Conservation considerations in the restoration of archaeological ceramics

Charles O. Diesen, Grant Goltz, and Paul S. Storch


This paper will explain and illustrate the post-excavation history of a prehistoric ceramic vessel that was recovered from a site in Northern Minnesota several years ago. We believe that the post-excavation and reconstruction history of this object is typical of what happens to many important archaeological objects. It points out the need for proper conservation training with archaeology and nature of ceramic reconstruction as a sub-specialty within conservation. One of the authors (Storch) made this point in a paper presented at the 1993 AIC Objects Specialty Group session on training in archaeological conservation. Other workers in the specialty of archaeological conservation also have discussed this in the literature (S. Koob, Conservation Notes, No. 21, 1996).

The object, known as the “Hannaford vessel,” was partially reconstructed in the field by the excavator (Goltz) using materials that had been recommended by the MHS Objects Conservation Lab via the MHS Archaeology Curator (Diesen); then disassembled and fully reconstructed in a lab by another, less experienced archaeologist who was not involved with the excavation. The second “reconstruction” was done without access to the original field notes and other documentation, which provided important and essential information, both for the reconstruction and interpretation of the site data.

The field mending and cross-fitting was done with Acryloid B-72 (Rohm and Haas) 1: 1 in acetone. When the vessel went to the other lab for further “processing,” it was photographed under lights that were apparently hotter than 45 degrees C, which is above the Tg of B-72. The adhesive softened, giving the archaeologist the impression that it was inadequate for reconstruction. At that point the vessel was disassembled and reconstructed with a proprietary cellulose nitrate-based adhesive, such as “Duco Cement” (DOW Chemical), which has a Tg of 54 degrees C. Also used was “Crayola Modeling Compound” (Binney and Smith, Inc.) to fill the lacunae, much as plaster has been traditionally used in ceramic reconstructions. The modeling compound is supplied as a dry powder to which water is added. The compound is kneaded until solid, then molded and pressed into the voids. The person doing the filling overlapped the compound onto the original vessel surfaces in many places; impressed their fingerprints into the fills; left dark smudge marks; and did not finish the fill surfaces in any way. The main problem with using this material as a filler in this way, is that it never completely hardens. It was developed by the manufacturer as a children’s non-toxic modeling alternative to clay; it was never meant for professional use in this application.

The final reconstruction was photographed standing upside-down on its rim for the illustration that was used on the cover of the final site report. The standard method of storage and display of cortical shaped prehistoric ceramic vessels that has been developed at the MHS is to create soldered wire supports that allow for the vessel to he supported in the correct use position.

The vessel was finally returned to the Objects Conservation Lab at the MHS for examination prior to being placed in the permanent collections. It was found that the lower quality adhesive had been used, and that the overall shape of the vessel was incorrect. The fill material was found to be distorted and soft. it was actually tearing away from the sherds in several areas. It was decided to reverse the adhesive and fill material and correct the shape. The cellulose nitrate was removed with acetone, and the fill was reversed with warm deionized water. Most of the “ghosting” caused by the compound was removed. The sherds were documented and matched with the original site note drawings and reconstructed with B-72 1:1 in acetone. The main sections were then joined with adhesive where possible. It was decided not to use a fill material in the lacunae. A interior form was reconstructed from polyurethane foam to serve as the support for the vessel. The various sections were oriented on the form to reconstruct the correct vessel shape, and were held to the form with stainless steel straight pins around their margins.

The method of “reconstructing” the entire vessel in this case was by digital imaging. Once the vessel was held onto the form, it was photographed in all positions. The images were then scanned into a PC, and the gaps were digitally filled using various graphics programs. The vessel is now stable and useable for research and measurement. It can be manipulated further if other sherds are recovered from future excavations at the site, without having to remove a solid filler material.

The main conclusion from this project is that the reconstruction of archaeological ceramics is a professional specialty that requires experience, skill and techniques that are usually not held by archaeologists as they are still currently trained in most academic anthropology programs. The development of a sub-specialty of ceramic reconstruction within professional conservation needs to be examined in-depth by both the archaeology and conservation professions. This should be done in the near future in order to avoid more potential damage to irreplaceable objects and their inherent information. This project served as an impetus to Diesen and Storch, along with the State Archaeologist, to produce educational materials, policies and forms that are now being used by contract archaeologists working in Minnesota before they can utilize archaeological objects collections for research and analysis.

1999 | St. Louis | Volume 6