Between the 13th and 19th cc. their “costume ideals’ underwent evident Mongolization and came under Lamaism influence at a later stage; so, it is not always simple to identify earlier perception of clothing. In the later period for Turks of the Sayan-Altai Mountains differentiation of the sexes was served by a color symbolism for clothing, and the red was associated to a healthy posterity (associations being obviously different at the beginning of the Middle Ages). An earring was only worn by some boys who often got ill but not by grown up men, etc. On the whole, earrings, rings and beads in plaits were undoubtedly associated to women’s subculture, which cannot be said about the Early Middle Ages. In traditional perception of clothing of the both sexes a special attention was paid to the edge of the hem and caftan flaps (see, for example, L’vova and others 1988, p.172-173), i.e., the details which are seldom accentuated in the most ancient Turkic carved stone images.

2 . Those sculptures were erected by early medieval Turkic peoples before the spread of foreign religious systems on their territories which strictly discouraged idol worship. In later burial steles (not anthropomorphic already) some ethnic groups included thoroughly detailed elements of women’s costume (including embroidery on headdresses and the tiniest pendants worn on plaits), all plaques and ornament of man’s plate decorated belt suspended accessories, design and ornament of stone hats and helmets made separately to be places on sarcophaguses, etc., such as, in particular, some Western Kazakh steles of Young Horde/Zhuz of the 18th – early 20th cc. (Tasmagambetov 2002, photo p. 288, 308, 334).

3 . If we suppose that elements of belts in detailed depictions were not colored, then it is not clear why this main symbol of man-warrior’s status for early Turkic peoples was practically non presented in carving in most statues of men (not everybody was honored to be embodied in a statue). For example, the most usual element of the belt for carved stone images belonging to the period which interests us and which is found on the vast territory of Kazakhstan is just… a scheme of its design consisting of two main parallel straps without any detailing (Ermolenko 2004, p. 25).

In ritual complexes of Turkic burials in the Altai there are also plate decorated plaques made of some pressed organic material and, probably, covered with gold foil (Tashanta III, barrow 10) (Kubarev 2005, p. 51, fig. 14, 15). Earlier, in the Scythian time nomads of this region put into graves belts with wooden plaques as well as mirrors, daggers, horse harness, torques and other items made of wood. However, the fact does not witness their wide use in everyday reality (compare: Kubarev 2008, p. 70). It does not also prove the statement that the ‘Pazyrykians’ of the Scythian time were fond of looking into wooden mirrors or preferred wood to metal for items which were to undergo great physical pressure. This is an argument to support the idea about an opinion widely spread in past that some dead in some cases were “afraid” of metal (Yatsenko 2006, p. 353-354). So, not so long ago in burial ceremonies Russians substituted metal details of the costume with wooden ones exactly for the same purpose (Zelenin 1991, p. 374). I observed belts of Turkic peoples with wooden details (for example, a Kazakh one in the exposition of Chemkent Museum of Regional Studies and History), but they are not typical for real costume and their application is not documented.

4 . See also the J. Mahler collection (Mahler 1959).

5 . In Late Sasanid Iran the situation was a little different. Thus, Abdul Kasim Firdowsi while writing ‘Shah-nama’ also used, as it is well known, official Persian chronicles (for sure, interpreting them for the sake of artistic epical description). One of the plots is – a wedding embassy to a Turkic Kaghan (Persian – Kaghan-i Chin) of Khusrow I Anushirvan shahunshah (531-579). There are some interesting details of the Kaghan’s daughters costume described there (Esin 1979, p. 452).

6 . For example, judging from the materials of carved stone images from Kazakhstan it is clear that craftsmen and customers tended to accentuate of all male accessories first of all the belt and then the earrings (Yermolenko 2004, p. 25, 27-28). Due to diplomatic gifts we know about the prestigiousness of these or those costume accessories (it is clear , more luxurious than usually) presented to Turkic neighbors (the gift of Kaghan Tung Chehu to the Chinese Emperor in 627) and by Byzantines to Turks (the gifts of Emperor Heracleios to the attendants of the western Kaghan in 622). They were, as the reader has already guessed, the same belt (a gold one) and the earrings (with pearls) (Yatsenko 2006a, p. 284, 300). What can greater or smaller popularity of this or that artistic micro element of the costume in carved stone images of a certain territory tell us about? About the degree of craftsmen’s skillfulness, the material possibilities or personal taste of the customer, about the social status of the latter one, about the local tribal specificity? Alas! It is extremely difficult to clear it up today.

7 . The special status of the belt for early Turkic peoples was, in my opinion, connected partly with a number of traditional views on the anatomical part of body (tights and waist) where it was worn. For men it was associated not only with sexuality but with heroic power and health (see the description of the batyr’s (epical hero) strong waist in the eposes of Turkic peoples and traditional wishes to men during certain rituals for South Siberian Turks like that; “Let your waist be strong!”). the role of the plate decorated belt in early Turkic statues is sometimes accentuated also by the gesture of a late aristocrat symbolically holding his belt with his left hand (Evtyukhova 1952, fig. 3-5; 45; Kubarev 1984, table II, No. 11; 1997, p. 100), and in the statue from Kukanda point in the Altai Mountains the left hand of the personage is tucked behind the belt (Evtyukhova 1952, No. 9; Kubarev1984, pl. VII, No. 48).

In early Turkic memorial inscriptions there are only short notices about the number and the material of plaques for the plate decorative belt, and – without any explanation (Dobzhansky 1990, p. 78). Usually, there were about 25 of plaques on the main belt (judging from authentic finds of the 7th-9th cc. in the Altai Mountains) (Kubarev 2005, p. 55), so mentioning the belt with 40-60 plaques means that it was wrapped round the waist two times or including in this number plaques of another form from pendant straps. Even after the disintegration of the First Kaghanate, in burial sites of the Altai Region of the 7th-9th cc., the belt was found not in every warrior’s grave but in 60% of adult men burials (compare.: ibid, p. 56). A special belt buckle of such belt plays a big role (Judging from the epitaphs there existed an expression “to deserve (a certain type of) the buckle, i.e. a higher post). Occasionally, the role of the buckle is accentuated with iconographic means. For example, in the carved stone image from Qiaoxia (the Chinese Altai) the plate decorated belt of the man is depicted in detail and is unbuckled (but it is fixed on his waist) at the same time he is holding a large buckle (separated from the belt!) in his left hand (Wang Linshan, Wang Bo 1996, p. 40, No. 78).

8 . The same items – a diadem (gold) and earrings (made of pearl) are presented to a western Turkic ruler (following his councilors’ advice) by Byzantine Emperor Heracleios in 622 during their personal meeting (Chichurov 1980, p.151, 159).

9 . At the same time the gilded plaques of the horse’s breast strap were depicted by the craftsman more thoroughly.

10 . The new electronic edition of my book on the costume of Iranian-speaking peoples in the Russian language, enlarged and supplied with colored photos is being made gradually during 2009 by the company X-lab on the ‘Folk Costume’ site:

11 . B.I. Marshak is inclined to consider the asymmetry of some details in banqueting scenes and scenes of leisure-time (loosening of the right or the left sleeves in turns, wearing of the sword either at the left or at the right side and so on) to be a mistake of the Chinese carver or a deliberate use of mirror compositions (Marshak 2004, p. 17, 20). From my point of view, it was not accidental for both Chinese carver and his Sogdian customers controlling him. Placing of the sword at the right side of the relaxing main personage (see also figure No. 36 the ‘Hall of Ambassadors’ in Samarkand), the same as loosening the left and the right sleeves in turns can be explained first of all by the situational comfort.

12 . In connection with the status of the white belt (compare, for example, the Altai peoples: during the ceremony “aiylchiga kur kurchaary” men- guests, to differ from women –guests, got monochrome (not motley) belts from the host and the most important of them is white (L’vova and others 1988, p. 183).

13 . Two rhombuses (clavis) of another cloth depicted on the caftan from this Kutcha wall painting are also presented in the image of the hunter in the petroglyphs from the Chaganka River in the Alatai Mountain Region (Kubarev 2005, p. 41).

14 . It should be mentioned that very popular in this terracotta caftans with two lapels (and also with one right lapel) originate from Western Xinjiang (Khotan) of the 2nd-4th cc. AD. They became popular in some regions and, later (in the 7th c.) were spread by Turks over many parts of Central Asia (Yatsenko 2006a, p. 204, 206, 269).

15 . Clothes with right overlapping, evidently started its intrusion to “barbarians” of Central Asia beginning from the 3rd-4th cc. AD, after the north part of china being conquered by nomads. There are several analogous personages known in the art of Kutcha, Khotan and Toumshouq in Xinjiang in the 5th-6th cc. (Yatsenko 2000, p. 360-361; see also Yatsenko 2006, p. 247). We see the earliest traces of such influence in burial of nobility in Turfan – one of eastern (close to China) oases of Xinjiang. Authentic garments with analogous overlapping found there are dated back to approximately the 4th-3rd cc. BC (the woman from M 25 grave in Subeshi: Lu Enguo 1999, p. 104).

16 . To define the circumstances of the initial phase in this process it is important to consider the edict of the Great Kaghan of the Turks in 590. according to ‘Tangshu’, after finishing a long lasting war with China and conquering frontier Turfan oasis the Great Kaghan decided to force the local Chinese nobility (the majority of the population in this oasis were already Chinese) wear traditional Turkic clothes – caftans with the left overlapping and the hairstyle of many plaits, referring to the fact that the ruler of the territory had married a Kaghan’s daughter (Chavannes 1903, p. 103; see also Yatsenko 2000, с. 361, 363, 367). It was the reaction to fast Sinicization of Turfan Region which was according to “Suishu” was the fatherland of early Turks. Later, the reforms of the western Kaghan Ton-Jabgu (618-630) including local nobility into the system of Kaghanate rule and giving it Turkic titles (Yatsenko 2006, p. 282) evidently played a special role in surface Turkization of the Sogdian costume.

17 . Being applied on the figurines in a negligent way the yellow paint of the caftan lining flowing down slightly colored the adjusting trouser leg edges and shoes.

18 . These details do not make me stop doubting in the authenticity of this terracotta.

19 . Later on in carved stone images from Mongolia we come across caftans overlapping to the right which was typical for neighboring China. Such are some of the statues from the memorial complex of Köl Tigin (died in731), and carved stone images from Ol’tu point and Middle Narin (Yevtyukhova 1952, p. 48; fig. 47, 2-3). In the latter statue we see one left lapel, which has an analogy only in the depiction of a noble Uighur in the Buddhist temple 9 in Bezeklik (Turfan oasis) (Yatsenko 2000, fig. 67, 1).

20 . The ground work of the caftan in one of the graves in Tuva was blue (Kubarev 2005, p. 40). The blue color of the caftan is documented for main Turkic personage No. 36 on the western wall of the ‘Hall of Ambassadors’ in ancient Samarkand (probably the Kaghan’s deputy in the capital of Sogd) in the scene of official reception. The color was, probably, associated with the head of the Pantheon, the God of Heaven - Tengri (Yatsenko 2007).

21 . On some Turks in Samarkand we see caftans of yellow color (golden silk? – see about its wide use at the turn of the 7th-8th cc. among quite common nomads of the Altai Region: Kubarev 2005, p. 27-30), presented in Chinese clay figurine No 3 only in the lining; in its turn, in Chinese depictions there is much more of white. However it is rather connected not with real color preferences but depended on the restricted choice of paints in Early Tang terracotta, and evidently а small number of depicted personages in each series.

22 . At the same time, just on the contrary, brown silk became popular at the periphery of the Kaghanate (in graves of not very rich people in the Altai Mountains) in the 7th-8th cc., the same as yellow-greenish and violet (purple!) which are absent in our collection (Kubarev 2005, p. 30).

23 . It is not by chance that Kang Peja, the ruler of Turfan Oasis (submitted to Turks at that time) and Chinese (Han) by nationality, addressing the emperor from the Sui Dynasty in 590 underlined that cutting plaits of local nobility would be “a great change” in life and returning to civilization (see: Chavannes 1903, p. 103).

24 . In the 7th-8th cc. in the Altai Region (burial site Ak-Koby) a silk ribbon 2cm high with a buckle was used for fastening; two ribbons 45cm long were sewn on at the stomach (Kubarev 2005, p. 28, 33). Sometimes we see another manner of fastening (with one large button and rather low – also at the stomach) for Turks in Afrasiab) (fig. 25, 14); an iron button is known in grave 23 of necropolis Kokel’ in Tuva. A deep overlapping is marked at that period in burials of the Altai Region, and in wall paintings in Samarkand (in barrow 20 in Barburgazy I – 24cm: Ibid, p. 33).

Some Rights Reserved © The Author(s) -- Los artículos son propiedad de sus autores. (Ley 11.723 de la Republica Argentina)

Valid HTML 4.01!
Any Browser!
Como lo hacemos?