The purpose of this exercise in sound archive arithmetic is, of course, not to deliver a ready-made calculation model for all kinds of selection. It is, at best, a clue to the solution for a small part of the total selection problem. Establishing reliable and effective criteria is probably a much more difficult problem to solve.

However, to get back to the beginning of the paper, it is important for any archive to establish the general policy with regard to the limits of its collection. Only when it is apparent that, even within those limits, the archive simply cannot cope with the amounts of records pouring in, it should consider a more energetic selection procedure. Even then, it is better to try a kind of coarse-mesh selection in order to lose as little time as possible on that stage of the total processing of records through the archive.

If, however, records enter the archive without any cohesion amongst themselves or without any connection with the collection already present, it is necessary to apply a fine-mesh selection. In this case, it is advisable to consider the ratio between the time necessary for selection and the time involved in further processing through the archive including the cataloguing process. A fine-mesh selection, which does not result in at least one quarter of the records being thrown out, can eventually end in a bad result in terms of costly hours. See table 6.

One final consideration. Negative selection does not always have to end with the destruction of the records. If space is no problem, one can, of course, store them in some part of the archive where they can do the least harm. One can also offer them to another archive. However, sometime it is definitely better to pull oneself together and have the records either thrown out or destroyed. If some archivists here or there still believes in miracles, the author is the last one to attempt to awaken them from their dreams. However, we can be very certain that the longer we wait, the less money will be available and the more our conscience will bother us. A well-established selection policy, consistently carried out, is the best solution.

Rolf Schuursma was the librarian of the Erasmus University, Rotterdam.
This paper was given at the IASA conference in Budapest in 1981.