John Wertime on “Primitive Pile Rugs”

On February 15, 2014, John Wertime


gave a Rug and Textile Appreciation Morning program on “Primitive Pile Rugs.

Wendel Swan introduced John effusively. 


I will not attempt to repeat that, but the TM’s Members’ Magazine described Wertime as a “collector, dealer and writer” who would “discuss the possible origins and evolution of pile weaving in the ancient New East.”  It said further that “special attention will be paid to the continuation of ‘fossilized’ forms among tribal and nomadic peoples into the twentieth century.”

John’s program reprised and elaborated on his treatment of this subject in his seminal article “Back to Basics,” in Hali 100.



John began by recounting how his personal history sparked his eventual interest in the likely origins of pile weaving.

Wertime lived in Iran in Teheran in 1968-76.  He was a founding member of the Teheran Rug Society and associated with a number of knowledgeable and influential people there, including Jenny Housego and Parviz Tanavoli.  John said that this time was a “golden era for Westerners, as a great many textiles became visible in the market.  He was interested in various areas (for example, metal work) but the great variety of types of flat weaves especially drew his attention.

Returning from Iran, he worked with Irene Emory, the great student of the structures of fabrics, and published a piece on flat-woven structures.  Then, an article on weft-wrapping led to an interest in pile structures. The discovery of the quite sophisticated pile Pazyryk rug dated to 300 B.C., suggested a much earlier history.  Wertime has drawn on the work of Elizabeth Barber,


whose research and writings have mapped a great deal of this earlier period.

Wertime acknowledged that the study of the origins of pile weaving is speculative, but thinks that a possible chronology can be constructed on plausible arguments.  He distinguished “chronological age” from “conceptual age.”  He also thinks that glimpses of some of the early pile structures may still be visible in rugs woven in the 20th century, and in some types being woven even today.  He quoted a passage to this effect from something written in 1931. 

The new does not always slay the old, and some most primitive method may survive for special use, surrounded by the methods of superior culture.”

The earliest items that resemble pile weaving might have been wild animal pelts.  It seems that the construction of particular shapes and types of animal hides required thread-like material (likely sinew from the same animal that provided the hide), the invention of the needle (7,000 B.C.), and the notion of, and skill in, sewing. 

The availability of plant-sourced fibers for use as string (e.g. flax, 7,000 B.C.) predated animal fibers (5,ooo B.C.) that could be used in this way.  Flax had to be processed (including spinning, 1500 B.C. or earlier ) to create thread.  Once thread was available, both embroidery and simple weaving were possible.  The earliest plant-based textiles (plaiting a basket fita in here) seem to be plain-weaves or weft-wrappings.  Particular knives used in “wrap and cut” of “knotted pile” (3,000 B.C) may indicate the existence of pile.  In both of these woven structures there is a set of warps through which wefts are interlaced.  Weaving, as distinguished from plaiting or twining, requires that the warps be under tension.  Embroidery customarily involves a ground fabric affixed to a rigid frame.  Both warps and wefts can be selected for color from the animal’s wool (dyeing not needed) and that, in turn, permits woven designs.

An early source of animal fibers might have been those taken from wild animals, but this was chancy and a more predictable source of animal fibers had to await the domestication of animals.  Wertime suggested that the Fertile Crescent was an early place where domestication of animals occurred.


The modern-day countries with significant territory within the Fertile Crescent are Iraq, Kuwait, Syria, Lebanon, Jordan, Israel, Palestine, Cyprus, and Egypt, besides the southeastern fringe of Turkey and the western fringes of Iran.

The Fertile Crescent was primarily an agricultural society and it became necessary, as the number of domesticated animals increased, to keep them from feeding on the crops.  This might be a major reason for the development of what we call nomadic pastoralists, who migrate with their flocks seasonally to fresh pastures.  The animals (often sheep and goats) had to eat, but the agriculturalists did not want them to eat their crops.  So some members of society took the sheep away from the crops to other places where they could graze.

Another progression also occurred in domesticated animals, sheep especially.  The earliest domesticated sheep (7,00 B.C.) did not have much wool  (largely kemp) and it was mostly dark.  Such sheep were not a particularly good source of animal fibers and so were mostly raised for their meat and milk.  Through thousands of years, sheep developed longer, thicker and whiter wool, and were then raised for their wool and milk. Wertime said that the nomadic shepherds only ate meat on special occasions.

Eventually domesticated sheep became an important source of animal fibers, and, as this occurred, wool was increasingly used to make textiles.  As with plant-sourced fibers like flax, wool had to be cleaned, carded, and spun before weaving could occur and the early animal-fiber textiles were used initially in the same simpler structures (plain weave and weft wrapping) that plant based fibers had been.  Barber points our that it was also important that the animal fibers (wool in particular) had more of some physical properties, for example, resilience and “stickiness” from its scales) that made it easier to weave, than plant-base fibers had, and this made them more satisfactory in some weaving applications.

With this background of the prerequisites of pile weaving, Wertime offered what he thinks is the likely progression from flat weave to woven pile.  Here I want to “mine” his Hali article because he gives in it a nicely comprehensible illustrative sequence (Wertime, in turn, credits Irene Emory for most of its illustrations).

Note:  This is a place where you really need to click on the image below, to get a print size that is readable.


This is Wertime’s sequence, that he believes is the most likely evolution from flat weave to pile.

Now Wertime moved to the pieces that he had in the room.  Sometimes, if he didn’t have a given piece in the room, he held up his Hali article.  I have scanned most of these latter instances and to show you a larger image.John5

He began with a rug not woven but constructed using naturally dyed strips of pelt (left page, very small images, provided in larger versions below).

Below is the front of this piece.  It is composed of strips of pelt, dip-dyed and then sewn together in mostly concentric squares.  Concentric squares, we shall see, persisted in woven pile rugs and may have had symbolic meaning.


Although this is a very old technique, the rug is not old.  Wertime described it as “Kirghiz or Uzbek pile rug.  Central Asia, 19th or 20th century.

Below is the back of this piece with Wertime’s actual article caption on the right. (Again, click to get a larger image.)


Wertime next moved to the first of the pieces he had in the room. 



This rug was woven in southeast Turkey near or in the city of Siirt.  The Siirt environs are mountainous and in one of the “fertile crescent” areas where the use of animal fibers in weaving probably originated. This rug was woven in two pieces and then sewn together.

Siirt rugs of this sort were woven using undyed soft goat hair


(this is a Siirt goat; goat hair may have been employed in weaving before sheep’s wool).  There are no dyes in it at all.  The differences in color are from variations in color in the coat of the goat. 

It is woven with a plain weave structure,

W1 back



but then one side of it is raked with a plant-sourced tool (called a “teasel”)


that pulls up the fibers on that side, to form what is called a “faux pile.”  Likely, the “faux” reference is to signal that the pile fibers are not the ends of “knots.”  They are pulled up continuous wefts.


Below is an image of the “teasel” plant as it occurs in nature.


The Siirt rugs we have seem not to be old (they are still being made) but their “faux” pile  structure may be one of the very oldest moves from flat-woven textiles to those with pile.

There was another Siirt piece in the room.  Notice that W1 and W2 both have cotton warps.  Wertime indicated that the early versions of this weave likely had linen warps.



This piece illustrates not just that colors and designs can be produced from various shades of un-dyed goat hair, but that the area under the niche displays a further permutation of design produced by pulling the wefts up systematically in different directions to form a subtle, additional diamond design.


Here is the plain weave back of W2.


The two rugs John treated next in his in his talk are the ones in the image below.

W3 and W4


These two rugs have extra weft looped pile on a balanced plain weave.  There is no wrapping of the looped pile weft around the warps.

Here are closer views of them.



W3 is dip-dyed mohair.

Both of these pieces were woven in the Karapinar area of central Anatolia, where they are called “tulus.”


W4 un-dyed mohair.


Rugs like W4are still being woven in Turkey.  There is archeological evidence from 700 B.C. of this type of structure being woven then.  Notice that the concentric squares design apes that of W1, the rug composed of animal pelts.

Here are some detail views of W3 and W4.




Back of W4 below


The next piece had long-pile and a wonderful green color.  It was woven in angora wool in central Turkey and then dip-dyed after it came off the loom.



Here are some details of W5.


This is the first rug we’ve seen here that has “knots” tied on warps, with many intervening rows of plain weave.


Now we moved to two Scandinavian pieces that Wendel Swan had brought in and that had structures of the primitive type.



The first of these was an early 19th century Swedish rug, woven in two sections and then pieced.



Wendel said that this rug was made for use on a bed.  Wertime added that all of the pieces shown in this program were made for use as bedding, not on the floor.

Wendel said that the blue is probably linen, as are the warps.  The image below is a detail of both the front and back of this piece.  It has 8 to 10 rows of weft between each row of knots.


Each knot if half looped and half open.  The knots are pressed forward so that the pattern shows only faintly on the back.  Wertime said that this structure is a form of weft-wrapping.


Here are some additional detail images of W6.




Wendel’s second Swedish rug, with a primitive pile structure, was this one, which he attributes to Bohuslan in Southwest Sweden.


1828 rya a W7

This rug has looped pile (with the loops intact) and is woven so that the pile pattern is not visible on the back.


This back structure is weft-faced and so is distinctive from that of some Central Asian rugs that are woven with symmetric knots on alternate raised warps, producing a similar opaque, but warp-faced, back.

As you can see this rug is inscribed and dated.



Here are some additional detail images of W7.



Wertime said that he had included a blue Tibetan rug with a similar structure in his Hali 100 article,


but the loops in the Tibetan piece have been cut.

TibetanNicholas Wright

The Hali 100 article also gave you a look at the back of this Tibetan piece, showing that it is also weft-faced.


The next two rugs were Kurdish, woven with very long pile in east Anatolia.



(Wertime said that Saul Barodofsky, who has traveled Turkey for over 25 years, reports, plausibly, that long-rugs like W8 (it is almost 12 feet long) were used as sleeping rugs.  One places the rug on the ground, lays down on it and pulls the other half up and over, forming a kind of “sleeping bag” without sewn up sides.)

A “half” detail of W8.  (Color differences are from camera, lighting and color reproduction sources.)


Both W8 and W9 were woven with wool from sheep.  This wool has been dyed in different colors. 



The reverse sides of both of these rugs are heavily decorated.  Either side can be used as the visible one.

(I do not have an actual photo of the back of W8, but you can see areas of flat weave that would also be visible on the back.)


I do have some images of the decorated back of W9.



Here are some additional details of the front of W9.





John took us, next, to two rugs from northern Afghanistan or Central Asia that are similar in color, their weft pile and design.

John comparing sleeping rugs

They are W10 and W11.





W10 has warps and wefts of goat hair.


Here is its back.


It is a not a soft goat hair.  Woven in strips.  May have been woven by nomads.




W11 is woven in one piece.

W11bIts back is like a kilim.



Here are some additional details of W11.





The next piece was the one below.  Wonderful color.



Long pile, all wool, Uzbek.  Warp-faced ground weave; three rows wefts between knots.


Traces of the front design are almost invisible on the back.


The next two pieces were Arab julkyrs, woven in Uzbekistan or northern Afghanistan in the late 19th or early 20th century.





Here is a detail of W15.


This is what the back of W15 looks like.


The last rug that Wertime had in the room was the Yuncu piece below, woven in western Turkey in the late 19th century.



This rug has both long and short pile with intervening areas of striped plain weave and brocade decoration.  It is woven with symmetrically knotted pile on wool, weft-faced plain weave ground.


Here are some detail images of W17


Below is the back of this piece.  The flat woven stripes and decorative designs are visible.



Wertime talked about another interesting rug that he treated in his Hali 100 article.  It on the far right in the passage of his article that he is holding up.




John believes that this beautiful rug was probably woven in east Afghanistan by Pashtuns in the 19th or early 20th century.  It was woven of wool and goat hair in a rare structure.

It was woven using a slip loop on a plain weave ground.  The pile loops are held in place only by the plain weave wefts immediately above and below the knot rows.  John said that if you pulled on the side of a given looped pile row the entire row would come out.

I had brought the last piece of the day, which was likely off-topic for what Wertime was treating, but it had a simple structure and seemed an example of “primitive pile.”   I wondered whether it might be seen as an example of another path to primitive pile.



This is the front of a child’s rain cape made from plant fibers (coir) in southwest China by a non-Han ethnic group, the Miao.

Here is its back.


It is knotted, not woven. There seem to be no wefts.


But it has a very long pile and is made with a plant fiber, something used to make textiles before animal hair. 

It looks like a small bear’s pelt.


Wertime said that he had tried, in this session, to suggest how we got from primitive woven rugs with pile like this,


to sophisticated pile rugs woven with “knots,” like the lovely Persian Safavid carpet, below, that appears in an article by Murray Eiland, Jr. in the same Hali 100 issue in which Werime’s “Back to Basics” article appears.


Wertime took questions,


And brought his session to the close.

The surge to the front began.








I want to thank John Wertime for this authoritative, yet conversational program.  Amy Rispin took an excellent set of notes and both Wendel Swan and Wertime did the editing.

I trust you have enjoyed this permutation of Wertime’s seminal “Back to Basics” article in Hali 100.


R. John Howe

Comments are closed.

%d bloggers like this: