Home Page Discussion


Salon du Tapis d'Orient

The Salon du Tapis d'Orient is a moderated discussion group in the manner of the 19th century salon devoted to oriental rugs and textiles and all aspects of their appreciation. Please include your full name and e-mail address in your posting.


Editor's Note: This is an abridged version of an article that appears in GHEREH (No. 22, p. 35). We appreciate their permission to reproduce this copyrighted material.

Old Motifs: The Bactria-Margiana Archaeological Complex as a Possible Source

by Christoph Huber

The possibility that a whole corpus of designs found on oriental carpets has roots going back some thousands of years is thrilling and exciting. To find analogies in the motifs from different works of art is like seeing a hidden tradition underlying the religious and cultural backgrounds of peoples, sometimes separated by considerable time and space.

In his beautiful book Tribal Rugs, James Opie emphasizes the importance of the so-called Luristan-Bronzes from the first millennium BC in the development of so much carpet motif(1). I would like to point to an other culture that may be of interest as a possible origin or intermediary of carpet ornaments.

During the last twenty years a hitherto unknown culture in Central Asia has been uncovered by archaeologists. This civilization of the Bronze Age is called Bactria-Margiana Archaeological Complex (BMAC) and is situated, as the name indicates, in Bactria on both sides of the Amu Darya and in Margiana, the Merv-Oasis. It can roughly be dated to the first quarter of the second millennium BC and has been considered to be a fifth center of Old World civilization (with Mesopotamia, Egypt, India and China)(2). The BMAC had strong connections not only to the Neolithic and Chalcolithic sites of the piedmont region of Southwest Turkmenistan(3), but also with Mesopotamia and Elam, Syria, Anatolia, Eastern and Northern Iran, Baluchistan, the Indus-Valley and even China(4).

The glyptic of the BMAC is of special interest in relation to carpets. Amulets and seals made of soft stone and pierced lengthwise often have a swastika engraved on one side (5).

Figures 1-4

Another interesting motif from a BMAC seal which can be followed through the centuries is the endless knot (Fig. 1)(6). It reappears on mediaeval metalwork(7), and together with the interlaced motif of the seal from Fig. 2 on a 14th century Italian painting depicting an Anatolian carpet(8). Additionally, it can be seen on a very important carpet from eastern Anatolia, now in the Vakiflar Museum in Istanbul(9), as well as on many 19th century Lori- and Bakhtiari weavings(10).

Another seal includes the kind of endless knot ornament(11) that is part of the Chinese symbol of happiness(12), connects the medallions on a 12th century bronze kettle from Samarqand(13), on a group of East Anatolian carpets from the 15th century(14), and on carpets in Timurid miniatures(15).

The above mentioned, almost ubiquitous ornament that is formed by two interlaced ovals appears in two forms. One is purely geometric (Fig. 2)(16) the other is formed by four snakes (Fig. 3)(17). The former is quite widespread on Islamic pottery(18) and metalwork(19) and can clearly be identified on Caucasian-(20) and Anatolian(21) carpets. To follow the successors of the latter, animated version is more difficult and perhaps more speculative. But again there seems to be a link to mediaeval metalwork from Khorasan(22) and to the Vakiflar Museum as a motif in the border of an Anatolian Carpet dated to the 16th - 17th century (Fig. 4)(23).

The fact that we can find a geometric and an animated version of the same ornament in the same cultural context is also interesting in relation to the S-motifs found on many oriental textiles. It has been asserted that they are derived from two-headed dragons and that those versions which show eyes and horns are the 'older' than the geometric ones(24).

There are a lot of (one-headed) dragons and snakes on BMAC seals and there may be a connection to representations of snakes on Turkmen carpets(25) and to the S-motif. But for the latter there are also possible geometric predecessors (Fig. 5)(26), which predate some of the suggested double-headed analogies by at least one thousand years. The tendency to add extra heads or other parts of an animal or a human body to a figure or an ornament, sometimes quite arbitrarily, is not restricted to the art of ancient Luristan or in the animal style of the steppe cultures. It occurs also in mediaeval Nishapur(27) and for example on Southwest Persian weavings of the 19th and 20th century(28). The common belief that the presence of animal heads in an ornament is an indication of its age is probably incorect.

Figure 5

The additional animal heads on a beautiful saddlebag(29) and some antique Luristan bronze cheek-plaques(30) led to reading the ornament of an important group of Lori bags(31) as a human figure sitting on a two-headed horse, which is itself again riding on another horse.

Based on a BMCA compartment seal and the design on a Luristan bronze, another interpretation seems more likely. On the compartment seal (Fig. 6)(32) we can recognize a female figure seated on a feline. The detail of the bronze center piece of a shield found in Luristan and dated to the 7th or 8th century BC (Fig. 7)(33) shows a human figure, again with raised arms, riding on a lion. The same can be detected on saddle bags some two and a half millennia later (Fig. 8)(34). There seem to be even further similarities. On every example from this group known to me, there is something beneath the mouth of the lions, which can probably be identified as floral on the 'degenerated' pieces dated to the 20th century(35). This floral element fits very well in P. R. S. Moorey's(36) interpretation of the whole ensemble as a representation of a fertility goddess. The feet of the lions, surprisingly, show no trace of claws and seem rather bovine. Curiously enough, that is also the case on the shield center as well as on some of the famous disk-pins from Luristan.

Figures 6 - 8

Lions, birds of prey, snakes and dragons depicted on the seal-amulets reflect an important part of the mythology of the BMAC(37). Even the representations of double-headed eagles (Fig. 9)(38) are found quite frequently. It is possible that the two-headed birds on a 19th century Arabatchi asmalyk (Fig. 10)(39) are actually derived from these models, thus marking a tradition lasting over 3,500 years(40).

Figures 9 - 10

Another traditional motif is two facing birds, sometimes sitting on a tree, which can be observed on both, stone- and metal seals from the BMAC(41). The same theme is reproduced on the Ordos bronzes from northwest China and on many early textiles from the near East(42). It can be observed on a 14th century Italian painting of a carpet from Anatolia(43), as well as on central Asian jewelry(44). Perhaps the pairs of birds so often found in the borders of Lori weavings(45) have the same roots.

Before looking at the more geometric kind of BMAC compartment seals, we should deal with a seal from clandestine excavations in Afghanistan (Fig. 11)(46). The similarity to the main motif of the Pazyryk carpet (Fig. 12) is clearly evident, despite the time difference of at least 1,000 years. A central Asian origin for the main ornament of this early carpet may seem unlikely. But the somewhat stepped 'flowers' of the seal are at least as close to the carpet as the 'Egyptian' ones represented on the 7th century BC 'threshold' carpet from Niniveh(47). Hints of a Central Asian origin of the Pazyryk carpet(48) and for some of its ornaments(49) support a possible link with the seal.

Figures 11 - 12

Of interest in this context are also the Bronze Age 'carpet knifes'(50) that have been found in South Turkmenistan, because they make likely a second, technical connection between Bronze Age Central Asia and carpets in addition to the ornaments(51) . Many of the BMAC compartment seals, which are sometimes well traveled, are based on a square or a cross inscribed in a lobed form (Fig. 13, 15)(52).

Figures 13 - 15

Based on other seals(53) and a seal impression from Shahdad(54), the lobed outline of these seals may be seen as four merged double-hooks (Fig. 14). And with these 'horns'(55) we enter a wide field of analogies with the world of carpets for which Fig. 16 and 17 may suffice as examples.

Figures 16 - 17

The remains of a highly important Late Bronze Age building from North Afghanistan, the palace of Dashly-III (Fig 18)(56) shows strong similarities with the discussed group of seals. Ignoring the area of the entrance, we get the picture of a well planed 'mandala'(57) (Fig. 19).

Figures 18 - 19

This strongly symmetric (and certainly not accidental) structure is reproduced on many carpets and other textiles from different times and areas (Fig. 20, 21, 22)(58).

Figures 20 - 22

As an indication that the similarities between the art of the Bronze Age and ornament designs are more then a mere coincidence, I presented various examples from a relatively narrow geographical area. As a hint for real continuity I pointed to possible links which help to bridge almost 4,000 years and show that in spite of the great ethnic changes in this 'crossroad of civilizations', there really are deep rooted traditions of motifs in Central Asia(59).

But on the other hand one has to be very careful to avoid jumping to conclusions. There is always more than one possibility (60). Finally, to show that the hints in this article are not to be construed as proofs, I would like to present two seal impressions from Susa. These could be seen as the source of Turkoman göls but their high age hardly can be reconciled with the material presented above.

Figure 23

These two seals (Fig. 23)(61) are described as crosses with two opposed animal-/birdheads on each end. The first one has additional animals (?) in the quarters between the arms of the cross. The center of the second seal impression has astonishing similarities with Turkoman weavings. Of course, this might be purely coincidental.









1. Opie, J., Tribal Rugs, London, 1992

James Opie also states that these ancient roots are best preserved in the weavings of Luristan thanks to the remoteness of their place of origin. He traces back the source of a multitude of motifs to the 'Animal Head Column', which is possibly related to the famous master-of-animals bronzes.

I do not presume to judge the validity of James Opie's arguments but, to encourage further discussion, I suggest alternative derivations for certain designs and point to their sometimes problematic nature.

2. Sarianidi, V. I., Preface in: Hiebert, F. T., Origins of the Bronze Age Oasis Civilization of Central Asia, Cambrigde MA, 1994, p. XXXI

3. P'yankova, L., Central Asia in the Bronze Age: sedentary and nomadic cultures, Antiquity 68 / 1994, pp. 355-72

4. Amiet, P., L'Age des échanges inter-iranien: 3500-1700 avant J.-C., Paris, 1986

5. Sarianidi, V. I., Die Kunst des Alten Afghanistan, Leipzig, 1986, Abb. 100

6. after: Sarianidi, V. I., Bactrian Centre of Ancient Art, Mesopotamia, 12 / 1977, Fig. 59 / 18

7. Bowl, Khorasan, early 13th century in: Melikian-Chirvani, A. S., Islamic Metalwork from the Iranian World, London, 1982, No. 26.

8. Erdmann, K., Der orientalische Knüpfteppich, Tübingen, 2. Aufl, 1960, Abb. 18

9. Balpinar, B.,Hirsch, U.,Vakiflar Museum Istanbul II: Teppiche - Carpets, Wesel, 1988, Pl. 62

10. Opie, J., Tribal Rugs, London, 1992, Fig. 4.16

11. Sarianidi, V. I., Soviet Excavations in Bactria: The Bronze Age, Fig. 11 / 9, in: Ligabue, G., Salvatori, S., eds., Bactria An ancient oasis civilization from the sands of Afghanistan, Venice, 1990

12. Opie, J., Tribal Rugs, London, 1992, Fig. 4.11

13. von Gladis, A., Islamische Metallarbeiten des 9. bis 15. Jahrhunderts, Abb. 219, in: Kalter, J., Pavaloi, M., eds., Usbekistan, Stuttgart, 1995

14. Balpinar, B.,Hirsch, U.,Vakiflar Museum Istanbul II: Teppiche - Carpets, Wesel, 1988, Pl. 37

15. Briggs, A., 1940, cited in: Pinner, R., Franses, M., Two Turkoman Carpets of the 15th century, in: Turkoman Studies I, London, 1980, Fig. 130.

16. after: Sarianidi, V. I., Die Kunst des Alten Afghanistan, Leipzig, 1986, Abb. 90. The same ornament, traced back to Anatolian and Syro-Hittite glyptics, can also be seen on a very interesting BMAC copper 'rattle-head' from the context of a ritual sacrifice ( Sarianidi, V. I., The Biblical Lamb and the Funeral Rites of Margiana and Bactria, Mesopotamia, 31 / 1996, pp. 33-48).

17. after: Sarianidi, V. I., Seal-Amulets of the Murghab Style, in: Kohl, Ph. L., ed., The Bronze Age Civilization of Central Asia, New York, 1981, Fig. 7. For a related example made of only two snakes and discovered at Susa, see: Amiet, P., L'Age des échanges inter-iranien: 3500-1700 avant J.-C., Paris, 1986, Fig. 65

18. Céramiques Islamiques dans les Collections Genevoises, Genève, 1981, Fig. 22. Kalter, J., Keramik des 9. bis 12. Jahrhunderts, Abb. 246, in: Kalter, J., Pavaloi, M., eds., Usbekistan, Stuttgart, 1995

19. Inkwell, Khorasan, 12th, 13th century, author's collection.

20. Shirvan, Late 19th century in: Tanavoli, P., Lion Rugs, Basel 1985, No. 57.

21. Balikesir, in: Zipper, K., Fritsche, C., Orientteppiche Bd. 3, Anatolische Teppiche, München, 1989, No. 9.

22. Lobed Cup, Khorasan, 10th century, Herat Museum; Rosewater-Sprinkler, Khorasan, 10th - 11th century in: Melikian-Chirvani, A. S., Islamic Metalwork from the Iranian World, London, 1982, Fig. 3; No. 5.

23. after: Balpinar, B.,Hirsch, U.,Vakiflar Museum Istanbul II: Teppiche - Carpets, Wesel, 1988, Pl. 28.

24. Opie, J., Tribal Rugs, London, 1992, pp. 52 ff

25. Rosetti, B., Die Turkmenen und ihre Teppiche, Berlin, 1992, pp. 205 ff. There is also the possibility of a relation between the gurbage ornament and the representations of frogs on BMAC seal-amulets.

26. BMAC compartment seal, author's collection. The S-border could also have developed from the segments of the 'cord motif' used as a border on many Near Eastern works of art (Amiet, P., L'Age des échanges inter-iranien: 3500-1700 avant J.-C., Paris, 1986, Fig. 87). This 'cord' can also often be observed separating the two registers of Syrian cylinder seals, and following them on the ones from the BMAC (Amiet, P., L'Age des échanges inter-iranien: 3500-1700 avant J.-C., Paris, 1986, Fig. 1192 / 3). This motif remains in use and can be seen, for example, on mediaval Islamic metalwork (Inkwell, 12th , 13th century, authors collection).

27. Richard Ettinghausen, Islamic Art and Archaeology, Collected Papers, Berlin, 1984, The Wade Cup, p. 363

28. See, for example, the human figures with hands in form of animal heads: Opie, J., Tribal Rugs, London, 1992, Fig. 4.42

29. Opie, J., Tribal Rugs, London, 1992, Fig. 4.30

30. Opie, J., Tribal Rugs, London, 1992, Fig. 4.28

31. Opie, J., Tribal Rugs, London, 1992, pp. 50f, pp. 122f.

32. After: Sarianidi, V. I., Reperti ineditti da tombe battriane depredate, Mesopotamia, 28 / 1993, Fig. 7.

33. After: 7000 Ans d'Art en Iran, Paris, 1961, Pl. XX

34. After a Luri bag face, author's collection. The picture of the goddess on a lion, common in Mesopotamia since very ancient times, had great influences, even to the Greek world (Möbius, H., Die Göttin auf dem Löwen, pp. 449 ­ 67, in: Wiessner, G., Festschrift für Wilhelm Eilers, Wiesbaden, 1967). A version which is remarkably close to those under discussion can be found in the four-armed Nanâ known from Sogdian paintings and woodcarvings excavated in Pendshikent (Mode, M., Sogdian Gods in Exile, South Asian Archaeology, 1991, Fig. 1d, e; Frumkin, G., Archaeology in Soviet Central Asia, Leiden / Köln, 1970, Pl. XXVI). Some medieval bronze Eastern Anatolian coins from the 12th and 13th centuries also show humans riding on felines (see Spengler, W., Sayles, W., Turkoman Figural Bronze Coins, Lodi, 1992, pp. 50, 127 ff., 145.

35. Opie, J., Tribal Rugs, London, 1992, p. 122, Fig. 7.22. James Opie's absolute dating of the weavings may stay untouched, but the question of degeneration has to be seen in a new light.

36. Moorey, P. R. S., Ancient Bronzes from Luristan, London, 1974, p. 33, about votive disk-pins found at Dum Surkh, which have strong connections with the shield center (Footnote 33; Fig. 6).

37. Francfort, H.-P., The Central Asian dimension of the symbolic system in Bactria and Margiana, Antiquity, 68 / 1994, pp. 406-18

38. BMAC compartment seal, author's collection.

39. Tzareva, E., Teppiche aus Mittelasien und Kasachstan, Leningrad, 1984, Nr. 109

40. Whether the animals in the Tauk Nuska göl, too, can be addressed as double-headed birds or animals is another open question. There are two points that have to be considered. First, unlike antique Iranian two-headed animal figurines, the 'heads' point in the same direction and in some cases the second 'head' at the back is not drawn in the same fashion as the first one (see for example: Pinner, R., The Animal Tree and the Great Bird in Myth and Folklore, Turkoman Studies I, London, 1980, pp. 244 f). Second, this additional 'head' is in harmony with the representations of tail and back feathers on Sasanian textiles, mediaeval animal carpets(a) and silver work (which seems to have common roots with cock appliques from the kurgans in Pazyryk(b)) as well as relatively recent carpets from different areas(c). From my point of view, the additional heads are actually tails and the (two legged) tauk nuska animals are normal chickens (or cocks). That is also exactly the explanation collected by V. G. Moshkova.

(a) Opie, J., Tribal Rugs, London, 1992, Fig. 15.6, 15.7

(b) Richard Ettinghausen, Islamic Art and Archaeology, Collected Papers, Berlin, 1984, Turkish Elements On Silver Objects of the Seljuq Period of Iran, pp. 1034-51

(c) Opie, J., Tribal Rugs, London, 1992, Fig. 4.18, 13.6, 14.1, 16.5, ...

41. Sarianidi, V. I., Seal-Amulets of the Murghab Style, in: Kohl, Ph. L., ed., The Bronze Age Civilization of Central Asia, New York, 1981, Fig. 17; Amiet, P., Bactriane Proto-Hisotrique, Syria, LIV / 1977, Pl. VI / 6, Fig. 21

42. For a further reading see: Pinner, R., The Animal Tree and the Great Bird in Myth and Folklore, Turkoman Studies I, London, 1980

43. Erdmann, K., Der orientalische Knüpfteppich, Tübingen, 2. Aufl, 1960, Abb. 8

44. See, for example, Janata, A., Schmuck in Afghanistan, p. 64

45. Opie, J., Tribal Rugs, London, 1992, Fig. 7.7 / 9, 7.12

46. After a foto from V. I. Sarianidi in: Brentjes, B., Probleme der baktrischen Bronzezeit, Iranica Antiqua 22 / 1987, Pl. 4a. For related pieces see: Pittman, H., Art of the Bronze Age, New York, Pl. 24j, and: Masson, V. M., Altyn-Depe, Philadelphia, 1988, Pl. XXIX / 10

47. Opie, J., Tribal Rugs, London, 1992, Fig. 3.8. For a Neo-Elamite version of the motif dating from the end of the second or the beginning of the first millennium, see: Collon, D., First Impressions: Cylinder Seals in the Ancient Near East, London, 1987, Nr 407, 408

48. Böhmer, H., Thompson, J., The Pazyryk Carpet: A technical discussion, Source, X, 4 / 1991

49. Hiebert, F. T., Pazyryk Chronology and Early Horse Nomads Reconsidered, Bulletin of the Asia Institute, New Series 6 / 1992

50. Khlopin, I. N., The Manufacture of Pile Carpets in Bronze Age Central Asia, Hali, 2 / 1982, pp. 116-18.

51. In this article we deal with Bronze Age materials, but according to Robert Pinner an even older design tradition related to Turkmen carpets cannot be excluded: Pinner, R., Decorative Designs on Prehistoric Turkmenian Ceramics, Hali, 2 / 82, pp. 118-19.

52. Fig. 13: BMAC compartment seal, author's collection. Fig. 15 after: Sarianidi, V. I., Soviet Excavations in Bactria: The Bronze Age, Fig. 13, in: Ligabue, G., Salvatori, S., eds., Bactria An ancient oasis civilization from the sands of Afghanistan, Venice, 1990. For the example from the Southeast Iranian Shahdad see: Hiebert, F. T., Lamberg-Karlovsky, C. C., Central Asia and the Indo-Iranian Borderlands, Iran, 30 / 1992, Pl. IIIb. For a related piece in the form of a cross excavated in the Mesopotamian Mari see: Beyer, D., Un nouveau Témoin des Relations entre Mari et le Monde Iranien au IIIème Millenaire, Iranica Antiqua, XXIV / 1989, pp. 109-20

53. Brentjes, B. Stempel- und Rollsiegel aus Baktrien und Chorasan, Baghdader Mitteilungen, 20 / 1989, Pl. 132 / 2, after a Foto from V. I. Sarianidi

54. Amiet, P., L'Age des échanges inter-iranien: 3500-1700 avant J.-C., Paris, 1986, Fig. 118

55. Nothing can be said here about the meaning of these 'hooks' or 'horns', if there is any at all. Considering the importance of ibex figures in the BMAC as pinheads, an interpretation as 'horns' cannot be excluded. If there would once be more support for a connection between carpet ornaments and the BMAC, one would have to take into account the religious background of these peoples, mentioned in the Rig-Veda (Parola, A., Margiana and the Aryan Problem, IASCCA Information Bulletin 19 / 1993, pp. 41 ­ 62). One would also have to examine the relevance of references to shamanic Sibira so often found in carpet books dealing with horns, eagles a.s.o.

56. after: Sarianidi, V. I., Die Kunst des Alten Afghanistan, Leipzig, 1986, p. 53

57. see: Brentjes, B., "Das Ur-Mandala" (?) from Daschly-3, Iranica Antiqua, XVIII / 1983. According to the linguist, Asko Parpola, this mandala is related to the Tantric Mahakali Yantra (Parpola, A., Margiana and the Aryan Problem, in, IASCCA Information Bulletin 19, Moscow, 1993. This could indicate that BMAC ornaments have survived for a long time, supporting the notion that some similarities are not accidental.

58. Fig. 20 after a carpet from Eastern Anatolia, 15th century in: Balpinar, B.,Hirsch, U.,Vakiflar Museum Istanbul II: Teppiche - Carpets, Wesel, 1988, Pl. 3. The very same design can be found on Caucasian carpets (Eder, D., Orientteppiche Bd. 1: Kaukasische Teppiche, München, 1979, Nr. 72) as well as on a Shahsavan khordjin (Tanavoli, P., Shahsavan, Fribourg, 1985, Fig. 36, Nr. 155) where it is interpreted as derived from the flower-and-bud-motif. The Turkmen variant of Fig. 21 is a bit crushed but still easily recognizable. Fig. 22 from an Uzbek embroidery in the collection of R. Christoffel, Steinmaur. Finally, Turkmen göls could be understood as derived from the same sources as the Dashly-mandala and the above mentioned compartment seals.

59. Another long lasting tradition, in the field of architecture and social organization, is described in: Lamberg-Karlowsky, C. C., The Bronze Age khanates of Central Asia, Antiquity 68 / 1994, pp. 398-405. The BMAC has to be seen as part of this motif tradition and as a link to the many other areas and cultures to which it maintained contacts. When we observe an ornament in the context of the BMAC, does this not necessarily mean that it also originated there. But the BMAC could be a good starting point for further research.

60. The negative forms of the Kaikalak can be misunderstood as animals (Opie, J., Tribal Rugs, London, 1992, Fig. 13.10-14), and ornaments that could be explained as having developed from the outline of the Dashly-palace (or the animal column) can be found on Near Eastern cylinder seals which are over one and a half millennia older ( Collon, D., First Impressions: Cylinder Seals in the Ancient Near East, London, 1987, Nr. 23, 62).

61. Le Brenton, Louis, A propos de cachets archaïques susiens, Revue d'Assyriologie et d'Archeologie orientale, 50 / 1956, Fig. 11, 2



Discussion