A cloud-band runner, a finely patterned Kuba rug and a flatwoven Soumak can all belong under the heading Caucasian rugs. What joins them is a regional weaving history, not one design or one construction. Their differences are part of the interest: a long runner can build its pattern through repetition, while a smaller rug can concentrate attention on a tightly organized field.
This guide separates the regional names from the weaving terms, then looks at actual examples. If you are starting with the wider subject, our guide to types of Oriental rugs places Caucasian weaving alongside other traditions.
What does “Caucasian rug” mean?
The term refers to rugs associated with the Caucasus, the region between the Black and Caspian Seas. It is a geographic umbrella, not the name of a single people. The Metropolitan Museum of Art describes a multiethnic weaving environment in which neighbouring communities shared designs. A rug's pattern therefore should not be treated as a certificate of its maker's ethnicity.
There are also several kinds of name in play. Kazak, Kuba, Shirvan and Karabagh are regional or market classifications. Soumak describes a weaving structure. Runner describes a format. These terms answer different questions, so “Caucasian Soumak runner” is not a contradiction: it combines an attribution, a technique and a shape.
Read a catalogue description as a set of claims with different levels of certainty. An institution may confidently identify wool pile but use “probably” for a district. That qualification is useful information, not something to remove when retelling the object's history.
A longer history than the familiar nineteenth-century rug
Nineteenth-century examples are important to the study of Caucasian rugs, but they are not the beginning of the story. The Met's Dragon Carpet is dated to the seventeenth century and attributed to present-day Azerbaijan, probably Quba or Karabakh. Its record identifies wool warp, weft and pile, with symmetric knots. The district remains qualified even in a detailed museum catalogue.
Look at how the angular ornament travels through the red field rather than gathering around a single isolated centre. This is a sprawling composition: follow one diagonal and another interrupts it, drawing the eye across the rug. It offers a different point of reference from the small geometric rug that readers may first associate with the region.
Kazak, Kuba, Shirvan, Ganja and Karabagh
Richard E. Wright's survey in Encyclopaedia Iranica offers three broad groupings for pile rugs. These are useful comparisons, not a complete list of local types or a formula for attributing every surviving piece.
- Kazak and Borchaly: western groupings often associated with larger, more spacious motifs, coarser yarn and deeper pile.
- Kuba and Shirvan: eastern groupings often associated with finer yarn, denser knotting and more closely arranged ornament.
- Ganja and Karabagh: groupings that can combine features of the others.
Spellings vary between catalogues. Kuba and Quba, or Karabagh and Karabakh, can appear in discussions of the same regional tradition. Keep a source's attribution with its object rather than silently sharpening a broad regional name into a precise village of manufacture.
Kazak: rhythm in a long runner
This runner's field is not dominated by a few widely spaced medallions. Small forms accumulate along its length, contained by a strong red border. The contrast between the dense field and the emphatic border gives the narrow format much of its character.
It is one example from our Kazak rug collection, not a template against which every Kazak-labelled piece should be measured.
Kuba: a floral field and an informative foundation
In this example, rows of floral forms give the field a compact rhythm. The archival photograph is useful for studying that arrangement, but it cannot tell us how the original colours look in person.
The museum records wool warp and pile, but a weft combining cotton and wool. This single documented object is enough to challenge the absolute claim that every Caucasian rug must have an entirely wool foundation. It does not establish the opposite rule that Kuba rugs always contain cotton.
Karabagh: following the cloud bands
The opening detail belongs to this runner. Seen as a whole, the hooked shapes become a repeated sequence rather than isolated symbols. Red, dark blue and lighter accents alternate across the field; the enclosing borders give that activity a clear edge.
In the detail, the stepped outlines dominate. Across the whole runner, the intervals between them become just as important. “Cloud band” is a useful name for discussing these forms, without assigning a fixed spiritual meaning to every hook.
Soumak and kilim: construction, not competing places
A piled rug and a Soumak may share angular motifs but achieve their surfaces differently. In Sarah B. Sherrill's account of flatweave structures, kilim or gelim is associated with weft-faced tapestry weaving. Soumak, also written sumak, uses weft wrapping: the working weft wraps around warp threads. It is not simply a pile rug with its wool cut very short.
There are different wrapping arrangements, and not all flatweaves are structurally identical. The distinction matters more than treating “flatweave” as a single texture. Our guide to kilims and related textiles explores that wider vocabulary.
In the example above, the medallions offer an easy visual comparison with the long runners. The broad composition reads differently, even though both use angular ornament. The catalogue's Soumak designation concerns the weave, not the number or shape of those medallions.
A historical counterpart appears in the Textile Museum of Canada's collection: a wool Soumak dated 1850–1860 from the Kuba area, attributed to Lezghi or Azerbaijani makers. Its record keeps place, cultural attribution and technique as separate fields. That is a helpful model for reading rug descriptions without collapsing several kinds of evidence into one name.
Why Shahsavan and Caucasian motifs can look familiar
Persian-language scholarship adds an important cross-border perspective. A comparative study by Fattaneh Mahmoudi and Mahnaz Shayesteh-far in Goljaam examines similarities between Shahsavan and Caucasian weavings. Its examples include hooked lozenges, crab-like forms and S-shaped motifs. The authors discuss the difficulty of distinguishing some earlier textiles when their patterns are so closely related.
The practical lesson is to resist turning a shared motif into an exclusive regional signature. A resemblance can be worth investigating without proving an exact origin. The study compares particular examples; its observations should not become a universal colour chart or an authentication test for every rug bearing a hooked diamond.
Reading the rug beyond its label
The examples here invite three different kinds of looking. The Dragon Carpet opens a historical comparison. The Kuba record shows the value of documented materials. The runners and Soumak show how strongly format and arrangement shape our experience of a design.
Start with what the object actually presents: a dense or open field, a repeated unit or a central focus, pile or a flatwoven surface. Then ask what the catalogue adds, and where it remains uncertain. This keeps the pleasure of looking at the rug connected to the evidence, without expecting one ornament to answer every question about its past.
For pieces in our catalogue, browse the Caucasian rug collection. The individual records provide the dimensions and condition information that a regional overview cannot supply.
Sources and further reading
- Richard E. Wright, “Caucasian Carpets”, Encyclopaedia Iranica. Regional groupings and the limits of classification.
- Sarah B. Sherrill, “Flat-Woven Carpets: Techniques and Structures”, Encyclopaedia Iranica. Tapestry weaving and weft wrapping.
- The Metropolitan Museum of Art: Dragon Carpet, 22.100.119, and White-Ground Kuba Rug, 22.100.5. Object dates, attributions, materials and Public Domain images.
- Textile Museum of Canada, Soumak rug, T94.2238. Kuba-area object dated 1850–1860.
- Fattaneh Mahmoudi and Mahnaz Shayesteh-far, comparative study of motifs in Shahsavan and Caucasian weavings, Goljaam, issue 11, pp. 25–40. Persian-language research, particularly pp. 36–40.
Leave a comment