Asian Arts | Articles

Separation Anxiety:
The Conservation of a 5th century Buddhist
Gandharan Manuscript


February 21, 2000

(click on small images to go to full page images with captions)

"Then, in the midst of the gods of the heaven of the thirty-three, a son of a god was dwelling in the Sudharma, the palace of the gods. In a huge heavenly mansion, surrounded by great divine opulence and great groups of divine young women, he played with them and made love. After he had enjoyed this divine opulence, during the night, he heard a voice:

"The son of a god will die on the seventh day. When he has died, he will be reborn again in the Indian continent, and there too he will expense seven states of rebirth. After experiencing seven states of rebirth, he will be reborn in Hell.
If even once in a hundred times he is reborn as a man, he will be poor and blind -- by him this was heard."

So begins the translation of one of the 5th century Buddhist sutras found within a recently discovered early manuscript.

The Project

Fig. 1

In the fall of 1998, a professional numismatist, with a specialty in Classical antiquity, brought an early manuscript to my studio for consultation. (Fig.1) The elongated leaves were brittle, compressed together, water damaged, and folded into a tight "S" curve, like a wad of dollar bills after the wash cycle. This manuscript was found in the Bhamiyan cave region in modern Afghanistan, purportedly from the 5th century, on birch bark, and written in the Kharoshti script.

The owner's hope was that the manuscript leaves could be opened and separated, in order to be translated. He explained to me that this manuscript was extremely rare, and that the owner hoped to discover any missing links to the history of early Buddhism.

Ethical Considerations
But before agreeing to take this project on, I had several philosophical and ethical questions. I am familiar with many types of Asian books and manuscripts; and have considerable experience examining and treating early Buddhist sutras on paper & Indian palm-leaf manuscripts. (Fig. 2) These Indian and Himalayan manuscript types provoke inquiry and ethical considerations, especially issues regarding the treatment of sacred objects. As with most ethical questions, there are a wide variety of answers, depending on the opinions of religious and museum specialists.

Fig. 2

When Buddhist sutras are found inside sculptures, conservators and curators often disagree on the appropriateness of removing these rolled-up documents from their consecrated compartments and treating them individually as works on paper. Additionally, some published accounts insist on the importance of having a religious official re-consecrate the sutras prior to insertion back into the sculptural cavity.

Indian palm-leaf manuscripts present fewer ethical problems for the conservator, as they are often collected individually for their illustrated images, and many beautiful examples exist in museums, libraries, and private collections. Palm leaf manuscripts were probably in use as early as the 2nd century, but no extant leaves survive earlier than the 10th century. Because palm-leaf is still used today in India for certain religious writings, much is known about the manufacture and treatment of the material.

Birch Bark

Fig 3

But what about birch bark? Prior to the day last fall when I first examined this manuscript, I was completely unfamiliar with the material as a support for writing. Admitting this to the client in our initial meeting, I took several slides, and was given a small bag of manuscript fragments (Fig. 3) in order to begin research and testing on an appropriate treatment methodology. To be honest, I didn't have very high hopes, and, as I talked to colleagues about the project -- I couldn't find anyone else in conservation familiar with this type of early manuscript. I was anxious about treating this object for many reasons: the object was purportedly from the 5th century, it might be a missing link to the history of Buddhism, and it was also extremely valuable. So, research and testing had to resolve whether this treatment would provide the desired results.

Initial research 
Before commencing a literature survey, I consulted several colleagues who had great experience with manuscripts, early non-paper collections, and Indian art in general. Colleagues familiar with papyrus were particularly helpful, and other colleagues tried to convince me that the manuscript must be on palm-leaf, which was my initial error as well. A particularly helpful consultation was with Nancy Turner, Manuscript Conservator at the J Paul Getty Museum. I brought my slides and my fragments up to her lab, and we examined them under the microscope, and they were, in fact, bark. We could clearly see the laminated layers of very thin barks and examine the inks and written inscriptions very clearly.

The literature survey was sparse but useful, with a great deal of historical data, but little information on conservation treatment.

History of birch bark as a material for writing
(Fig. 4) The research in Butterworth's Conservation of Manuscripts and Paintings of South-East Asia, revealed that birch-bark (called bhoja-patra) was a primary writing material along with palm-leaf in India before paper. Birch bark was mentioned as a writing material by the Greek historian Q. Curtius, noting its wide use during Alexander's invasion by Hindus. Early extant manuscripts date to the 2nd and 3rd centuries, written in the Kharoshti script. Fragments survive from a range of time periods, and the material is described throughout Indian literature. Bhoja-Patra's use diminished in the Mughal period when paper replaced it as a writing material, but it still has a sacred status in India today.

Fig 4

Despite its status as a sacred material, all of the literature focused on bhoja-patra's use as a support for writing. Buddhist sutras were written by monks, and often sponsored by donors, who might be high lamas or officials. The manuscripts are not illustrated and have no history of association with consecrated sculptures or images. Bhoja-patras were literally the Ancient database of Buddhism in India.

The manuscript brought to my studio, now called the "Los Angeles manuscript" was found in the Bamiyan cave region of modern Afghanistan. In ancient times, this area was part of Gandhara, the region invaded by Alexander the Great in 326 BCE. Gandhara became a second Holy Land of Buddhism, and most extant sculpture from the region took the form of Buddhist cult objects, Buddhas and Boddhisattvas, (Fig. 5) like this Boddhisattva sculpture from the 3rd century in the Norton Simon Museum, or architectural ornament of Buddhist monasteries. Because of the strong influence of the Greeks and Romans, Gandaran sculpture reveals many Western classical elements, including the treatment of the robe and its heavy folds, and the physiognomy of the central figure. Interaction through caravan trade routes and the Silk Route in particular maintained these stylistic exchanges.

Fig 5

Fig 6

Bamiyan Caves
One of the greatest sites in Gandhara is Bamiyan, a mountain valley in north-central Afghanistan. (Fig. 6) A high cliff forms one side of the valley and is honeycombed with monastic dwellings, bounded on either side by colossal Buddhas in niches cut out of the rock. The largest, depicted here, is 174 feet high, and in the Gandharan style. (Fig. 7) The region is full of monasteries, stupas, and caves, some with rich interior depositories of paintings, murals, and manuscripts.

Fig. 7 Fig. 8

In the studio
The literature states that the inner bark of the birch tree was used for writing. (Fig. 8) After being peeled off the tree, the bark was dried. Oil was then applied over it and it was polished. Layers were joined together by a natural gum. Finally it was cut to a suitable size and kept in between wooden covers. The ink used for writing on birch bark was "Indian black, a carbon ink. It was prepared by burning almond shells to charcoal, which was then boiled with cow's urine. This ink is said to have a special brilliance and is fast to washing." Tests have shown birch bark sheets to be typically 0.2 - 0.5 mm thick, and contain a cellulose content of 38%. Additionally, birch bark is highly soluble in organic solvents, but not soluble in cold water.

Fig 9

Initial research complete, I began to test the fragments in the studio. (Fig. 9) The inks proved stable, as predicted in the literature, and the birch bark itself became supple in high humidity and contact with water, without staining. Bolstered by these results, I agreed to take on the project.

Once we had the manuscript in the studio, we began detailed written and photographic documentation prior to treatment. (Fig. 10) I had decided to use humidity to attempt to open, separate, and flatten the leaves. But before I describe the steps employed, I would like to read the only historical account I could find in the literature for separation. 

It reads: 

"In the 1930's, the Musee Guimet in Paris had acquired bundles of birch bark found at Bamiyan in Afghanistan. Any attempt to open the sheets was resulting in the breaking of the sheets into small bits. Water vapor had no effect on separating them. Finally, hot paraffin oil was used to soften and separate the stuck sheets. The reason for choosing paraffin was its clarity and preservative quality. The fragments were immersed in cold oil which was then heated on a slow fire for some time until a light smoke started coming from the oil. In this condition it was possible to detach the leaves from one another with the help of a pair of tweezers. The mud split up easily and each piece was cleaned, drained and laid on a sheet of glass. Each fragment, along with the oil, was sealed in glass. The edges of the glass sheets were sealed with paraffin wax."

They go on to point out the primary disadvantage of this treatment is their weight and bulky storage requirements, but clearly there are other disadvantages!

Fig. 10 Fig. 11

The Conservation Treatment
Since the inks were stable, and the bark impervious to water staining, I began humidifying the manuscript using an ultrasonic humidification chamber made from a photo tray and plastic sheeting and maintaining 80% relative humidity (RH). (Fig. 11) While watching carefully for excess precipitation, the manuscript underwent humidification in this fashion for several days. To my great satisfaction, the leaves began to relax, but were not wet. I continued to humidify the manuscript in this fashion for another 72 or more hours, which allowed me to begin carefully manipulating the "block".

Fig 12

Fig 13

As you can see, I was able to begin to open the entire manuscript from its tight "S" configuration to a looser shape. (Fig. 12) Maintaining a minimum of 70 - 80% RH, the entire block not only began to open, but the outer leaves began to separate as well. (Fig 13) Soon I was able to begin separating the outer most leaves with the mechanical aid of a Teflon coated microspatula and removing them completely from the block. Despite all of this humidification, the leaves were still extremely brittle and fragile, and great care had to be employed. Also -- most leaves were split down the center, and had many other fractures and loose attachments. At many times in this project, more than one pair of hands and eyes were needed, both to handle the delicate leaves, and double check accurate collation of a manuscript in a non-Roman alphabet.

Once the first group of 6 outer leaves were removed, they were placed on a damp blotter and Gore-Tex layer, and gently held in place until the top of the "blotter sandwich" could be placed on top. Needless to say, this first group's successful separation was extremely exciting around the studio, and I called the owner with the initial good news.

We continued the treatment, and subsequent leaves also separated in this fashion, followed by further humidification once flat, followed by flattening between dry cotton waterleaf blotters. In the end, the leaves opened up to approximately 2 inches high by 14 inches wide, from the original 2 x 3 inch folded object.


Fig 14

Since the goal of this project was translation of the manuscript(s), I recommended Mylar encapsulation without repair of the individual leaves. This way, the leaves would remain flat and could be handled, and there would not be any repair tissues obscuring any text. We designed a 14 x 18" Mylar package, dividing each sheet into four rows to accommodate the manuscript leaves. Each package contained two Mylar sheets and horizontal rows of double stick tape. (Fig 14) The bottom Mylar sheet was clamped to the work table, and, working together, we would bring over each manuscript leaf one at a time, in order; lift the tape paper, and seal down the "row" to lock in the leaf. I was so afraid of losing the order or flipping the leaves, since I don't read Kharoshti or Sanskrit, that we were extremely careful about collation.

Fig 15

In the end, the treatment was a great success. (Fig. 15) All of the leaves separated, and remained flat in their encapsulated packages. There were fragments and areas of birch-bark leaf loss, just like any ancient material like papyrus, but all the scraps and fragments were saved and encapsulated in the precise order of the original folded object.

After the final meeting with the very pleased owner, and the work was released from the studio, the manuscript was taken away for translation. After being photographed, transparencies were sent to Richard Salomon, the author of Ancient Buddhist Scrolls from Gandhara, a publication of the British Library. This book came out in June 1999, six months after my research and treatment. In the book, Mr. Saloman states: 

"As the Dead Sea Scrolls have changed our understanding of Judaism and early Christianity, so a set of 29 scroll fragments acquired in 1994 by the British Library promise to improve our  knowledge of the history of Buddhism".

This surprising discovery in 1994 revealed over two dozen texts which had not been identified with previously known texts in other Buddhist languages and traditions.

After Mr. Saloman's initial consultation, it was sent to Dr. Gregory Schopen at UCLA, who translated the manuscript. This was the result: the 40-odd leaves or fragments were two complete books, or texts, and the technical name for the script is "upright Calligraphic Gupta". The larger text was a previously known Buddhist sutra and the equivalent of 37-40 leaves. The transliteration and translation were sent to me along with the information that the "smaller of the two books, which Dr. Schopen has dubbed the "Los Angeles Manuscript", consists of a seven page incantation apparently unknown in contemporary manuscripts and thus of significantly more interest to scholars." This manuscript is now in the process of being published by Dr. Schopen, and is every bit the "missing link" to history as the owner had hoped.

I would like to thank the many colleagues and associates who assisted me in my research and treatment, and in the preparation of this paper. In particular, thanks to aNancy Turner, Manuscript Conservator at the J. Paul Getty Museum, for her consultation, and my studio associate Micol Hebron. In addition, I would like to thank Dr. Page duBois, Professor of Classical Studies and Comparative Literature at UC San Diego, Indian art specialists Christine Knoke at the Norton Simon Museum and Dr. Stephen Little at the Art Institute of Chicago, and the LACMA Conservation Center for use of their library and resources. 

Susan Sayre Batton
Art Conservation & Consulting Services
Los Angeles, California

Agrawal, O.P., Conservation of Manuscripts of Southeast Asia. 

Agrawal,. O.P., "Investigations for preservation of birch-bark manuscripts", 
Preprints of ICOM Committee for Conservation, Vith Triennial Meeting, Ottawa (1981) Paper.

Filliozat, Jean, "Manuscripts on birch-bark (Bhurja patra) and their 
preservation", The Indian Archives, 1 (1947) 102-8

Harle, J.C., The Art and Architecture of the Indian Subcontinent. 
Penguin Books.

Asian Arts | Articles