Knotendichte selbst prüfen — und was sie wirklich aussagt

Knot density is the one quality figure in a rug that you can verify yourself, without expertise and without instruments. It takes a ruler, two minutes and the willingness to turn the rug over.

That is worth pausing on, because almost everything else — the wool, the dyeing, the origin — needs either years of handling or a laboratory. Which makes it all the more important to understand what the figure tells you and, above all, what it does not. Buy on knot density alone and you will regularly buy past your own requirement.

Counting it yourself

Counting always happens on the back. There every knot sits as a small visible pair of dots, while on the face it disappears into the pile.

How to go about it:

1. Turn the rug over and pick a spot in the open field — not near the border, not inside a medallion. Pattern can change the density locally.
2. Lay a ruler across and measure 10 centimetres (4 in) of width. Count the knots along that line. On fine pieces a magnifier, or the zoom on your phone, makes this much easier.
3. Do the same over 10 centimetres (4 in) of length, at right angles to the first measurement.
4. Multiply the two counts, then multiply by 100. That gives you knots per square metre. For kpsi, divide each count by four first — that turns it into knots per inch — and multiply those two figures instead.

That the two counts rarely match is normal and not a fault: knots usually sit closer together in the warp direction than in the weft direction. To a trained eye the ratio between them even hints at the weaving region.

A word on accuracy: count in two or three places and take the average. On a hand-knotted rug the density varies across the field — that too is normal, because it was made by hand and not by a machine.

Which knot — and why it moves the number

While counting you will notice that the knots do not look the same everywhere. There are essentially two techniques, and they are the reason two rugs of the same density can read as different degrees of fineness.

The asymmetric knot — commonly called the Persian or Senneh knot — loops the yarn around one warp thread and passes it behind the second. It allows higher densities and finer outlines, because the rows of knots can be set closer together. On the back you recognise it by the two ends of a knot sitting unevenly rather than symmetrically side by side.

The symmetric knot — Turkish or Ghiordes knot — wraps both warp threads evenly. It grips more firmly and stands up better to wear, but in very fine patterns it tends towards stepped outlines. On the back it forms a regular grid that is easy to count.

One further term is worth knowing, because it turns up in English-language trade descriptions: the jufti knot, tied over four warp threads instead of two. It covers the same ground with half the knots and roughly halves the weaving time. It does not inflate the figure you measure — that figure is genuinely lower — but there is less wool in the pile, and the pile thins sooner.

In practice this means a symmetrically knotted rug at 200,000 knots is not worse than an asymmetric one at the same figure — it was made for something else. Robust nomadic and village work is largely symmetric, fine workshop pieces largely asymmetric. Compare the number alone and you are comparing two things that have different jobs.

The units — and why they disagree

Two dealers can describe the same rug and quote completely different numbers without either of them exaggerating. The units are the reason.

The conversion that matters most is the one between kpsi and knots per square metre, because it comes up constantly in international trade. A square metre holds around 1,550 square inches — so 160 kpsi is roughly 248,000 knots per m². The same quality, two numbers that look as though they have nothing to do with each other.

Raj calls for caution. It counts the knots across seven centimetres of width. From 50 raj you can derive around 700 knots per running metre and, if the density in the other direction is similar, an order of magnitude of roughly half a million knots per square metre. That is an approximation, not a conversion — it assumes warp and weft density are close, which is not always the case.

What the density does not tell you

This is the real point of the article. Knot density is one factor, not the factor — and about the four things below it says nothing whatsoever.

The wool

Wool from the back of a highland sheep, shorn in autumn, is a different fibre from belly wool, and different again from wool recovered from skins. The difference lies in the lanolin content and the staple length, and it decides whether the rug still has lustre after ten years or has gone dull. A dense rug in poor wool ages worse than a coarser one in good wool.

The dyeing

Whether plant dyes or synthetic dyes were used, whether the colours are lightfast, whether they bleed in washing — none of that has anything to do with the number of knots. All of it decides what the rug looks like in twenty years.

The spinning

Hand-spun wool is uneven, and that unevenness is exactly what produces the living depth of colour you see in good pieces — the fibre takes dye differently along its length. Mill-spun yarn is more even and reads flatter. Both can be worked at the same density.

How cleanly it was executed

Do the rows run straight? Are the selvedges firm and even? Is the rug square, or does it pull to one side? A very dense rug that was woven crooked stays crooked for the rest of its life.

Which density for what

The useful question is not “how fine can it go” but “what is the rug for”. Framed that way, the answer almost always differs from the one the prestige of the highest number would suggest.

The practical consequence: a rug that is walked on every day gains little from very high density. It gains from good wool. A very fine piece, by contrast, only opens up its motifs where you look at it rather than walk across it. How these gradations relate to one another is set out on Hand-knotted rugs.

How to spot a number being dressed up

Three warning signs that can be checked without any expertise:

A density figure without a unit. “800,000 knots” is not a statement — per square metre, per rug, per square inch? The number on its own says nothing, and sometimes that is the intention.

A density figure without a material figure. Density and fibre belong together. Quote the one and leave out the other, and you are emphasising the metric that sounds good while omitting the one that is harder to dress up.

A headline figure with no connection to the actual piece. “Up to X knots” describes what a house can have made in principle — not what the rug in front of you has. Ask for the density of that one rug. Anyone who cannot name it has not measured it.

That applies explicitly to figures on this website, and to every other. The number that counts is the one you can count for yourself on the back.

When you need certainty

Counting yourself answers the question of density. It does not answer the question of value — that requires material, dyeing, age, condition and origin to be judged together. Where that matters, with an inherited piece or before insuring one, a written appraisal is the right route.

And if you simply want to know what different densities feel like: that is a question you do not read, you pick up. Two rugs side by side, turned over, a hand run across both — the difference is obvious in two seconds and still abstract after two paragraphs of text. The traditional collections are a good place to start.