...that is the question! Have you decided, dear reader? I suspect you have. Everybody who comes into contact with rugs and carpets sooner or later will have to make up his or her mind on this question.
Opinion on restoration lies between two extremes: to restore whatever can be afforded without any pangs of conscience, and to never artificially change the extant condition apart from some cleaning and conservation. After many discussions on the subject of restoring rugs, I have found that individual opinion seems to vary in a wide range from a pragmatic decision at an individual case to the dogmatic belief to leave everything as it is. As I am not impartial in the matter, to write on this subject I have tried to consider relevant aspects and to do justice even to the extremes.
I think it might be useful to begin with a few statements which are obvious and non-judgmental. Later we might come to a discriminative conclusion:
* Antique pre-1800 rugs and carpets almost always show wear and damage which often has been more or less skillfully repaired at some time in the past.
* The older a carpet, the more likely is extensive damage or wear to be.
* At the time of their making, most carpets and rugs were made to be used, not collected.
* If damage occurs through use which results in a hazard to personal safety, such as holes and cuts, a decision has to be made to either repair the piece or replace or even discard it.
Fashionable fragments, mainly of Anatolian origin, which had been (illegally) removed from Turkish mosques by local pickers, were sold to specialist dealers in the West and remained unrestored. Those dealers, in turn, convinced leading collectors that wear and missing parts should be ignored if the rug fragment has age and noble provenance. Stabilizing the piece to prevent further deterioration respects the integrity of the weaving.
| A specialist at work on a densely knotted Kashan carpet, Izmer |
Before 1800 no market for tribal goods existed, neither in Europe nor in the United States, and hence, when they wore out from use, functional pieces like saddle bags, mafrash, horse covers, and tent bands were discarded by those who used them, the nomads and villagers. It was not until the 19th century, when traveling dealers, not museum curators or art historians, recognized the beauty of nomadic folk art, that those more or less damaged pieces entered Western collections, were preserved from being totally ruined, and collectors faced the decision to restore or not to restore.
Peter Stone, in his 1981 book, Oriental Rug Repair, wrote in his introductory comments: "The primary requisites are attention to detail and patience." He continued, "Knowledge of oriental rug repair techniques can be helpful even to those who don't intend to apply those techniques. Dealers and collectors can use this information to evaluate the repairability and price of worn or damaged rugs on the market. Perhaps the greatest reward in repairing oriental rugs is the knowledge that one is preserving something beautiful. And that the effort of repair and restoration benefits everyone who will find pleasure in viewing the rug."
Stone discerns three steps of improvement. Conservation, which is the softest treatment, is "preserving an object with as little change to the object as possible. The goal is to prevent further damage or deterioration." Deterioration may be due to chemical changes, insect or microbial attack, exposure to heat, light, or moisture or from the way an item is displayed or stored. Restoration on the other end is "an attempt to return an object to its original or first condition." Only an approximation is feasible, even if the restorer uses the same type of animal or vegetal fibers, the same spin and ply and the same construction features. The restored rug is by definition not identical to the original rug, regardless of whether the restored parts are invisible from the front or -- much more critical -- from the reverse side.
The struggle for perfection sometimes has caused much more damage to other antique textiles. In order to obtain naturally dyed woolen yarn for restoration, fragmented Anatolian kilims have been taken apart to recover their old wool, destroying forever the great works of art they might have been. Nowadays kilims are highly respected, artistically and commercially prized, and their brutal treatment and use as wrapping material has been abandoned. Yet the dismantling of old pieces to gain wool for restoration continues, namely of warp-faced jajims, which are undervalued, the literature having so far taken little notice of this special type of blanket. The author has saved two dozen of them from perishing and acquired them in the final hour.
 | George Jevremoviç, David Bamford, and Bob Mann discuss rug restoration at Izmir, April, 1995 |
Some 1500 years later, when the Chinese Jingdezhen kilns of the Ming dynasty produced enormous quantities of vases, cups, dishes, etc., the blue and white ware so highly valued today, many of these proverbially Ming vases came on the market, often only very slightly damaged. The Chinese are skilled and artistic craftsmen; they were and still are also masters of restoration. As the Chinese taste strives for perfection in works of art, the restorers aim at the invisibility of their work. So hairline cracks, chips to rims, or defects in glaze disappear under artful restoration, and the beauty of a piece, even at close inspection, is excellently preserved. This repair is not and should not be detrimental to the market value; only a specialist as knowledgable and skilled as the artist restorer can detect later additions. A Ming moon flask in perfect condition, worth $400,000, would be unsalable to a Chinese collector, however, if one handle is lost. Thermoluminiscence (TL) analysis, an expensive scientific method, can easily reveal later manipulation to ceramics. Bronze vessels from the Shang dynasty, about 3000-4000 years old, usually are not so often subject to extensive restoration but to careful conservation. Buried in tombs, they withstood the ages unharmed.
Key points for restoration seem to be age, integrity of the object, and cost of restoration related to total worth.
When we come to rugs and textiles, the adrenalin level rises and arguing becomes more subtle as most readers are concerned in one way or the other. With regard to textiles, there is very little controversy and greater agreement; usually they are more difficult to restore than are rugs, soumak or kilim flatweaves. For instance, a Safavid double weave cloth has such a complex structure that it is nearly impossible to reweave a significant area. In most cases, a restoration would not make much sense as silk cloth often was woven for garments in an endless repeat design, the pattern repeat not more than 20-40 cm [8-16 in.] in length and width. Silk brocades several hundreds years old have withstood the ravages of time as fragments, and nobody would consider a treatment of them other than conservation.
With the repair/restoration of rugs and carpets, we find ready consent when it comes to decorative rugs. Holes or dog-earred corners are restored without a compromised scruple, as long as the correlation between the cost of restoration and the restoration of the rug's retail cost is strongly favorable. With antique collectible rugs and carpets, whether for museums or private collections, a decision to restore or not to restore evokes disparate views. Speaking generally, one can say that museum curators hesitate to change the present condition, even if there are ugly repairs, refraining from adding even a single lost knot. On the other hand, antique rug dealers prefer a more complete state of preservation, as does the public in general and many collector clients in particular. Collectible rugs made after 1800 and with moderate damage (like holes, wear creases, cuts, etc.) will be restored without much reluctance; many dealers themselves do this business. Look at The Markarian Album, page 21, and you will see master Richard at work. On the other hand, rug books like Grote-Hasenbalg, Ulrich Schürmann, Eberhart Herrmann, almost never illustrate a heavily damaged rug or even a fragment. All those beautiful late 18th and 19th century rugs in the books certainly could not have been off the loom yet remained in mint condition; but completeness has been a value in itself and so restoration was no sacrilege.
 | Hasan Opcin, director of Woven Legends Restorations in Izmir welcomes Maurizio Battilossi of Italy and Armand Deroyan of France |
Pioneers in collecting fragments are E. Heinrich Kirchheim and Christopher Alexander. While still in the early days of collecting his Orient Stars, Kirchheim acquired excellently preserved (and restored) Caucasian rugs primarily from the early 19th century. When he gradually turned to older rugs, diligent and persuasive dealers praised the beauty of rug fragments and helped him to acquire exceptional examples. Rug world rumor, unsubstantiated by this writer, has it that there was a basement room in Konya full of wrecked rugs and, when a European dealer came every six weeks or so, more fragments were offered (and sold) to their delight; prices soared and, if there was not enough stuff, a knife allowed one piece to become two. I always wondered that this was not only possible but also glorified as an extraordinary kind of art. Although I never have encountered fragments of paintings by Albrecht Dürer or Frans Hals, I must be fair; every dealer has the right to sell what is salable, and every collector to pursue his individual taste. And, to do justice to the looting of mosques: had the fragments not been removed from where they had lain for centuries, we might never have had the opportunity to enjoy their glorious remains in great exhibitions; instead, they would have been ground to powder and corroded to carbon dioxide.
To sum up, one can say that rugs from after the mid-18th century may be restored to an extent that a dealer or collector finds suitable, recreates most of the original impression, is economically reasonable, and makes the owner happy -- provided the craftmanship and the quality of the restoration work is such that restoration cannot be recognized at first and second glance even after years.
As mentioned earlier, I am prejudiced towards restoration of rugs. The essence of a collectible rug, in my opinion, is its appearance, not its keratin-based animal hair. It is color and design. The latter are emanations of the human mind, a coincidence of the skill and the artistic imagination of the individual weaver, stemming from a coherent flow of tradition. In a rug, as in many other works of art, materializes the culture of a tribe, the customs and oral reports within a family, and, last but not least, the personal ingenuity, ability, and artistic sensibility of the individual weaver. If the image of a rug is partly lost by later, unknowledgeable users, I myself feel obliged not only to spend money for acquiring a rug but also to bring back as much and as appropriate as possible of the original appearance. It was not the intention of the weaver to show parts of the foundation to her family or her tribe; those parts which were lost should be repiled.
The patterns of most rugs and carpets are governed by rules of symmetry. These rules enable a skilled restorer to reweave the missing parts of a rug. Reduced ends, holes, and cut-off or partly damaged selvedges at full length and overcasting the remainder with brown wool, or, even worse, adding to a valuable rug whose selvedge is missing the selvedge of a different destroyed rug: these bad treatments can be eradicated.
Unfortunately, almost all buyers of decorative rugs know nothing about the criteria of restoration and are unable to discern whether a carpet is original or heavily repaired; they usually do not ask their dealer about damage or manipulation. When a dealer takes advantage of a buyer's inexperience, his actions ultimately hurt the carpet trade and its reputation. If trust is lost, it is lost once and forever. It would be worthwhile to report on cases of bad repair, painted pile or foundation, fake knots, and other questionable practices, and how to detect them.
In the commerce of antique rugs, it is a must that restoration be executed with the utmost skill in the art, and the purchaser must be informed about the extent of the restoration. Many of us know that this part of a deal is often neglected. When the innocent buyer wishes to sell a rug which was praised as a treasure when he bought it, he finds the experienced buyer -- dealer, auction house representative, or veteran collector -- on his knees, scrutinizing every square inch of the proffered rug, viewing it at closest range, and deftly caressing it with the educated palm of his hand, then repeating the same procedure on the reverse side, before he then shows the baffled owner all the many small repairs which will bring the purchase price down. It is not so disappointing that an antique rug is restored but that it is not skillfully done: easily detectable repairs, made with inappropriate material and colors, and thus eventually the weaver blamed of a botched-up job.
Since the mid-1980s I have had fascinating talks with the Persian rug dealer Parviz Rasta of Hannover, who is an unsurpassed restorer of antique rugs. Discussions on strategies and the results of restorations lasted sometimes until midnight and often took place twice a week. I learned much from his shared knowledge, his techniques, his choice of wool, spinning and plying wool threads, searches for fitting color; unfortunately, I learned much which came too late to save me from some of my earlier acquisitions. Parviz sometimes removes two or three days of repiling work if he finds the color or wool lustre did not meet exactly his high expectations. For my part, I had to learn that estimates of the duration of a specific restoration mostly fall short of reality, and so the cost turned out to be much higher than my financial situation would allow. Looking onto a damaged rug is to see just the superficial, while looking into the rug during the restoration process is a profound learning experience.
Over the years, I have faced the problem, with Parviz and other restorers, where costs and duration of restoration became increasingly unpredictable; this was particularly the case with Parviz's projects as he suffered great pain from a slipped disc, resulting from his years of restoring while sitting on the floor. Expense soared and sometimes six to 12 months passed before I could get my restored rug back.
On a vacation trip through Turkey, I came in touch with two antique rug restorers in Istanbul; Nuri A. and Nebil B. who did great work, and their prices seemed to be more moderate than back home. But there still was a major complication in sending rugs to Istanbul for restoration: Turkish law does not allow the importation of handmade carpets from abroad. For newly made products, this measure is understandable, but it also holds for rugs temporarily imported for restoration. Exemptions are made only if complicated applications are made to Ankara government officials. The application process alone can last up to three months and, in the meantime, the rugs are retained at the Turkish customs storage in Istanbul, adding additional cost and a considerable deposit to be paid in hard currency. When later the restored rug is ready for return shipment, the deposit is refunded but not the same hard currency; rather, it is refunded in Turkish lira, which means, at an annual inflation rate of about 100 percent, that one faces a heavy loss when the repaid deposit is converted from inflated lira into one's domestic hard currency. To get around this, many dealers pack their rugs in suitcases and fly into Istanbul with them since airline tickets from Western European cities to Istanbul are cheap, especially on Turkish airlines. There is some risk in getting through Turkish customs, but the risk seems low compared to the ever rising costs of restoration in Europe.
This was the situation when, at the end of 1992, I discovered Woven Legends Restoration, Inc. through correspondence with Bob Mann, based in Cenver, Colorado, owner George Jevremovic's United States manager. Our ties became stronger after the first trial runs and the 1993 I.C.O.C. in Hamburg, where the restoration facility in Izmir had a stand displaying all kinds of restored rugs. Fortunately, it was not necessary for me to send my rugs to Denver; instead, they could be sent directly to the customs-free zone at Gaziemir-Izmir, Turkey. The manager of the Izmir operation, Hasan Opcin, himself an excellent restorer of antique rugs, has accumulated such so much experience with the restoration of thousands of rugs and carpets from all countries of origin that no question remained unanswered. No restoration problem is tackled without thorough investigation of the object, and good advice is provided when an owner feels uncertain or unsure about the best solution.
In April 1995, five Europeans with previous experience with the Izmir facility met in the Customs Free Zone close to the airport. Greeting them were George Jevremovic, who flew in from Eastern Anatolia where the Black Mountain Looms carpets are being woven. Also on hand were Bob Mann from Denver, who coordinates all restoration activities in the United States, and Hasan Opcin, director of the Izmir facility, who played host to the party. The other visitors from Europe were Maurizio Battilossi of Torino, Italy; Besim Jr., representing Austria and Switzerland; Armand Deroyan, France, and David Bamford of the U.K.
 | Silent working at Woven Legends Restoration, Izmir |
The first floor of the hall has been extensively fireproofed. A guard is posted in the hall 24 hours a day, even on weekends, to guard against theft or burglary, while on the ground floor iron doors and an alarm system are the first line of security for the facility. On the ground floor, a small office is shared by the director of the operation, Hasan Opcin, Ms. Didar Arikan, business manager, and Akin Altuntas, bookkeeper. A large, locked, storage room houses raw machine-spun and handspun wool yarns from many sources of the Middle East, Asia and commercial sources in Europe and the U.S.A.
The in-house wash plant can accommodate carpets as large as 15x24 feet (4.5m x 7.2m). Washing is conducted with great care, with special attention to the fastness of dyes and the stability of structure. The washing agent is a neutral synthetic detergent which is later removed by throrough rinsing of the carpet. The carpets are hung on a bar to dry at moderate, controlled temperature. Besides the craftsmanship of its weavers, one of the most important aspects of restoration at Izmir has to do with dyes: the dyeing process itself and the selection of the dyed wool. All shades necessary for the work are made in the dye shop, using highly lightfast and washproof dyestuffs from Ciba-Geigy Co. Though vegetal dyed wool is also used, it comprises a minority among the 4000 (+ or -) shades, which are stored in shelfboxes in the restoration hall.
To the sophisticated customer, it might seem strange that a perfectly performed restoration is made using "synthetic" dyes. Critics may think of Orange IV or Ponceau 2R, dyestuffs they hate in rugs from the last quarter of the 19th century. I must confess that I, too, had an uneasy feeling when I learned that almost all restorers use synthetically dyed yarn (except my friend Parviz Rasta who had accumulated his heaps of wool from damaged old mafrash or kilims). The modern dyestuffs, however, are soluble (chromium) pre-metallized chemical compounds which fix to the wool fiber as fast as do mordant dyes from natural sources. The only difference in the dyeing process is that synthetic dyestuffs for wool are metallized in a vessel in a chemical plant to form a metal complex, while with vegetal dyes the insoluble metal complex is built up on the surface of the mordanted (metal salt treated) wool yarn in a kettle. The different shades of synthetic dyes stem from different constitutions of the basic, non-metallized dye molecule. The various shades of vegetal dyed wool are created by an uncountable mixture of unpure alum with traces of accompanying metal salts from copper, iron, tin, etc. and a variety of secret pre-treatment recipes for the mordant process and a mix of varying proportions of the basic dye and its derivatives. The chromophor group (color-giving part of the dye molecule) in synthetic dyes for wool is predominantly the well known Azo-group (-N=N-) and the so-called anthrachinone three-ring molecule for a wide range of reds, while for vegetal dyes azo compounds, to my knowledge, never occur.
Vegetal dyes are not chemically pure substances; they are, from plant to plant and from area to area, mixtures of chemically related derivates which may change their composition during extraction from roots, stems or leaves and during the dyeing process. This makes it extremely difficult to produce large quantities of a specific shade for restoration. It causes no problem if vegetal dyed wool is used for the production of modern handmade rugs, such as the Azeri or Black Mountain Looms carpets. Little differences in shades of wool create a welcome abrash, but the use of vegetal dyed wool would be prohibitive for restoration. For an experienced dye master, it is easily feasible to reproduce all hues of colors from an antique rug by synthetic dyes.
Once satisfied that the quality of restoration will be excellent, the customer's next concern, and that of Woven Legends Restoration as well, is an estimate of the cost of repair. This process involves determining the number of work days needed prior to the beginning of restoration and a survey is made either in the home country or at the Izmir facility. Representatives in the United States, in Europe, and in Izmir are trained to estimate in advance what time might be needed to bring the rug or carpet to the condition the customer has demanded. If the number of work days is fixed, it is the basis for the later invoice. The number of work days for which a customer is billed is not changed from the estimate, even if during restoration a previously unforeseen problem is discovered, for instance, brittle wefts, and the projected duration of the work cannot be kept. Therefore, a vital prerequisite is a written work order on which the customer states specifically what he wishes to have done: will the rug be washed and/or stretched; which kinds of damage will be repaired and which will be left untreated; will the missing end border be rewoven or only the worn places and the holes closed, etc.
I tried to estimate rugs and carpets for two days during my stay; the rugs having already been estimated by the company. With 60 restorers at work, rugs from all countries of origin and age were available. From the many examples, I chose to estimate only two. And now let's have a little fun.
 | The top end of the Senneh showing the need for repair to the end border |
Our second illustrated example is a detail of a Yürük rug, 264x124 cm. This rug is heavily damaged and you can see where big holes have been filled with a newly made foundation; some extensive areas of repiling need to be done. How many work days would you guess this part needs?
 | The bottom end of the Senneh showing which has been repaired to its full width |
To estimate cost, my technique was to calculate the duration of a restoration via unit operations i.e. so and so many hours/days to insert 100 warps of thick or thin wool into a foundation of zero, 60 or 90 degrees depressed warps. Or repiling 100 knots into a dense or coarse foundation. Or adding a selvedge weft line by weft line per 10 cm etc. I figured that knowledge of these unit operation durations could be compiled during a year of work in Izmir. The result would give a customer the basis for a rough calculation of approximately what the restoration of a damaged rug in question would cost and thus facilitate the decision of what to restore and what not to restore.
The specialists in Izmir believe that my approach is a theoretical one which will never work in practice, even though they themselves analyze and estimate the different damages individually in a single rug. For an exact estimation of work days as the unchangeable basis of the invoice, only the direct investigation and feeling of a piece allows a reliable quotation to the customer.
And now, are you ready for the answers?
The Senneh rug would requires four days to insert the warps for each end, and 20 days each for repiling 28 rows, and for color match and clipping, an additional six days; in total 54 work days.
 | This detail of a Yürük rug shows where a new foundation has already closed holes; repiling is yet to be done. |
So, how did you do on our quiz? Did you come close, or have you had to rethink the rug repair equation?
