C. Steckner
(FOCAS, Cologne, Germany)

Obviously the ultimate database to protect and to recover artefacts would be a kind of genotype archive to allow their cloning. The database then would be the basis to identify the object, as well as to reconstruct it.
However, only a new and specific database related definition of the archaeological artefact will allow such a process of identification and replication. For the database the facts are no more represented by the verbal description and measured size nor by photographs or drawings, but by a given set of data, which allows both to understand and to generate its description as well as its photorealistic and non-photorealistic computer graphics.
For the new recursive definition of the archaeological artefact a second aspect has to be taken in account. Any description and any measurement taken from the original artefact will have some variance, as well as any pattern to identify the right set of data representing the original artefact – or even a fraction of a lost complete artefact. The provenience, the size, the classification, the reference or identification number may be uncertain or wrong. Most of these aspects of a plausibility check are covered by the statistics of shape and following methods relating all sorts of descriptors to the possibly complete artefact.
The paper will present an outline of the methods to build such a database system and will give an example of its realisation.