Abstract
This paper examines the various steps already taken, or currently planned, to build the basis of primary information for art documentaion. It presents the very informative examples of methods taken at the National Art Library, the Victoria & Albert Museum, and the Witt Computer Index of the Courtauld Institute of Art.
It points out that there are variations in data describing each work of art in art history and considers what demands this fact poses on the data formats used at libraries. Some constructive examples of projects to computerize research in the fields of art history and collection management are presented.
