re: A preservation experience

["Waters, Donald" <DJW@mellon.org>]

Jim O'Donnell was good to record for us all on this list his "preservation
experience" and I sympathize on several different levels with his implicit
plea that being "gobsmacked," while thrilling on occasion, is not
desirable as a regular feature of the scholarly workplace, especially when
undertaking what one would think is a routine consultation of the
scholarly record.

In a thoughtful response, Eileen Fenton offered several criteria that she
thinks are necessary to identify those organizations that might qualify as
archives of the digital scholarly record, and which could be trusted to
help prevent one from being gobsmacked on a regular basis.

Tony Watkinson's response was reasonable enough, pointing to the rich body
of work on the preservation of digital materials that CLIR has published,
and which serves as an excellent source for deeper investigation of the
criteria that Ms. Fenton listed.  David Goodman also made a useful
suggestion that an additional criterion should be added to Ms. Fenton's
list, namely that multiple methods of preservation for any single item
would be most desirable.  

However, Mr. Goodman's contribution made me wonder how the particular
permutation he thinks is necessary ("independent of funding agencies,
independent of publishers, independent of libraries") is to be afforded.  
Jan Velterop then added to the thread that BioMed Central has all of its
articles "archived" in four places: the US, Germany, France, and the
Netherlands.  He went on to say that only the archive in the Netherlands
has committed "to transposing the material to future formats if and when
necessary."

Intrigued by Mr. Velterop's assertions, I followed his references.  What I
found differs in some significant ways from his claims.  The German
archive, for example, advertises itself simply as a mirror site for BioMed
Central.  It has been a long time since I have heard anyone claim that a
mirror site qualifies as an archive; it certainly does not qualify in the
sense in which an archive is defined in Ms. Fenton's message.  On the
other hand, I am truly puzzled by Mr. Velterop's argument that the
National Library of Medicine's PubMed Central, his U.S. archive, is not
committed to the future migration of content.  NLM performed an
extraordinary service to all interested in scholarly communications by
creating an archival document type definition (DTD), which it announced
earlier this year (http://www.nlm.nih.gov/news/electronic_archiving.html).  
My understanding is that NLM made this investment precisely to ensure that
the journal content for which it takes responsibility could be migrated
forward at minimal cost as technology changes.  

Moreover, NLM's participation agreement
(http://www.pubmedcentral.nih.gov/pmcdoc/pmcagree.doc), which Mr. Velterop
presumably signed, says that NLM will "archive the content and provide
that the content will remain accessible in the future to the Participant
and other users of PubMed Central whether or not this Agreement is still
in effect."  Perhaps I am mistaken, but this clause sounds like a
long-term archival commitment to some form of migration or other mechanism
to keep the content alive in the face of technological changes.

My takeaway?  It seems hard to escape the wisdom of Eileen Fenton's
message that there is still much confusion about what we mean by
archiving, and that we all need to be very rigorous in evaluating claims
about what archiving is or is not.  These evaluations must reference some
commonly understood criteria, such as those that Ms. Fenton suggested.  
Otherwise, it will be easy to make glib claims about the future of the
scholarly record, and difficult for scholars in the future to avoid being
gobsmacked.

Don Waters