HOME      Ordering WOVEN STRUCTURES      Woven Structures Update


7.A - Offset Knotting:  Where and Why?

Although offset knotting in rugs or bags has sometimes been noted in passing, there has rarely been much discussion of reasons for the feature or its possible use in making attributions. On these pages, we have posted examples from around the rug-producing areas so that they can be examined and compared.  

Jaf Kurd bags from western Persia (right), offer the most familiar examples of offset knotting. Since each rug knot is tied on two warps, it is easy for a weaver to stagger knots in successive rows. In this way she makes steeper and smoother diagonals, as shown in the diagram. Each design "step," normally the width of one knot, is reduced to the size of a half knot. Because offset knots are a disadvantage where straight verticals are desired, however, the weaver of this bag produced the diamond lattice-work field with offset knots, the border with knots aligned vertically in normal fashion.

In this instance, the weaver offset her knotting to copy a common kilim and brocade motif in pile. Unlike the large hooked figures featured on classic central or western Anatolian kilims, those woven by Kurds in both Eastern Anatolia and across the borders in Iraq and western Iran were made with smaller pattern units. To copy intricate tapestry motifs with normally aligned knotting would result in awkward, spindly latch-hook arms with steps proportionately too large along two sides. By offsetting knots, the Kurdish weaver made narrow stems only the width of two knots -- yet where properly articulated, stems that are smooth- edged and clear.  

Actually, we don't have to speculate on the development of the Jaf patterning above, as immediate predecessors exist. The identical motif appears in some of the purest reciprocal brocading of Eastern Anatolia -- work produced by both Kurds and others. In the brocaded chuval at the right, pairs of two-warp design units were offset consistently. The placement of these two-warp units could be copied precisely with offset knots. (See Tracking the Archetype for a discussion of the slit-tapestry origins of this kind of hooked motif.) 

We are usually quick to attach Jaf labels to knotted-pile bags displaying offset knotting along with hooked motifs in lattice arrangements; yet the weavings vary enough in their materials and color palettes to raise questions about such automatic attributions. The piece of Jerry Silverman's shown below was the topic of a Turkotek discussion a few months ago (Salon 61), but the group came to no definite conclusions as to its origin. This bag combines an offset knotted-pile hooked lattice motif of the Jaf variety with a soumak section featuring a motif rarely seen in Kurdish work. 

Knotted pile and soumak bag face. Jerry Silverman.

The motif in the Jaf bag below, and in detail at the right, is certainly close to that in the piece of undetermined provenance above. This version of the small hooked medallion surely evolved directly from slit-tapestry kilim patterning. Every element in the field is formed with diagonals, and all of the knotting is offset.  Because this weave has nearly twice the number of knots vertically as horizontally, only when knots were offset could smooth, steep diagonals be produced. The same is true of the first Jaf example above. 

Jaf Kurd bag face. Daniel Deschuyteneer.
For an excellent discussion of Jaf bags, see "Diamonds in the Pile," by Mark Hopkins, in the June/July 1989 issue of Oriental Rug Review, pp. 6-12. 

     Other Examples from West Persia and Anatolia

A group of rugs with hexagonal motifs clearly derived from kilim forms has been the subject of some controversy. The rug at the left, with nearly all of its field done in offset knotting, would seem to be from western Persia. Some related rugs with this design have single heavy wool wefts between rows of knotting, and also have overcast selvages -- features rare in Anatolian products. The rugs have, nevertheless, been attributed variously to Persia and Eastern Anatolia. They have sometimes been called Sanjabi or Jaf Kurd. On one example that I examined recently, the ends were finished with obliquely wrapped bands. (See the End Finishes Project, Obliquely Wrapped Bands for a rug belonging to Allan Arthur.) The finish on the piece shown at the left may be the same; it is impossible to determine without a closer look. 

Offset knotting in the rug's field. These motifs are copies of kilim designs with narrow pattern units, and are made entirely with diagonals.  [Auctioned on eBay, June 2001, Number 1438510017. Said to have an all wool foundation, with 25 symmetrical knots per square inch.]
A rug in the Vakiflar Museum, Istanbul (No. 169 D), is an early example of the genre represented above. It has the same hexagonal motifs arranged in bands, but with an even more common kilim motif in the center of each. The entire field was executed in offset knotting. The white triangular drops along the hexagon edges in the previous piece seem to be later additions. The minor motif used between the hexagons, is similar to that in the Jaf bag above. 

This rug has been identified tentatively by Nazan Olçer as from "Eastern Anatolia (?), 17th century." It is Plate 59 in Nazan Olcer et al, Turkish Carpets from the 13th-18th Centuries, Istanbul. 

(On the following pages you will find other early examples of Kurdish offset knotting.)  
Another group of Kurdish rugs with offset knotting, in a pattern sometimes dubbed "baklava design," comes from eastern Anatolia. These are sometimes attributed to the area south of Malatya, other times to the Sarkisla area, northwest of Malatya. Instead of the pattern of hexagons seen above, these rugs have allover diamond- shaped hooked medallions, usually arranged in panels, but occasionally in continuous field arrangements. These rugs typically have reinforced selvages and color palettes that include a deep apricot, cochineal magenta, and deep blue -- a palette quite different from the Kurdish rugs with hexagonal patterning. 

As with the previous examples, the pattern parts in this design are so narrow that they could only have been articulated in pile by the use of offset knotting. In many of these rugs, pattern detail is nearly obscured by long, shaggy pile. 

Again, a parallel is found in the flatweaves widely produced by nomads in the area -- narrow hooked medallions traditionally arranged in two wide bands on decorated storage sacks. The piece at the left is a rare variety of tapestry offset on paired warps.  Nearly identical patterning was produced in reciprocal brocading in the Malatya area. In each case, the  latch hook stems were composed of pairs of two-warp offset units -- the feature imitated by the pile weavers. In each of the structures, hooks are delineated by contrasting outlines.  

Tapestry offset on paired warps decorates an ala çuval from the Malatya area of eastern central Anatolia. Similar patterning was produced in reciprocal brocading. In this example, each tiny segment was produced with about six very thin wefts. 

Detail of the Anatolian Kurdish rug [eBay, Number 1438503749, June 2001.] The back side is shown below.
In both the detail above and the back view below, we see offset knotting throughout the field of hooked diamond medallions. In contrast, the white undulating borders have been produced with regular, vertically aligned knotting. 
It is easy to compare the stepped motifs in the white vertical border with the smooth diagonal forms of the white knotted hooked medallion in the lower left. The outer vertical border is also done with regularly aligned knotting. 
An unusual Luri rug from Western Persia, circa 1950, shows extensive use of offset knotting to shape design elements.  Here there can be no question about the weaver's intent, as diamonds of two contrasting shapes appear in the detail below. The knot count in this rug is balanced:  it has 5 knots horizontally and 5 knots vertically.  Thus in the right-hand diamond, diagonals incline at a 45 degree angle. By offsetting knots in the diamond on the left, the form has been stretched vertically. This weaver also chose to elongate the motif in the rug's vertical borders. Among scattered field motifs, she drew small birds, water pitchers and even small human figures using the steeper diagonals created by offsetting knots. 

Daniel examined this Lori rug and mentioned that he found it full of irregularities: weft inlays, discontinuous wefts, discontinuous knotting, stacked knots and overlapping knots. Since this weaver took an irreverent and experimental approach to the rug's execution, it is perhaps not surprising that she also freely offset knots in her design.  
In northwestern Anatolia, reciprocal brocading was used extensively by Turkic nomads for decorating storage sacks. Harald Böhmer, in a 1990 ICOC lecture, pointed out striking similarities between the patterning in Karakeçili sacks (below) and one type of knotted-pile rug with offset knotting -- rugs like those shown in details at the right. The carpets were knotted with offsets to exactly duplicate the storage sack motifs produced with offset two-warp units; they were apparently produced in the early 19th century, when the first of these nomads settled.  As with the reciprocal brocade piece and Jaf Kurd bag at the top of this page, there can be no doubt that the brocaded designs directly inspired the pile products. Although the diagonals are steeper in this carpet than on the brocaded sack, offset knotting was not used for this reason. Instead, it allowed the weaver to produce the design with identical pattern units: All through the design each offset interlacing of two warp pairs was represented by two offset knots. 

Note: The dominant flatweave structure below has sometimes been described as "wrapping" or soumak. That is incorrect; the patterning is composed of interlaced supplementary wefts. It is reciprocal brocading -- the majority of the patterning in this case done with two-warp units. Soumak is used only for selected horizontal lines in the design.  

Detail of Anatolian carpet shown by Adil Besim in the Milan ICOC fair, 1999.  

Detail of a village carpet, James Opie, Tribal Rugs, Portland, 1992, Plate 15.16. 

The published version of Dr. Böhmer's  lecture appears in Eiland, Pinner and Denny, eds., Oriental Carpet and Textile Studies, Vol. IV, Berkeley, 1993, pp. 57-66. In that paper he also shows several other examples of carpet motifs that have derived from the brocaded ala çuval of western Anatolia.   

The northeastern Anatolian sacks are discussed and shown in photos in Pinkwart and Steiner, Bergama Çuvallari, Wesel, 1991. The illustration here is from Plate 11.

These particular carpets were also discussed in a  Turkotek Salon, Number 31, presented by Daniel Deschuyteneer.   

1690 Johnson Road NE, Atlanta, GA  30306  USA

E-mail:  marlam@mindspring.com
Phone:  404-872-3356

Woven Structures Update