On Thu, 3 Aug 1995 DCORBETT <[log in to unmask]> said:


>A little thrill from seeing the handwriting a former and fabled colleague?
I
>don't think so. Perhaps a small amount of annoyance tinged with a helping
heap
>of frustration when the handwritten record is completely illegible,
whether due
>to lousy handwriting or fading ink.

>Don't get me wrong, I think hand-written records are fine, but if you have
the
>capability both financially and technologically to allow for a
computerized
>collections catalogue, do it. It is much easier to track changes on a
computer
>than on manually.

Ordinarily, I'd agree.  Properly constructed automated forms are better
than hand-written ones.  However, my recent experience analyzing the
accession cards of a well-known American art museum, proved to me how
important manuscript notations and other stylistic criteria found in
accession cards may be.

In this museum, curators were able to determine by handwriting and by the
style of the typewriter used to create an accession card which former
curator was responsible for the analysis, attribution and other information
therein contained.  Some curators were highly regarded, other's not.
Handwriting recognition allowed users to attach a level of validity and
trust to some information, and doubt and distrust to other information.

If converted directly to records in a computer database, as has often been
done with nonspecialized staff or by OCR techniques, much if not all of
this valuable information would be lost or irretrievable -- indeed, the
meaning of the text may be entirely lost, even though the new electronic
record is a faithful copy of the paper record.  Since this data is highly
dependent upon institutional culture and memory, it is clearly necessary 1)
to preserve the paper files, 2) to make sure they remain accessible, and 3)
to encourage current curators to document their knowledge in whatever
database is used to record the accession card data.

What is required to do this is the ability to attribute the data in the
record and to collect it historically.  This means that the database must
be able to record multiple versions of the same data, with attribution to
its authorship and date of creation.  Many database systems on today's
market or those home grown do not acknowledge that the information stored
in museums is dynamic and evolving in ways that require overlapping
attributions and descriptions such as I described above.  Some systems
provide these features.

An article appeared within the last year in the New Yorker Magazine that
decried the wholesale destruction of manual library cards, and the
consequential loss of the accumulated comments that librarians added to
them over the decades.  The same applies to museum accession cards.  Today,
most registrars and curators think of these items as the old fashioned way
to obtain current data, but these records often preserve the intellectual
history of the institution and its collection and ought to be granted a
special place in the museum's archival and information access programs.

In our quest to be neater, more efficient, intelligent and inventive with
our new electronic information systems, and as we replace our handy cards
and paper files with mutable microscopic magnetic fields, let us not forget
that important pieces of the baby may have been left floating in the
bathwater.
______________________________________

Robert A. Baron
Museum Computer Consultant
P.O. Box 93, Larchmont, NY 10538
[log in to unmask]