Archive Preservation Considerations

From: Alan Lesgold <al+_at_pitt.edu>
Date: Tue, 25 Aug 1998 22:27:28 +0100

The Walker article, combined with Harnad's comments, present a picture that
seems ideal. However, it does not attend sufficiently to the issue of archival
preservation. Traditional journals were archived in part by printing numerous
paper copies and sending them many different places. Except for the early part
of the present century, paper has proven to be a wonderful storage medium. I
have read paper documents 2,000 years old. Have you successfully read any
20-year-old electronic documents? My point is that we face an urgent need to
develop and finance enduring archival preservation for at least the modest
share of our scholarly output that is still worth reading after a decade.

It is not yet clear that the best archival strategies are, but I suspect that
they include the following aspects:

1. Multiple sites store independent copies of journals in formats with
relatively long lifetimes, such as CD-ROM or DVD.

2. Support for those sites comes from a continuing income stream provided by
libraries and scholarly societies. The income stream must be sufficient to
support copying of all media every five years as well as conservative
amortization of server hardware to assure up-to-date operations and high
accessibility.

3. Escrowing of copies of all journals so that the "investors" gain access to
them if the server organization fails. This is standard practice when companies
buy software and seems prudent for archival servers as well.

The basic principles are (a) that any new approach to journal distribution must
be part of a systematic effort that attends to all of the functions and
systemic relationships in the current approach; and (b) that payment streams
roughly match required functions. In particular, schemes that include
panglossian assumptions about costs must be avoided, unless we expect all
future archival retrieval to be an archeological activity. Libraries are
abandoning their archival roles out of financial necessity, leaving the
publishing marketplace to provide this function. This will make the costs
explicit rather than burying them in several thousand college budgets, but it
won't make the costs go away.

Societies that go to simplistic giveaway arrangements for articles will likely
be dooming those articles to very short lifetimes.

========================================
Alan M. Lesgold
Professor of Psychology and Intelligent Systems
Executive Associate Director, Learning Research and Development Center
University of Pittsburgh
3939 O'Hara Street
Pittsburgh, Pennsylvania 15260
Email: AL+_at_pitt.edu -- Phone 1-412-624-7045 -- FAX 1-412-624-9149
Received on Tue Aug 25 1998 - 19:17:43 BST