Since in his introduction Vincent has mentioned technical weaving factors that I talked about previously, I had probably better clarify. There are so many variables that can affect design compression and elongation in a weaving, I'll try to separate them.

MATERIALS

Cotton warps (as in Indian and Persian workshop carpets) are affected very little by changes in humidity and are stable, not stretchy.
Wool warps (more often used by village or peasant weavers) are inherently much less stable, are elastic, and are affected by changes in humidity, although they are superior in other ways.

LOOM VARIATIONS

The way the warp is attached to the two loom beams and the way warp tension is adjusted has a lot to do with the control the weaver has over her product. With a ROLLER-BEAM loom and fancy ratchets, she should be able to keep the tension almost constant. If the warp is CIRCULAR, and wound directly on the loom so that it goes around both end beams, the control exercised by the weaver is limited and tension is adjusted less frequently. In this case, as the weaving progresses, the warp may either stretch (due to humidity and constant pressure and agitation) or may tighten (due to inadequate weft ease and warp take-up). With looms on which the warp is wound separately and attached to the beams with HEADING CORDS, there is also much less control over warp tension, although there is usually some provision for adjustments. Thus the resulting elongation or compression of designs has more to do with the type of loom used and warping methods than with place or time of production. (If you'd like to see a variety of heading cords, you can take a look at the END FINISHES PROJECT on my website: www.marlamallett.com/end.htm.)

WEAVING PROCESSES

First, we need to remember that weavers engaged in making a rug are often unable to see much of their completed work. Finished portions are often rolled onto a lower beam, or pulled around the lower beam and up behind the loom. When weeks or even months are involved in the production, how are village weavers going to remember exactly how many rows of knots or wefts they used in each portion of the design? I've never seen any country weaver write down that stuff.

There is more likely to be vertical variation in tapestry-woven kilims than knotted-pile rugs. Knotted-pile motifs often require a set number of knots for their execution, while individual sections of tapestry-woven designs permit more flexibility. In other words, a pile motif may require 10 knots vertically, while the wefts within a similar tapestry motif may vary from 35 to 40. No weaver is going to sit and count those wefts; instead, she's going to guess when it looks right. Vertical variations within either may result from technical changes, although vertical variations in tapestry pieces are more likely to result from the more intuitive weaving process.

With kilims woven in two parts and seamed in the middle, it is the most natural thing in the world for the weaver to simply lay the completed section beside the loom in order to match the parts. She can lay the finished piece on the ground if her loom is horizontal, or even hang it over the loom, if her loom is upright. If she puts the model on the right (with the motif's center toward the new weaving), both pieces will be woven lower end first; if she turns her half-kilim lengthwise and puts it on her left, one piece will be woven bottom first, the other top first. On the WOVEN STRUCTURES UPDATE - 3 page of my website (www.marlamallett.com/up-three.htm) I have shown one way to determine which end is UP on a tapestry-woven kilim. Of course warp loops or heading cords also tell us which end was woven first.

Both warp and weft tension can affect the way a design elongates or compresses. Knots and wefts tend to pack down more closely if the WARP is TIGHT, and the design tends to compress. Conversely, if the WARP becomes LOOSE, the design tends to elongate.

A more taut WEFT means that the design tends to elongate. If the weft is instead given lots of ease, so that it can follow a sinuous path, knots compact better and the design tends to compact.

There are, then, lots of variables: the kind of warp and weft material, the kind of loom and tension-controlling provisions, the kind of weave structure and weave balance. Especially if different weavers work on a rug, we can expect changes.

For me, there is also an aesthetic component to this subject. Most artists, whatever their media, know intuitively that exact repetition is boring. Thus you rarely see artists and craftsmen measuring parts, but merely "eye-balling" them. (One exception would be when matching parts along a seam.) They know that slight variations give their work life and vitality. Whoops... that takes us back to internal elems, doesn't it!

Best,

Marla

Thank you for participating.
You say: "Knots and wefts tend to pack down more closely if the warp is tight, and the design tends to compress." What bothers me is the word tend. If the warp is tight the knots and wefts will need more warp space. They will be pushed upward. If the warp is adjusted because of elongation of the warp, (unwinding), the design will not elongate, the warps are stretched but this doesn't show in the design. The warps aren't made of rubber. If the weft tension is adjusted, no elongation visible. It would be a funny thing if this wasn't noticed in the pre-19'th century production.

Best regards,
Vincent

Sorry that I didn't make myself clear. This warp-tension stuff is something that every weaver learns about very quickly--especially struggling beginners! Just about everyone who has set up a loom has had difficulties with getting the warp tension even all across the loom. In areas where the warps are a little looser than the rest, the fell of the cloth (the edge of the weaving where the last weft is) curves UPWARD (toward the open warps). In other words, the wefts (and knots if there are any) do NOT pack down as well in those areas of loose warps. So to correct the problem, if the warps have been tied on at the front beam, the weaver simply reties the loosest ones, tightening them a little. Then the fell of the cloth straightens out. The TIGHTER the warp tension, the more firmly the weaving compacts...and conversely, the LOOSER the warp tension, the more the weave elongates. Of course all of this is much more evident on a loom with a reed and beater, where the beating is perfectly consistent across the weaving's width.

If you don't like my word "tends", just leave that out. The statement that bothered you is still true: Knots and wefts pack down more closely if the WARP is TIGHT, and the design compresses. Conversely, if the WARP becomes LOOSE, the design elongates.

Marla

If I understand it right, you're talking about loose and tight warps in one loom. This isn't the subject here. That's really sloppy. I'm trying to establish some ground for all the warps, same tension and getting more tensed as the work progresses. I was trying to look up some previous postings on the board, because we did adress this problem before. I couldn't find them. The postings "tend" to vannish from the board. In restoring a hole in a rug, the problem with warp tension is very obvious because the space that needs mending is limmited. If warps are to tight, the wefts can't get trough easy. The wefts need to be more curved and loose. If the wefts are pulled true straight, they will eat warpspace that is needed for knots/design. Design is lost. If the space wasn't limmited, the design could be knotted but it would be elongated.
That's why I'm stuborn about this.

Best regards,
Vincent

I understand quite well that your subject is warp tension, etc. as it applies to the entire rug, and actually may change during the course of the rug production. I was attempting to illustrate how warp tension differences affect the weave in an instance when one can clearly see differences side by side.

Of course weft ease plays a primary role as well in determining the weave density...It's one of the first things that all restorers learn, isn't it...that it is essential to insert wefts with ease so that they do not fill up a small area too fast!

Marla

There seems to be a difference.
-In one loom the knots and wefts tend to pack down more closely if the warp is tight compared to the warps that are loose in the same loom.
-In one loom the knots and wefts tend to decompress if the warps are getting tighter,(because of eating warpspace) and aren't adjusted correctly. This is elongation of design.
"Most people think it's because of the hammering down of the wefts and the weight of the finished rug, the bottom halve is compressed."
I think you said this in context with salon 50: "The blunted side".

Think this has been a very educational posting, Thank you Marla.

Best regards,
Vincent