Hi Patricia, Myriam,

I built the current Collections Management System for the Sixth Floor Museum
in Dallas.  This database is the closest I have come to working with a
MARC-based system for Museum purposes.  Essentially the system is
MARC-mappable, rather than MARC-compatible.  That is, the record types and
structures inside the database make it possible to export the collection
records into MARC format.

For those who are interested, I'll give a brief summary of what MARC is:

MARC stands for Machine Readable Cataloging, and was developed as a
communications format for the exchange of bibliographic data between library
computer systems.  There are a number of implementations of MARC, including
USMARC (used in the United States), CANMARC (used in Canada), and UKMARC (used
in Great Britain).

As a communications format, MARC was originally designed to aid in the
transfer of data on magnetic tape reels, and also to facilitate the printing
of standard catalog cards for manual library catalog systems.  MARC is not a
database or information retrieval system standard, however nearly all
automated library catalog systems utilize the MARC format for storage of data,
in addition to all of the major Online Library Record Vendors, such as OCLC
and RLIN.

A MARC record involves 3 component elements: the record structure (determined
by a number of International Standards, including ANSI Z39.2 and ISO 2709);
the content designation (the codes used to tag elements of data within a
record, essentially developed by the Library of Congress); and the data
content of a record (the actual information being coded in a MARC record - a
book or other object).

Typically, data content is structured to comply to data formatting standards
not included in the MARC specifications.  In the US, this usually means AACR2r
- Anglo-American Cataloging Rules (2nd ed. Revised), but also includes the
Library of Congress Subject Headings, LOC or Dewey Decimal Classification
Numbers, and other tools.

MARC format allows for 3 primary record types:  a "bibliographic" record (use
"bibliographic" loosely, to refer to any source of information which can be
represented in a catalog record), an authority record and a holdings record.
Authority records describe names and terms which need to be standardized for
retrieval of data.  These include personal, corporate or geographic names, as
well as controlled vocabulary terms.  The Getty ULAN - Union List of Artist
Names, is an example of a list of Authority Records.  Holdings records are
used to indicate the number and locations of copies of a book or other
resource cataloged in the bibliographic record.  On national bibliographic
utilities, such as OCLC, each institution that owns a copy of a given book has
a separate Holdings record which is linked back to the central bibliographic
record.

Each MARC record (regardless of type) has 3 structural components. These are,
the record Leader, Directory, and the Variable Fields.  The Leader and
Directory contain data which specifies how the record is to be processed by
the computer system in which it is stored.  There can be any number of
Variable Fields, which contain the actual substance of the catalog record
(e.g., the Title, Main Entry, Subject Headings, etc. contained in a record). A
variable field is identified by a 3-digit numerical tag (e.g., the '245' MARC
tag indicates the Title statement of a bibliographic record).  The tag is
followed by a 2-character identifier code, which describes the formatting or
content of the data contained in the field.

After the identifier codes, a MARC field is divided into Subfields, which are
specific to each type of field, and are defined by the content standards
controlling the record (e.g., AACR2 Cataloging Rules).  For example, MARC
field 300 - Physical Description, contains subfields for the Extent of an item
(the # of pages in a book), physical details, dimensions, and accompanying
materials, as prescribed by AACR2.

As you can see, MARC is a highly complex standard, which provides for very
concise data management.  So much work has been done to develop MARC that it
provides an extremely rich medium for representing intellectual resources
data.  Because of this and the standard's nearly universal acceptance in the
library community, it is potentially an extremely valuable tool for managing
and sharing Museum Collections data.  However, MARC does have flaws and
drawbacks, and to implement it in the Museum environment would require a great
deal of discipline, training and commitment on the part of the institution.
There are only a few institutions thus far who have utilized MARC for
cataloging their Museum Collections.  If I remember correctly, one is the
Inventory of American Sculpture at the NMAA.

There are a number of sources for information on MARC on the World-Wide Web:

The Library of Congress maintains a USMARC home page at:

http://lcweb.loc.gov/marc/

You can also order publications on Canadian MARC (CANMARC) from the National
Library of Canada:

http://www.nlc-bnc.ca/index.htm

-Kevin Comerford
 Media Archives Manager
 Microsoft Corporation