[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Archiving information within databases



A good discussion.

First, an observation.  From my experience with archival records and IT
systems, it's very important to avoid generalizations.  In working with a
group that combines the two disciplines (and throws in records managers and
librarians, too boot), people often stumble into misunderstanding because
they are using the same word with different meangings.  In Arizona, we're
assembling a thesaurus to help this cross-discipline communication
(http://rpm.lib.az.us/alert/thesaurus/).  The thesaurus suggests phrases
that can be used for greater clarity.

One of the best examples is the very word archives.  For archivists and
records managers, it's the three to five percent of records that we're
going to keep permanently.  For IT people, it's the stuff on backup tapes.
So when IT people hear that archives are to be kept permanently, they often
think that archivists want to keep all those backup tapes.  The language is
so ingrained in both communities, that even after being sensitized to the
problem people tend to forget to translate.


I think the question of identifying the record is much more practical than
theoretical.  We *generally* don't keep every piece of paper in a business
process; we tend to keep a final record that serves as a capstone and
memorializes the process.  Hence, I don't know that we want to preserve
every transaction within an automated business system that's stored in the
database.  We want to store the equivalent view that serves as the memorial
record.

My argument is based on the assumption (which I'm willing to be abused of)
that saving the database system in functional form is desirable.  Let me
use an example from the college and university arena (with a nod to Rob
Spindler at Arizona State University, who helped me develop this).

A fine arts major gives a musical performance in lieu of a written thesis.
The score is the record that is preserved.  The performance may also be
recorded (ah--there's that problem word--), and the recording may also be a
record (or a CD or a videotape).  However, the performance is necessarily
transient; there's no way that any person can have a *fully* authentic,
genuine, exact reproduction of the performance (you can't step in the same
river twice).  The records allow for a recreation of the original
performance, but not an exact duplicate.  We hope that the records (the
score, the recording) provide sufficient information for us to understand
the original performance; but the score may lack dynamic marks and
conductor's notes and the recordings may be low fidelity or from a single
perspective.

In my mind, the database is equivalent to the performance.  The record is
the thing preserved to memorialize the event.

With databases, we do not desire to preserve the full functionality of the
original software. (Complicated by the fact that researchers are likely to
want to ask different questions of the data than the original system could
answer.)  Rather, we want to memorialize how the data was used by the
creators of the database system.

All of which leads up to the point that one must define the record when the
system is designed to ensure that the record is fixed.  If existing data is
lost when information is updated (there's only one field for phone number,
and the old phone number is lost when the individual gets a new phone),
we've lost the prior record.  *Which may be okay, depending on value of
that information.*  We want to make sure that we know what information
needs to be memorialized in a record, and design the system to ensure that
happens.


-- Richard Pearce-Moses
  rpm@lib.az.us

A posting from the Archives & Archivists LISTSERV List!

To subscribe or unsubscribe, send e-mail to listserv@listserv.muohio.edu
      In body of message:  SUB ARCHIVES firstname lastname
                    *or*:  UNSUB ARCHIVES
To post a message, send e-mail to archives@listserv.muohio.edu

Or to do *anything* (and enjoy doing it!), use the web interface at
     http://listserv.muohio.edu/archives/archives.html

Problems?  Send e-mail to Robert F Schmidt <rschmidt@lib.muohio.edu>