David Stronach

University of California, Berkeley


or all the considerable interest that has been taken over the years in the nature and uses of the yurt — in, for example, its wide distribution (which stretches from Mongolia to Anatolia), in its prefabricated, eminently portable elements, and in the variety of different terms that are used to define its component parts — very little has been done to try to uncover the more remote history of this long-lived, highly adaptable type of dwelling. Thus, while Peter Andrews’ Nomad Tent Types in the Middle East, mentions all significant references to yurts in the Middle East that occur in documents of early Islamic date or in travellers’ accounts, these and other sources cited in this magisterial work1 are not enough to carry the story of the yurt back to any moment before 700 CE.2

With reference to older attestations of the form, it may well be appropriate for future investigators to continue to interrogate Chinese literary sources. In addition, others may wish to explore the possible relevance of Inner Asia’s far-flung, variously dated petroglyphs. At the moment, however, I am chiefly concerned to draw attention to the testimony of once buried evidence which, to the best of my knowledge, has never been consulted in any detail in the present context — and which now appears to extend the chronological horizon of the yurt back to at least 600 BCE.

But before this and other matters engage our attention, a note on nomenclature is in order. The term yurt appears to be something of a misnomer. Of the main nomadic groups that live in yurts not one of them uses the term to describe this kind of portable structure. Instead, in Turkic languages, the term can mean “territory” or “camp site” but never “tent” (Andrews 1997: 5). Indeed, the real Turkic name for a tent can be ev, öy or üy, each of which simply means “dwelling.” Yet what appears to have begun as an error in Russian usage currently equates with a broadly accepted word in English.3 Accordingly, some use of the term may not be out of order; and all the more so in the present context since most ancient representations of yurts depict covered dwellings. In other words, while modern researchers in the field presumably have every opportunity to distinguish between “ribbed tents” and “trellis tents” (to use two of the terms that Peter Andrews derives from the structural elements of two somewhat differently constructed types of yurt), the archaeologist only rarely has this luxury. As a rule there is nothing to go on except for an artist’s impression of a yurt’s covered profile — and this in itself may not be a reliable guide to the nature of the frame that was employed.

Characteristics of a yurt

Fig. 1. A yurt being dismantled at Achikh Tash,
Southern Kyrgyzstan.

A short description of the key characteristics of the above-mentioned ribbed and trellis tents should perhaps also preface the archaeological notes which follow. To begin with, the characteristic wooden frame of a domed “ribbed tent” consists of long struts that bear directly on the ground at one end and which unite radially in a roof wheel at the top. In addition, the lower end of each strut is customarily secured by a peg driven into the ground (Andrews 1997: 179 ff.). For the more evolved “trellis tent” — so named for its most characteristic feature — I will very largely borrow, in an abbreviated form, from Peter Andrew’s description of the Türkmen tent of Khorasan (1973: 94 ff.) as well as from his descriptions of other trellis tents that are found in the general region of northern Iran and Afghanistan (1997: 25 ff). The tent consists of four principal elements (Fig. 1):

  1. The wall frame, made up of several flexible lengths of trellis (each with an open-work pattern of crossing wooden laths) which, when they are held in place by the restraint of a number of encircling bindings, create a cylinder up to about five and a half meters in diameter and about one and a half metres in height.
  2. The door frame which is introduced on one side of the trellis wall.
  3. The roof wheel, which is about two meters in diameter, which is pierced radially with slots to receive the roof struts and which always possesses an arrangement of spokes to support the wheel’s separate felt cover.
  4. A set of curved struts, each about two and a half meters in length, which span the space between the top of the trellis wall and the rim of the above-mentioned roof wheel (which is customarily suspended some three meters above the level of the floor).


With reference to the presence of several external woven restraints, the upper part of the trellis wall (Fig. 2) is “encircled by several broad girths, woven from wool ...while the

Fig. 2. Interior of a yurt, Achikh Tash.

struts are held firmly at the correct spacing by a much narrower girth which is wrapped around each in turn” (Andrews 1997: 95).4 The dome and the upper part of the walls are covered by two large felts, cut so as to leave most of the roof wheel exposed. The latter opening is then covered by a smaller felt, the forward part of which is usually folded back in order to leave a smoke hole which takes up the front third of the roof wheel. At the last, the walls are hung with four rectangular felts that nearly reach the ground and the open doorway is provided with either twin wooden door leaves or a felt flap.5


While the long struts of an important ribbed tent could no doubt be laboriously shaped in such a way as to provide partly vertical side walls beneath a domed top, the great advantage of the presence of a trellis appears to have been that it ensured, with a minimum expenditure of effort, the initial verticality of the side walls (not to mention a suitable, vertical unit to which a doorframe could be attached). Furthermore, the roof struts in this superior design could be relatively short and it was often only necessary to go the trouble of bending them at one point near their lower end.6


As far as the internal appointments of a traditional yurt are concerned, those dispositions that are still in evidence in many parts of northeastern Iran and Afghanistan may serve as as a broad guide to the way interior domestic space is often organized. Wherever other factors are equal, the doorway faces south. The men’s side is then to the west and the women’s to the east. The hearth stands at the center of the tent, but a little forward of the exact center in order to lie directly under the smoke hole. In addition, the interior of the tent is often conceived of as having four distinct quarters with the hearth at the center. The place of honor (or the reception area) is located towards the rear. This is where (at least in Iran) a brocaded rug can cover the standard floor felts and where the adjoining north wall may display say, two wall-bags of superior quality (Andrews 1997: 77).


Finally, the range of adjustments that can be made to the coverings of a yurt in order to accommodate changes in climate and temperature are many and various. In hot weather, for example, the wall felts can be raised by as much as 50 cm so that “air can enter the tent through the top and flow out through the gap at the periphery” (Andrews 1997: 73). At such times too the cane screens that are often attached to the outside face of the trellis wall (see note 4, above) serve to “filter the glare of the sun on the dry ground outside” apart from offering protection from wind-blown dust and debris (Andrews 1997: 74). In addition, the smoke hole, which is closed at night in winter, spring and fall, is left open all summer long. Further, since the smoke hole normally faces south (with any protruding,

folded felt located on the north side), this opening is positioned in such a way as to admit the rays of the sun. As Andrews has remarked, this arrangement provides a “patch of light on the wall or the furnishings, which moves predictably around the periphery according to the time of day.” In other words the interior of the tent becomes “a sun dial and the position is used to tell the time for prayers or meals” (Andrews 1997: 74).


In cold weather, the main external felts are duly lowered until their bottom edges touch the ground. Furthermore, the hearth is lit; and cooking — an outdoor undertaking in the warmer months — becomes an indoor activity. In sum, this portable type of dwelling seems to have more than deserved its longevity. It was regulable for extreme changes in climate; it could boast rich hangings to indicate elite status both inside and out; its standardised, prefabricated parts made it swiftly repairable; and, as the dwelling of choice for pastoral nomads occupying a broad belt of territory approaching a quarter of the span of the globe’s surface, it could be speedily assembled or disassembled for conveyance on camels, horses or donkeys or even, at times, on open carts (Gervers and Schlepp 1997: 101).


The depiction of a yurt on an engraved bronze bowl of c. 600 BCE

This paper owes its initial inspiration to the recovery of a totally unexpected image from a surprisingly early archaeological context. In brief, the year 1982 saw the chance discovery of the “Arjan tomb,” a rich burial of Neo-Elamite date that came to light not far from the ruins of Arjan, a Sasanian and medieval township deep in the Zagros mountains of southern Iran at a point 10 km north of Behbahan and 250 km southeast of Susa (Tohidi and Khalilian 1982). In the course of recording the tomb the excavators recovered a number of precious and non-precious metal objects, at least four of which are now known to have carried the same unvarying legend, “Kidin-Hutran, son of Kurlush.” And while Kidin-Hutran appears to represent a hitherto unknown local Elamite ruler, the Elamite cuneiform script that was used to write this short text can be reliably ascribed to an interval between the mid-seventh and the mid-sixth centuries BCE (cf. Vallat 1984: 4). In line with this finding, moreover, a series of independent clues provided by the iconography and style of the main objects suggests a parallel date which most scholars would now place either late in the 7th century or at some point early in the 6th century.7


Fig. 3. The engraved design on the
interior of the Arjan bowl.

When the first detailed description of the tomb and its contents appeared in English in 1985 the bronze bowl had still not been treated by the conservators at the National Museum in Tehran. Accordingly, it was merely described as a “large shallow bowl, 43.5 cm in diameter and 8.5 cm deep” and was listed as one of thirteen bronze vessels recovered from the floor of the tomb (Alizadeh 1985: 55). The subsequent treatment of the object (Vatandust 1988) revealed the existence of one of more stunning artifacts from the tomb: namely, a vessel with Kidin-Hutran’s inscription etched on the exterior (just below the rim) and with five concentric registers of engraved decoration distributed across the surface of the shallow interior (Fig. 3).


Such a scheme of decoration — with its notably relaxed and lively character — can be broadly related to the bronze (and sometimes even gold) “Phoenician bowls” of the Mediterranean and the Near East, which remained in production as late as the second half of the 7th century BCE (Markoe 1986). At the same time, however, the Arjan bowl cannot be taken to be the product of a distant workshop. As Yousef Majidzadeh was the first to point out, a large number of specifically Elamite elements are visible in the bowl’s multiple incised images (Majidzadeh 1992: 136-138). And as I have sought to stress elsewhere (Stronach forthcoming), this circumstance implies that an engraver who was working for a local patron — presumably Kidin-Hutran himself — drew up the intricate designs that make this vessel such an extraordinary “window” on one limited region of southwestern Iran in the years shortly before Cyrus the Great (559-530 BCE) conquered the Medes and founded the first Persian empire.


As Figure 3 indicates, the broad outer register of the Arjan bowl includes a prominent representation of the basic wooden elements of a circular, domed “ribbed tent”. The tent is shown without its customary felt covering in an illustration that was clearly intended to reveal the structure’s characteristic, long curved struts and all-important roof wheel. Indeed, this last item is deliberately shown in an unreal, upright position, i.e. in an “aspective view” in order to stress its vital role.


The doorway in the incised design is also of special interest; for, while modern yurts are often equipped (as has just been noted) with double wooden doors that are side-hinged, the Arjan tent appears to document the presence of a single, broad wooden door that was top-hinged.8 Very conceivably this latter design had the same advantage in an emergency as a felt door flap: it could be closed in a split second. As an enlarged and slightly modified view indicates, the door was customarily propped open by a tall pole with a lion- or wolf-headed finial — a distinction no doubt reserved for a residence of high status (Fig. 4).9


The presence of two intriguing objects of identical appearance inside the yurt (unusually shaped water jars or, as it is tempting to suppose, twin incense burners), not to mention the presence of a number of attendants either inside the tent or in the vicinity of the shaded doorway, may have been intended to demonstrate that this portable dwelling not only served as a kind of portable hunting lodge (in keeping with the bowl’s adjacent references to hunting and banqueting) but also as a setting for formal audiences when Kidin-Hutran was “on tour” in the back-country of his mountainous kingdom. Indeed, it is more than likely that the internal details

Fig. 4. A detail of the yurt or “ribbed tent” in the
outer register of the Arjan bowl. The lower parts of
a number of struts have been deleted in order to
provide a clear view of the internal appointments.
At right, in a location that also placed him at the
focal point of an adjacent banquet scene,
Kidin-Hutran sips wine from a deep vessel with a
flaring rim while seated on a high-backed throne
with a single visible cervine (gazelle-headed?) finial.

illustrate the fittings of the tent as these would have appeared when viewed from the open doorway. Thus, on entering the royal tent, a visitor would have taken in the elaborate wall-hanging (or floor carpet?) associated with the place of honor as well as the flanking positions occupied by the two probable incense burners.10


Needless to say, a number of intriguing questions are necessarily posed by the inclusion of Kidin-Hutran’s yurt in what appears to have been, at least to some extent, a record of this local ruler’s characteristic activities. It has to be acknowledged, for example, that the yurt may once have been a common tented form in the highlands of southern Iran, in which case it could have been brought there at the time that the Persians first entered the region somewhere near the beginning of the first millennium BCE. Alternatively, if such an explanation should fail to find adequate confirmation in the fullness of time, it would at least seem difficult to deny that portable dwellings of this type must have been present in the steppes of Central Asia from a date prior to 600 BCE; and, in this event, the long-established conventions of gift-exchange between rulers both great and small could always have chanced to bring this exotic indication of status all the way to Kidin-Hutran’s southern domain.


At all events it is now decidedly difficult to suppose that the yurt was an exclusively Turkish invention, and that tents of this kind made their first appearance in the vicinity of Iran, as Andrews once suggested (1973: 94), “as the homes of Türkmen nomads all descended from the Oghuz tribes” after these tribes had crossed the Amu Darya (the Oxus) in the eleventh century. On the other hand Peter Andrews’ inclination to view the ribbed tent as “an ancient type, as old as, if not older than the trellis tent” (1997: 179) can now be shown to have been entirely correct. Kidin-Hutran’s tent was presumably representative of one of the more superior designs that was available at the time of his reign and, as such, the incised design in the Arjan bowl strongly suggests that the introduction of the otherwise dominant trellis tent had still not occurred.


A yurt in a wall painting of the lst century CE

Fig. 5. A wall painting found in a tomb of the
1st century CE at Panticapaeum, near Kerch.

One further hint that nomadic peoples of Iranian origin used yurt-like structures in the course of their migrations across the endless grasslands of Asia comes from the extreme western limit of this investigation. I refer to the presence of what may well have been a felt-covered framed tent (Fig. 5) in a no longer extant wall painting found in a Sarmatian tomb of the first century CE. The tomb came to light in the city of Panticapaeum (in the vicinity of modern Kerch, in the Crimea) and the painting itself has been in the published domain for more than eighty years (Rostovtzeff 1922: 160 ff. and pl. 28,1).

In his description of the painting Rostovtzeff observed that “the scene is an idyllic one. The dead man, armed, followed by a retainer, is riding towards his family residence, a tent of true nomadic type (my emphasis). His household, wife, children, and servants, are assembled in the tent and beside it, under the shade of a single tree; beside the tree is his long spear, and his quiver hangs from a branch.” He goes on, “The interpretation is easy: the gentleman is a landed proprietor, who spends most of his time in town: in summer, during the harvest season, he goes out to the steppes, armed, and accompanied by armed servants; taking his family with him. He supervises the work in the fields, and defends his labourers and harvesters from the attacks of neighbours who live beyond the fortified lines; Taurians from the mountains, ferocious footsoldiers; Scythians from the plains, horsemen and landowners. Who knows? perhaps he raids a little himself.”11

Whatever credence one may wish to place in Rostovstzeff’s vivid interpretation, the chief point in the present context is that, in spite of its unusual, square-shouldered appearance and strangely prominent ventilation hole, the felt-covered structure in Figure 5 is, in all probability, the second earliest known depiction of a yurt. Indeed the prominent “shoulders” that appear in the painting might represent an uncertain attempt to stress the presence of an inward-leaning trellis wall. At the very least this carefully delineated structure appears to represent a tented dwelling of some quality.

If close attention is paid to the exaggerated scale of the chair and its occupant (the supposed “wife” of Rostovzteff’s narrative), there is a good chance that an enthroned goddess is represented: one attended, in fact, by a number of individuals, each of whom is depicted (following time-honored norms of differential, hierarchical scaling) at a decidedly smaller scale. And although the goddess herself is shown in a frontal as opposed to a side view — in what was already a much-used artistic convention by the first century CE — her proximity to the yurt calls for special notice. That is to say that her position may well have been intended to underline the special relationship that existed between the goddess and the deceased (whose body, in this reconstruction, can be understood — notwithstanding his parallel, active, equestrian representation — to lie, suitably mourned, inside the tent).

Fig. 6. A detail of the lower register in a large felt
carpet or wall-hanging of the 3rd century BCE
from barrow 5 at Pazyryk.

As far as the structure’s prominent roof opening is concerned, it is appropriate to stress that William of Rubruck’s thirteenth century description of Tartar tents included a reference to structures “from which projects a neck like a chimney” (Gervers and Schlepp 1997: 105). Furthermore, the likelihood that this elite Sarmatian tent also had something like a square base is again not unparalled in the long history of the yurt. As late as 1935 Owen Lattimore was able to photograph a yurt of a similar, more or less square design which was used in Inner Mongolia to celebrate “the Sacrifice of Chinggis Khan at Ejen Horo” (Gervers and Schlepp 1997: 114 and fig. 16).

Finally, with regard to the Panticapaeum painting, there would seem to be a distinct possibility, as not a few others have surmised, that the composition represents a retelling of a well-known legend that was already possibly alluded to in the celebrated felt carpet or wall-hanging (Fig. 6) from barrow 5 at Pazyryk (Rudenko 1970: 13ff). It is true that the repeated elements in this latter design of the 3rd century BCE (cf. Mallory et al. 2002: 210) are reduced to a single horseman, who wears his bow-case on his left side “as if prepared for war” (Rudenko 1970: 275); to the rider’s slim, long-tailed horse; to an elaborate “tree” with abundant blossoms; and to an enigmatic enthroned figure who, with a shaven head and no facial hair, is usually taken to be a goddess (cf. Stronach 2002: 389 and fig. 10). But at the very least these similarities oblige us to continue to weigh the character of the principal participants in the Panticapaeum wall painting — and, hence, the status of the depicted yurt.

Yurts in Sogdian funerary reliefs of the second half of the 6th century CE

The last body of once buried evidence that calls for close consideration comes from the eastern limit of this survey. It is chiefly owed to recent archaeological discoveries from north China, most of which have only begun to be described in print within the past ten years. As readers of Étienne de la Vaissière’s article in the previous issue of The Silk Road will recall, the period of the fifth and sixth centuries marked a peak in Sogdian emigration to China. It was a time when the Sogdians were deeply involved in the caravan trade between China and the West; and, at least by the latter part of the 6th century, “most of the main towns of northern China” had a resident Sogdian community in which each community was customarily headed by a Sabao (or chief caravaneer) who was also granted mandarinal rank in the official Chinese hierarchy (Vaissière 2003: 24).

By the second half of the 6th century numbers of Sogdian officials of this high status appear to have been in a position to order Chinese-style stone funerary beds for their relatively capacious tombs. The carved and painted vertical panels that were an integral part of such beds (cf. Marshak 2001: fig. 12) provided ample space, moreover, for the owner to record elements of his Sogdian way of life (including his continuing devotion to Zoroastrianism) as well as evidence of the extent to which he and his family were integrated into Chinese society.

From the standpoint of the present enquiry, however, such panels are of particular value because they allowed the owner to document something of the nature of his peripatetic ambassadorial duties at a time when relations between the Sogdians and the Western Turks were of great importance. This element is illustrated with striking realism in the case of the superbly preserved funerary couch of An Jia, a Sogdian Sabao who flourished under the Northern Zhou and who was buried at Xian in 579 A.D.12

Fig. 7. Drawing of an elite yurt
depicted in a funerary relief from
the tomb of the Sogdian Sabao,
An Jia (d. 579 CE).

Descended from a family that originated in Bukhara (Marshak 2001: 244), An Jia had extensive dealings with the Turkic khagans of his day. In one panel An Jia is depicted, for example, in intense negotiations with a khagan inside the doorway of the latter’s yurt (Fig. 7). In the illustration in question the appearance of the yurt is reduced to little more than an elegant frame for the animated discourse of the two principals; nonetheless the near-vertical sides of the tent strongly suggest that it could have benefitted from the presence of a trellis wall.13 Beyond this, the elite rank of the yurt is indicated by the fact that it had a covering of tiger skins.14 In addition, the inner side of the open doorway had a curtain of fine quality (perhaps suggesting a use of silk) and the floor of the yurt appears to have been at least partly covered by a long-fringed circular carpet.

At least one other elite yurt with a tiger-skin covering is illustrated in a panel that stood on the left side of An Jia’s bed (Marshak 2001: fig. 18) and still other panels record the lavish nature of the hunts and banquets that An Jia organized for the entertainment of his Turkic counterparts. Indeed, whatever diplomatic considerations may have occasioned these proceedings, An Jia’s record of his exceptional life leave us with a clear impression that he was notably taken by the exotic ways of the khagans.

Chinese testimony

If An Jia was fascinated by such matters, he was not alone. Quite apart from the fact that the Chinese taste for the exotic reached unprecedented heights during the heady days of the Tang dynasty (c. 618 - 917 CE), members of the highest ranks of Chinese society appear to have found unusual pleasure in exploring, especially in the winter but in certain cases even in the summer as well (see note 9, above), the attractions of an urban, tented existence. In the capital, Luoyang, where the leading literati of the 9th century frequently occupied grand villas with extensive grounds, the celebrated poet, Bai Juyi (772-846), not only set up a yurt in the front courtyard of his Luoyang villa, but he wrote a poem, in 833, in praise of the virtues of his tented abode. Through Bai Juyi’s personal vision, then, we learn — most engagingly — of the advantages of a yurt:15

The Sky-Blue Yurt

Bai Juyi

The finest felt from a flock of a thousand sheep, stretched over a frame shaped like the extended bows of a hundred soldiers.

Ribs of the healthiest willow, its color dyed to saturation with the freshest indigo.

Made in the north according to a Rong invention, it moved south following the migration of slaves.16

When the typhoon blows it does not shake, when a storm pours it gets even stronger.

After Marshak 2001: fig. 14a.

With a roof that is highest at the center, it is a four-sided circle without corners.

With its side door open wide, the air inside remains warm.

Though it comes from far beyond the passes, now it rests securely in the front courtyard.

Though it casts a lonely shadow during nights brilliantly illuminated by the moon, its value doubles in years when the winter is bitterly cold.

Softness and warmth envelop the felt hangings and rugs; the tinkling of jade enfolds the sounds of pipes and strings.

It is most convenient after the earth has been covered with frost, and it is the best match when snow fills the sky.

Positioned at an angle is the low chair for singing, evenly disposed are the small mats for dancing.

When I have leisure time I lift open the curtain and enter the yurt, and when I am drunk I wrap myself up in a cover and sleep there.

Behind me an iron lamp-stand that bears a candle; a silver incense censer that flames is suspended from the ceiling.

Kept deep within is the flame that lasts till dawn; stored inside is the fragrant smoke that lasts till evening.

When the animal-shaped charcoal is close by, fox furs can be cast aside.

When the ink-stone is warm it melts the frozen ink and when the pitcher is heated it becomes a stream in springtime.

An orchid canopy will barely attract a hermit and a thatched hut is inferior for meditating.

(But invited to my yurt) an impoverished monk responds with praise, and a threadbare scholar stays in place, unwilling to leave.

Guests are greeted with it, descendants will hand it down to posterity.

The Wang family boasts of their antiques, but they have nothing to equal this Sky-Blue Yurt.17

Concluding remarks

If we may work backwards from the latest evidence just cited, Bai Juyi’s testimony is important. It supplements, in many vital ways, the visual representations of elite yurts that occur in the new-found Sogdian reliefs. In particular, Bai Juyi’s poem indicates that the more significant yurts of the second half of the first millennium CE were of considerable size (as witness the places reserved — at least in cosmopolitan Luoyang — for such activities as singing and dancing); and that such satisfying, logically designed structures (such as were most at home in more northerly climes) were at once luxurious and far more impervious to the assaults of winter than a contemporary Chinese mansion.

Accordingly, context alone can be seen to explain the greatly abbreviated, almost coded depictions that appear in the Sogdian reliefs. Context is all; and it is clear that the tents in question were only meant to be read as “atmospheric settings” for the actions of the principal protagonists. At the same time these 6th century Sogdian carvings provide precious evidence of the extent to which dwellings of this kind were unquestionably in wide-spread use at this juncture among the Western Turks.18

As far as the more ancient history of the framed tent is concerned, both the excavated evidence from Kerch and that from Arjan, deep in southern Iran, can be said to underscore an already acknowledged Iranian perspective. On the one hand the Sarmatians were an Iranian-speaking people and on the other hand a number of the objects from the Neo-Elamite tomb at Arjan document the extent to which the Elamites were adjusting to the habits and tastes of their immediate Persian neighbors in the years before and after 600 BCE.19 Long before the wholesale adoption of the yurt by the Turks, in other words, there may have been an extended period during which peoples of Iranian origin made prior use of the form.

Interestingly enough, Andrews himself stresses that, while the framed tent has a known history of “1300 years,” it also has “several centuries” of unknown history before that (1997: 12). At one point, for example, he goes out of his way to stress “the need to master the technique of wood-bending” in order to create such a tent. Then, after pointing out that the techniques in question were definitely available in the time of Chinggis Khan (as evidenced by the remains of the trellis tent mentioned in footnote 18, above), he goes on to admit that the techniques also existed much earlier “as indicated by the cartwheels... found at Pazyryk” (Andrews 1997: 25).

The extent to which conceivably yurt-related innovations can be said to have been present at Pazyryk is of course one of high interest. In this context the structural and decorative similarities between the distinctive wood and leather shields from Pazyryk (Rudenko 1970: pl. 144) and the decorated cane or reed screens that regularly complement today’s trellis tents is decidedly striking. In addition, a box-like wooden cabin that was mounted on one of the Pazyryk carts is known to have been partly covered by black felt (see e.g. Rudenko 1970: fig. 17 and pl. 131). However, if any attempt should be made to speculate on the ethnicity of those who were buried at Pazyryk — or on the degree to which they might have enjoyed indirect communications with the heartland of the Achaemenid empire as early as the 5th century BCE — it is necessary to be aware of the fact that the ethnic identity of those who were buried in Pazyryk’s frozen barrows remains uncertain,20 and that the cumulative evidence from a series of revised radiocarbon determinations, dendrochronological indications, and art historical considerations now combines to to suggest that the date of the Pazyryk culture falls “in or near the third century BC” (Mallory et al. 2002: 210).

* * *

In conclusion, if I may take the no doubt rash step of providing a tentative timetable for the evolution of the yurt (which, for all we know, may still owe its first putative beginnings to the distant moment at which pastoral nomadism began to take hold in Inner Asia), I will limit myself to a few interim reflections. To begin with it is not difficult to concur with Peter Andrews’ contention that the simple “bender tent” of his overall classification21 could have been in existence by the second half of the second millennium BCE and that it could have been employed by, among others, “Iranic nomads” (1997: 5-6). The next advance was surely the creation of the ribbed tent — the first quintessential form of domed yurt — which conceivably evolved early in the first millennium BCE since it would appear to have been widely distributed by 600 BCE. As for the admirable trellis tent, which still remains in regular use over a very substantial area, this was almost certainly present, as we have seen, by 560 CE. Furthermore, if the tall, rectangular wickerwork shields of the Achaemenid Persians (Schmidt 1953: 225 and pl. 136; Briant 2002: 195), not to mention the similar, if shorter, shields of those who were interred in the barrows at Pazyryk, should be in any way related to the cane screens that were presumably a necessary complement to even the earliest trellis tents (see note 4, above) there could be a case, in my view, for suggesting that the earliest examples of this most evolved form of yurt were introduced at a date not far removed from the middle years of the first millennium BCE.

About the author

David Stronach was educated at Gordonstoun and St. John’s College, Cambridge. Between 1957 and 1959 he was a Fellow of the British Institute of Archaeology at Ankara and the British School of Archaeology in Iraq and was, successively, an assistant on the excavations of Seton Lloyd, Max Mallowan and Sir Mortimer Wheeler. In 1960 he was named British Academy Archaeological Attaché in Iran and in the following year he began a nineteen-year term as the Director of the British Institute of Persian Studies. Since 1981 he has been Professor of Near Eastern Archaeology at the University of California, Berkeley. He has conducted excavations at various sites in the Near East including Pasargadae, Tepe Nush-i Jan and Nineveh. His publications include papers on the early history of wine, textiles, and the Persian garden. In January of 2004 he was awarded the Gold Medal of the Archaeological Institute of America for Distinguished Archaeological Achievement.



  1. Such as the Wen Chi scroll of 12th century date (Andrews 1997: 12) or other evidence which suggests that elements of the current tent terminology of the Türkmen of Iran could be as much as 1200 years old (Andrews 1997: 215).
  2. In exploring a topic that has frequently taken me into areas of enquiry that lie outside my customary “bounds” I have not seldom profited from the promptings of others. In this context I particularly wish to acknowledge the extent to which this paper is indebted to the rigorous and devoted field studies of Peter and Mügül Andrews. More recently, I have received valued help from Elizabeth Baughan and Alma Kunanbayev, and, most especially, from Boris Marshak and Jeffrey Riegel. This said, I alone am responsible for the tenor of the remarks that follow.
  3. In which context the noted anthropologist, William Irons, refers without equivocation to the “yurts” of the more nomadic component of the present-day Yomut Türkmen of northeastern Iran. See, for example, Irons 1975: 26, 36.
  4. Flexible cane screens (composed of tightly connected, vertical lengths of sunflower stems or cane) are also very often fastened, in the manner of a tall sheath, to the external face of the lattice wall of a trellis tent (see e.g. Andrews 1997: 48). The sometimes striking decorative qualities of screens of this type (as they are still produced, using reeds rather than cane, by, for example, the Kirghiz) are now treated in detail in Sommer 1996.
  5. In the latter instance the felt flap can be backed by a mat composed of “canes laid horizontally and bound with vertical goat hair lines.” This arrangement allows the flap to be rolled up, with the felt face outwards, when the doorway is open; equally, in an emergency, the felt flap can be dropped in an instant (Andrews 1997: 67).
  6. Outside Iran, in Mongolia in particular, straight roof struts also regularly serve to bridge the space between the top of the trellis wall and the roof wheel. See, for example, Gervers and Schlepp 1997: fig. 11.
  7. See especially Boehmer 1989: 142-3; Curtis 1995: 22; and Stronach 2003: 252-5.
  8. This design may even throw useful light on a longstanding puzzle connected with the anatomy of the 13th century Mongol tent. The top-hinged Arjan door could account, for example, for the phrase “... let them lift for you the wide door” which is found (in evident reference to an elite tent) in paragraph 37 of The Secret History of the Mongols. For prior discussion and references, see Gervers and Schlepp 1997: 97.
  9. The presence of a lion- or wolf-headed finial finds an unexpected parallel in a much later context that derives from 7th century China. There the eccentric Tang prince, Li Cheng-Chien, who elected to live in a yurt on the grounds of his palace on a permanent basis, is said to have enjoyed sitting in front of his tent under a “wolf’s head ensign” (Schafer 1963: 29).
  10. With reference to Kidin-Hutran’s use of his yurt as a mobile “hunting lodge,” compare the way in which a Mongol ruler of the first half of the 13th century is said to have moved his felt tent “to follow the hunt” in an activity in which he regularly took “his officers and retinue with him.” (For references, see Gervers and Schlepp 1997: 99.) For the known deployment of precious incense burners in tents of diverse kinds, see both Plutarch’s vivid description (Alexander 20.12-13) of Alexander the Great’s visit to the vast, captured tent of Darius III (a description discussed at greater length in Stronach 2004: 718, note 42) and the thirteenth rhymed couplet in the poem, “The Sky-Blue Yurt” by the eminent Tang poet, Bai Juyi, which appears on p. 14, above.
  11. Fig. 8. Detail of an elite yurt in
    a Sogdian funerary
    relief of c. 560 CE.

    It is frustrating that Rostovtzeff makes no mention of the way in which the creation of a large wall-niche apparently destroyed part of the original painting (Fig. 5) or to the fact that a second wall-niche appears to be closely associated with an inscription, in Greek, which refers to Anthesterios, son of Ktesippos (Fig. 5). Indeed, it is difficult to decide whether these omissions stem from Rostovtzeff’s innate awareness of the extent to which the Sarmatians chose to “percolate into the populations of the Greek cities” on the northern rim of the Black Sea, where they adopted “the Greek language and some Greek customs” (1922: 120) or whether his silence was meant to indicate that these very possibly secondary manifestations had no place in his analysis.
  12. Marshak 2001: 244-252. In this same article, which has been justly singled out as “the main reference for the Sogdian funerary reliefs found in northern China” (Grenet 2003: 35), the author initially illustrates and discusses a set of Sogdian reliefs now housed in the Miho Museum in Japan. Since one of these reliefs shows a long-haired Turkish ruler seated, at ease, in the doorway of his yurt (Marshak 2001: fig. 8a) and since Professor Marshak believes the Miho reliefs to be the earliest in the series — dated, that is, to the 560’s (Marshak, personal communication) — this specific yurt deserves to be counted, if only by a decade or two, as the oldest so far attested in these Sogdian documents (Fig. 8).
  13. I am indebted to Jasmine Shahbandi for the drawing in Figure 7. The very slightly impressionistic treatment of the scene is intentional.
  14. Marshak 2001: 249; the skins in question were presumably those of the Siberian Tiger. Given the normally robust internal structure of any framed tent, I also think it likely that the horizontal red band near the top of the tent and the vertical red “flaps” on either side of the open doorway were chiefly decorative embellishments (see especially the color plate in Marshak 2001: fig. 14a), even if a structural function cannot be ruled out entirely.
  15. Unreserved thanks are owed to my colleague Jeffrey Riegel, Professor of Chinese at the University of Californa, Berkeley, who prepared, with great generosity, and at short notice, the following translation of Bai Juyi’s poem. He comments that the poem, composed in twenty rhymed couplets, is probably of the Tang dynasty sub-genre “in praise of things.” The initial task of tracking down the poem, the importance of which was first drawn to my attention by Boris Marshak, was greatly facilitated by the unstinted help of Lynn Xu.
  16. Urmat Mamytov and his family at their yurt near
    Karakichi Pass, not far from
    Lake Sonkyol, Kyrgyzstan.

    The term “Rong” was used by the Chinese of the Tang period to refer to non-Chinese populations beyond their western borders (personal communication from Jeffrey Riegel). It is of interest that Bai Juyi refers to his yurt as one that was made “in the north” while also referring to it as a Rong, i.e. western, invention. But since Turkish power to the north and northwest of the Tang capital was effectively consolidated by the time that Bai Juyi had earned his prominence no serious contradiction exists.
  17. The poem may be located in its original form in the Bai Juyi ji jianjiao (Annotated and Collated Edition of Bai Juyi’s Collected Works), Vol. 4 (Shanghai: Guji chubanshe, 1988): 2134-6.
  18. Given the predictable stress on status in the records that are available to us, there is little hope that extensive evidence will ever be available where the tented structures of commoners are concerned. One exception is known, however; and it appears, importantly, to fortify contemporary evidence which suggests that the tents of any single tribal group, elite or otherwise, will normally be of the same type. The case in question concerns the wooden elements of an unmistakable trellis tent from the grave of a commoner who was buried in the Khentei Mountains of Mongolia in the time of Chinggis Khan. While this simple grave provides the earliest incontrovertible evidence for the existence of the trellis tent (Andrews 1997: 25), it could also be said to lend circumstantial support to the view, expressed above, that the yurts in the various Sogdian depictions were probably already of this improved design.
  19. The name of Kidin-Hutran’s father, Kurlush, even suggests that he himself was of Persian ancestry. See Vallat 1984: 4; Potts 1999: 303; and, most recently, Alvarez-Mon 2004: 232.
  20. Against a backdrop of dates obtained from Chinese or other historical sources, the population has been variously defined as originating from the Issedons, Wusun, Yüezhi or Saka (Mallory et al. 2002: 204).
  21. Such a tent is described, in brief, as having “supple wooden rods... stuck into the ground opposite one another, bent to meet as an arch, and fastened at the top” (Andrews 1997: 5).

From the Editor | Archaeological Explorations of Bronze Age Pastoral Societies in the Mountains of Eastern Eurasia | On the Antiquity of the Yurt: Evidence from Arjan and Elsewhere | The Burial Rite: an Expression of Sogdian Beliefs and Practices | Palmyra as a Caravan City | The “Ancient Tea and Horse Caravan Road,” the “Silk Road” of Southwest China | Klavdiia Antipina — a Tribute to the Ethnographer of the Kyrgyz | Mongolia: a different view | British Library Symposium on “The Kingdom of Khotan to AD 1000: A Meeting of Cultures” | Guidelines for Contributors