Anatolian kilim

What a kilim is, technically

A kilim is a flat weave, and more precisely a weft-faced fabric. The warp threads, wool in traditional Anatolia, are stretched on the loom and will stay invisible. The coloured woollen wefts pass over and under, are beaten down with a comb, and end up covering the warp completely. The fabric is the design; there is nothing on top of it.

Three immediate consequences. There is no knot: nobody knotted anything on this rug. There is no pile: nothing stands above the surface. And both faces are right sides: a kilim is reversible, identical on both sides, and it has no back where work could be hidden. It also weighs a fraction of a knotted rug of the same size.

Alongside the kilim proper, Anatolia has other flat weaves, embroidered or wrapped on the same structure: the cicim, the zili, the soumak. The soumak, where the weft is wrapped around the warps, is not a kilim in the strict sense, but it is handled and washed in the same way.

The slit: what makes the design and what weakens it

The dominant technique is slit-weave, which has given the type its name. Where two areas of colour meet along a vertical line, each weft turns back on its own warp without crossing its neighbour. An open slit is left between them. That is what gives the kilim outlines of a crispness the knotted rug does not reach.

It is also why the weavers avoid long verticals: a vertical twenty centimetres long would leave a slit twenty centimetres long. They shift the line by one warp at each pass and obtain a stepped diagonal. The well-known sawtooth design of the kilim is not a decorative choice, it is a structural constraint that became a style.

The slit is therefore normal, and it is the weak point. A pull, a twist, a rug dragged by one corner or wrung out roughly lengthens those slits millimetre by millimetre, until two areas separate outright.

Two halves sewn down the middle

Nomadic looms are narrow. Many Anatolian kilims were therefore woven as two strips, then sewn edge to edge down their length. The seam is visible down the middle, and you can often see that the two halves do not match exactly: a motif out of line by a few centimetres, a slightly different shade because the two strips were not woven at the same moment.

This is not a fault and it is not a repaired rug. It is how it was made. The central seam is, on the other hand, a mechanical weakness and very often the first thing to open. It is taken up again with a needle, in the original stitch.

Regions, motifs, colours

The great names come from Konya, Karapinar, Obruk, Aksaray and Cappadocia in the centre; from Malatya, Sivas, Erzurum and Reyhanli in the east and the south-east; from Bergama, Balikesir and Manisa in the west. Each area has its formats and its ranges, but the repertoire of motifs is largely shared: the elibelinde, a female figure with her hands on her hips; the kocboynuzu, ram's horns; the eye and the amulet; the comb; the scorpion; and, on prayer kilims, the mihrab.

The old palette comes from madder, indigo, the yellows of weld and vine leaf, walnut husk, and naturally ivory or brown wools. Synthetic dyes appear from the 1870s-1880s and become general thereafter: that is the main dividing line for dating a piece. Whole eighteenth-century kilims are rare and often fragmentary; what circulates is overwhelmingly nineteenth century and later.

The value of a kilim, and what destroys it

An old Anatolian kilim, in natural dyes, with its original ends and its whole format, has a real market, held by collectors and specialist dealers. That market is as demanding as it is fragile.

What destroys the value, in order of frequency: a large kilim cut up to make cushions or table runners; fragments assembled and sold as a whole piece, which shows in raking light; general stitching in synthetic thread over the whole surface; and above all a kilim glued to a backing cloth. Gluing is almost always irreversible: taking it off carries fibre away.

One point should be added, because the question comes up: we buy antique kilims. Their value is not read off a knot density, which does not exist here, but off the design, the age, the fineness of the weave and the quality of the dyes — it is a narrower and more specialised market than that of the knotted rug, and we say so at the valuation.

In the workshop: nothing to repile, everything to reweave

Since there is no pile, there is nothing to repile. No knot to redo, no tuft to reset. Restoring a kilim is a reweaving job: replacing the broken warps, rebuilding the wefts one by one with a needle or on a loom, closing the open slits in the original stitch. It is slow work, and over a wide gap the repair will stay visible on both sides, because there is no pile to cover it. We say so beforehand.

Washing follows the same logic. The dust has not sunk to the bottom of a pile: it is caught in the fabric and it goes right through it, so the dusting is done on both faces. The bath is cold, with neutral soap. A kilim is never twisted to drain it and never wrung out roughly: that is the movement that lengthens the slits. Drying is strictly flat, on a rigid surface, with a gentle, symmetrical re-tensioning while the piece is damp.

Rippling comes from that. A knotted rug has a pile that absorbs differences of tension; a kilim has none, and the least irregularity reads on the surface. If the warp is cotton and the wefts are wool, the two materials do not shrink in the same way as they dry, and a wave caught in the drying becomes permanent. Washing a kilim is in fact judged as much on the drying as on the bath.

One last point, often overlooked: a kilim hung on the wall from two corner nails distorts under its own weight within a few months. It must be hung from a sleeve sewn across the full width, never from isolated points.