Restoration and Conservation of a Karen Bronze drum

Restoration and Conservation of a Karen Bronze drum

Bookmark

By Hall Conservation

Historical and cultural background

To the Karen people of Burma (now Myanmar), their bronze drums were of great value, but in a much broader sense than we in the modern west define value. The drums cost a lot to buy and they confirm the status of the owner, but whereas that is all that conspicuous consumption does for the modern westerner, to the Karen, a drum was a safeguard against ill-fortune, it gave the owner power over others, it gave power over nature, it summoned up ancestor spirits when they were needed, it kept the nasty spirits happy, it could foretell the future, it could make warriors more courageous, it made ceremonies more effective, it was an insurance policy. Such was the power of the drum that enemies would make pre-emptive strikes to steal a drum in case it was going to be used against them, thereby prolonging and intensifying a bitter inter-village or clan feud.

Karen origins are obscure, and even the name given to them is a colonial one imposed collectively to a wide range of subgroups. Possibly they migrated into Burma around A.D. 600-700 from China. They settled high in the Burmese hills, traditionally lived in longhouses and believed in capricious spirits that dwell in sentient and inanimate objects around them. At what point the Karen began to acquire and regard the frog drum as their most precious possession is, like much of their history, not known with any certainty, but it is a long tradition of at least 1,000 years and is integral to their animist beliefs because the sound of the drum pleases the spirits. In ancient times it was the custom that on the death of a chief his best wives, elephants, horses and most valuable possessions were buried with him, but during the 16c this practice was forbidden, although it did continue in token form. The drums sometimes had a part broken or cut off to be buried with the body, or miniature versions were substituted. Surviving drums were devalued because objects associated with the dead became inauspicious. However, paradoxically, old drums were highly prized and were deemed to develop a sweeter sound with age, so there were exceptions, perhaps the drums were passed on before each successive owner died, either as gifts or sold. Although the perceived benefits of ownership of a drum were much more than mere wealth, they were also regarded as a form of financial ‘savings’ and if the owner fell on hard times, a drum might be sold. If the owner had fallen on hard times, selling it might have been justified on the grounds that it had stopped working. Drums were classed as auspicious or inauspicious depending on a number of factors, the sweetness or harshness of it’s tone, whether the sound carried well and whether it conferred all the blessings that were expected of it or not. Sometimes inauspicious drums might be deliberately broken, but even an inauspicious drum would still have a market value, after all, one wealthy man’s inauspicious drum would be better than no drum at all to a man socially down the scale but on the way up. Not all the drums were owned by an individual, many were owned by a family or the extended family of the clan, so a number of individuals would benefit where there was collective ownership. Individual ownership though conferred the greatest prestige, a man with one drum had higher status than a man with seven elephants.

Bronze drum production reached it’s apogee during the 19c when teak logging was at it’s peak in Karen territory and the value of the timber was high. The Karen themselves did not make the drums because they did not acquire the technology, the actual makers were another tribe, the Shan (also referred to as ’Red Karen’). The Shan, despite their expertise in making the instruments did not themselves use them in their own rituals and ceremonies. The centre of production was the Shan quarter in the town of Nwe Daung, a Karen town where the Shan artisans were made to settle by the Karen in what must have been a comfortable ghetto with a surrounding ditch and hedge. The Shan metalworkers (they were also skilled blacksmiths, silversmiths and jewellers) were in effect compliant prisoners of the Karen, it is not clear whether the ditch and hedge was intended to keep them in or protect them against others, whichever, the two communities mingled freely and there was free access in and out. There is also an implied respect by the Karen for the Shan because despite the nature of the relationship, the Shan were allowed to keep secret their knowledge of the crucial element in drum production, lost wax casting. If the relationship had begun on a basis of dominance by the Karen over the Shan, it metamorphosed into a symbiotic one where each benefited from the other. The Shan had a captive clientele in the Karen, and the Karen had dependent suppliers in the Shan.

One thousand years of tradition ended abruptly in the early 20c. By the latter part of the 19c the Karen State was in political turmoil and was tearing itself apart by internecine warfare when they embarked on a disastrous war with Britain in 1888 to 1889. During the conflict, Nwe Daung was destroyed and the Shan workshops razed to the ground along with the rest of the town. The craftworkers were dispersed and neither Nwe Duang or anywhere else re-established itself as a centre for the production of the drums. The disruption to the lives, religious practices, economic and trading patterns of the Karen was so extreme that former reverence for the drums rapidly withered away to the extent that by the 1920’s the use of the drums had died out.

Manufacture

Each master bronze caster had their own yard or compound in which the whole process of manufacture of the drums was carried out by a technique that has some similarities to bell casting in the West. Building the mould started from the inside and gradually worked outwards; first a substantial hollow clay core was built up from clay mixed with rice husks, which reinforced the clay and at the same time opened up the body to stop it cracking while it is moist and drying out. When the basic shape of the core was satisfactory it was coated in a layer of slip, a creamy slurry of clay mixed with water and, to give it additional body, sieved cow dung. It could then be trued up by turning on a lathe. When the form and surface was satisfactory, a rice water glue was painted on to it to consolidate the surface and prepare it for the next stage.

Wax was then applied to the core’s consolidated surface in an even thickness and the decoration applied to that layer, this formed the wax positive necessary to make casting by the ‘lost wax’ process. The decoration was made by stamps or dies, engraved rollers, spatulas and scrapers, the raised decoration, the frogs, other animals, raised beads and handles by direct modelling. The vertical strips that look like joints are purely decorative, but do refer to an early form of the drum that was made in sections and had to be joined. When the decoration was completed, sprues, which at that stage were rods of solid wax were added. The sprues became the drain holes for the wax as it melted out, then inlets for the molten bronze when it was poured in and outlets for the air as it was driven out by the molten bronze. With the sprues attached, the first layer of outer mould case could be applied. This was another layer of fine slip, backed initially by more fine slip with rice husks added and then the main body of the mould made from clay mixed with horse dung, rice husks or rice straw to bind it together. Mould making was completed by pushing chaplets, iron rods or nails through the outer case, through the wax and into the core. Chaplets prevented the core and outer case slipping out of place when the wax had melted out.

With the mould complete, it would be allowed to slowly dry, then, with it upside down so that the wax could drain out, a kiln was built around it and fired, gently at first and then progressively building up the heat. During the heating, as the wax melted out, the rice husks and straw burnt away which left the mould porous enough so that all the mechanically and chemically combined water could be driven off. Heating the mould is a critical stage in lost wax casting as trapped wax or water can blow back the molten bronze at 1,500º C or rupture the mould. An overheated or underheated mould adversely affects the quality of the cast.

While the mould was being brought up to temperature the bronze would be prepared. Ingots of copper, tin and scrap were placed in a crucible in the furnace. The alloy is typically copper with around 16% tin plus a little lead and zinc, but as each batch was prepared individually for each pour and as each master would have his own preference, no two are quite the same. The high proportion of tin, 16%, (art bronze is 90% copper to 10% tin) provides the sweetness and increases the sustain in an instrument. Tests were carried out as the melt progressed and the proportions of the metal ingredients adjusted. Co-ordination was the key to success at this stage. When the mould was ready, the alloy had to have the right proportion of it’s constituents and be up to temperature, the mould kiln was opened, the mould was turned the right way up, part buried in a pit filled with sand or fine sieved earth, dross floating on the bronze was skimmed off and the metal poured into the mould. The mould was then completely buried so that cooling was slowed down over several days.

When the cooled mould and it’s cast were excavated, the mould was broken open and the casting checked for flaws. There was always a possibility that the mould might have broken internally and the flow of metal had been obstructed. If the wax coating was too thin it might prevent the metal flowing to every part. If the metal or the mould were not hot enough, that would cause the metal to run sluggishly or freeze prematurely. Assessing the temperature, like mixing the alloy, was down to the eye and experience of the master.

A good casting at this stage was too thick and rigid to allow the instrument to vibrate sufficiently so that the essential harmonics could develop, it is simply not technically possible to cast the bronze thin enough. To reduce the thickness and tune the body and tympanum, the casting was mounted on a lathe and the inner surfaces were turned using hand held chisels - the ‘chatter’ marks are clearly visible. The walls of the body are turned down to less than 1 mm in places. Simply holding the chisel steady enough to remove metal in a controlled way within such an awkward form is impressive, but judging the thickness to such a degree of fineness by eye demonstrates consummate skill.

The sound

The drums were hung just above ground level by one handle from a convenient support. Steadying the drum with his toes on the other handle, the drummer had both hands free to play it. The tympanum was struck with a soft headed beater which, in the centre, produced a mellow bell-like tone which hardened as the tympanum was struck closer to the edge. The body of the drum was also used to produce another sound, described as being something like a cymbal. The body had it’s own beater, a bundle of fine, straight twigs or split bamboo, like a modern drummers brushes. It would be possible for the drummer to control the way the vibrations fluctuate through the drum by the way it was beaten to produce harmonics that would float under and around the basic notes. The sounds that could be got from one highly regarded drum were described by a Karen elder as; gu-mu gu-mu, hoo-hoo hoo-hoo, nge-wo nge-wo, reh reh reh reh, ro ro ro ro, waw waw waw waw, pur-r-r-r. Exactly how that translates into the actual sounds made the drum is beyond our experience, but it does indicate a range of quality of sounds with a various rates of attack, sustain and decay.

The damaged drum

This was the first of three Karen drums that Hall Conservation has worked on, but by far, the most badly damaged, the tympanum was in good condition, but the body of the drum was severely damaged. It was badly distorted and there are a number of horizontal splits that follow the decorative lines at their thinnest cross section, a roughly Y shaped tear and a series of fractures around an approximately palm-sized hole. A short section of the plain band at the lower edge of the body of the drum was detached and a crude repair had been attempted by stitching the part back with wire through drilled holes. This form of repair had been carried out across other splits and at the Y shaped fracture, some sort of resin had been applied also.

In addition to the obvious damage there are a small number of small holes in the body of the drum, these are minor flaws in the casting.

If we now turn our attention to the damaged drum which is the subject of this web publication and consider how it came to be so severely damaged, the intuitive response is to assume that it was accidental, after all, who would deliberately inflict so much damage and why? When an object is accidentally damaged, the pressure points are usually from consistent directions and possible to interpret, for example, it falls or something falls on it, but the damage here seems to be inflicted from several different directions. We are told that for various reasons, the Karen themselves deliberately damaged the drums, but Western accounts of deliberate damage focus on puncturing of the tympanum. This would destroy the tensioning of a skin drum (to which Westerners are more accustomed), but would not be as catastrophic to a rigid tympanum (cf the use of a guitar body, a hard box with a hole in it, for percussive effect in Flamenco music). The acoustical study of a Karen drum carried out by Laura M. Nickerson and Thomas D. Rossing of the Physics Department, Northern Illinois University demonstrates convincingly that whole drum is essential to the sound it is intended to produce, as the harmonics are set up throughout the body. It is therefore arguable that the most effective way to render the drum useless would be to wreck the body, which also happens to be the weakest part of the instrument. This is not an exhaustively researched explanation of the damage, but plausible and knowledge of Karen culture is far from complete.

Acknowledgements

The above is the distilled from; The Karen Bronze Drums of Burma, Richard M. Cooler - Studies in Asian Art and Archeology; Frog Drums and their importance in Karen Culture, Sylvia Frazer-Lu - Arts of Asia, September/October 1983, L’Asie Exotique, and Acoustics of Karen Bronze Drums

Laura M. Nickerson and Thomas D. Rossing

The ‘Arts of Asia’ publication is has a very detailed account of the Karen and the making of the drums with very good archive drawings and photographs. There is also a very good description of the Karen at the Northern Illinois University Centre for Burma Studies

Repairs

The high tin content that gives the instrument a sweeter ring and helps produce the lovely rich, dark patina also makes the metal more brittle and susceptible to the kind of fracturing found here. The reason for this is that the tin does not completely amalgamate with the copper but forms tiny beads within a matrix of copper. It is rather like a sponge, soaked in water and then frozen, if the sponge is bent, the ice will not deform but the sponge will stretch and tear around it, which is comparable with what has happened to the alloy of the drum.

When copper is deformed it rapidly becomes stressed as the grains break into smaller sized grains. This makes it stiffer because the reduced sized grains cannot slide over each other so easily and it is more liable to fracture, so copper will bend in one direction, but be more resistant when reversing the bend and more prone to breakage. A technique to overcome this is annealing, the metal is heated to dull red and allowed to slowly cool. The process recombines the grains of the copper and it becomes pliable again. This also works with the lower tin content bronzes, but in the case of high tin content bronzes, the tin melts out as ‘tin sweat’ when the alloy is heated (melting point of tin around 230º C, annealing temperature of copper around double that), so an attempt to anneal the material to make it more workable would cause damage. This presents a dilemma, annealing will cause damage and not annealing means that working the metal will cause damage by either causing creating fractures, microfractures or additional stress.

Before deciding on whether to anneal or not, a test heating was carried out on a small detached sample to determine which course of action would be the least damaging. The chosen compromise was to attempt reshaping without annealing because although there was a high risk of further breakages it was considered that this would cause less changes overall to the integrity of the drum.

The two images below show a small section before heating and beads of ‘tin sweat’ before annealing temperature was reached.

In practice, by moving the distorted sections progressively a little at a time appeared to reduce stressing to adjacent sections, so that additional visible fracturing was minimal. The two images above show the extent of the missing sections and splits when the deformation was realigned. Reshaping can never be completely accurate because the mechanics of damage cause stretching and cannot be pulled back, but with reshaping at a satisfactory stage, the fractures could be locked in place. This was done with patches of copper sheet as bridges across the fractures and gaps. Our preferred technique for fixing the copper sheet to a material of this type is soft soldering, but it proved impractical due to a problem come across previously on a number of occasions - test areas were stripped back to bright, apparently clean metal, but the solder would not ‘tin’ adequately across the surface. Probably, because the metal is porous, grease and dirt have been drawn into the porosity, only to be driven out on heating, thereby preventing the solder taking evenly across the surface. Instead of solder the reinforcing plates were fixed by epoxy putty. Dirt and grease are barriers to epoxy resins, but because heating is not part of the process it will remain within the porosity rather than flood out across the surface, leaving a greater clean surface area for adhesion. For added reinforcement, slightly work-hardened copper tubing was used to stiffen the most damaged sections, which are most vulnerable to damage in future.

Where the plain copper patches were visible, the surface decoration was reproduced by lifting moulds from complete areas and then pressing them into a thin skim of polyester resin filler over the copper sheet. This is similar to the method the Karen used themselves to emboss the repeat decoration into the wax positives.

The entire drum was washed in conservation grade detergent with 10% IMF in water, some small blue paint splashes on the tympanum were lifted off with acetone. The resins that were visible were touched in using artist’s acrylic paint and the whole drum was given two coats of Renaissance microcrystalline wax.

Hall Conservation Ltd is the hub of an international network of highly skilled and knowledgeable conservators and provide the highest quality professional conservation and restoration service. The company is capable of successfully completing a wide range of conservation projects. From a pool of specialists Hall Conservation can bring together a conservation team suited to the requirements of any specific project; incorporating all fine arts, decorative arts and architectural conservation work on any scale, anywhere within Europe or beyond.