Page 133 - Building Digital Libraries

CHAPTER 6

the way that we both interact with and describe materials in the new and
emerging digital repository environments.

MARC
As previously noted, MARC was originally developed in 1969 by the
Library of Congress as a standard method for transferring bibliographic
data between systems on electromagnetic tapes. Today, MARC is the lingua
franca for transmitting data between ILS systems within the library commu-
nity. And while very few digital repositories handle MARC records directly,
a handful can export their bibliographic metadata in MARC or provide an
easy path to the creation of MARC records for stored content.
Within the library community, MARC itself is often misunderstood.
When many think MARC, they are often thinking of the fields and tags
defined by the physical rules governing the input of data, like RDA. Figure
6.1 shows a representation of a MARC21 record of an electronic thesis from
Ohio State University.
These rules define what data elements can be placed within a par-
ticular MARC field, creating specific “flavors” of MARC. For example, one
could find flavors for MARC21 (a merging of USMARC, CANMARC and
UKMARC), CHMARC, FINMARC, UNIMARC, and so on. Currently, the
Library of Congress recognizes approximately forty different MARC flavors
7
that are actively utilized around the world.
Technically, MARC is nothing more than a binary data format made up
of three distinct parts: (1) the leader, (2) the directory, and (3) the biblio-
graphic data. The leader makes up the first twenty-four bytes and contains
information about the MARC record itself. The leader will include the total
length of the record, note the start position of the field data, the character
encoding, the record type, and the encoding level of the record. However,
it is within the leader that one of MARC’s most glaring limitations becomes
visible. A MARC record reserves just five bytes to define the total length of

FIGURE 6.1
MARC21 Example

118