Page 131 - Building Digital Libraries

CHAPTER 6

Metadata Primitives

Considering the significant investment that organizations make towards
metadata creation, it seems fitting to start this chapter by looking at why
the ability to use a heterogeneous metadata system is important. The most
obvious reason for this is that libraries function within a very homogenous
environment. Since 1969, when the Library of Congress first released the
specifications for the MARC framework, libraries have optimized their
workflows and systems to support MARC record creation. Even when utiliz-
ing non-MARC formats, libraries still tend to rely on the traditional AACR2
or Resource and Descriptive Access (RDA) guidelines when dealing with
metadata creation. In general, librarians want to maintain the homogenous
nature of their bibliographic content, while experimenting with non-MARC
formats. However, no single metadata element set accommodates the func-
tional requirements of all applications, and as the Web dissolves boundar-
ies, it becomes increasingly important to be able to also cross discovery
boundaries. For many organizations, the digital repository represents the
1
first significant foray into the world of non-MARC, semantic web-centric,
metadata. And while many may find themselves wishing that the American
Library Association or the Library of Congress would step up and create
a single, universal XML metadata schema that adopts current cataloging
rules (AACR2, RDA), we find ourselves in a place where multiple standard
2
formats have been developed to allow description to happen in the schema
that is best suited for a particular type of material.
In 2002, Roy Tennant wrote his now famous (or infamous, depending
3
on your perspective) column calling for the death of the MARC metadata
format. He argued that libraries would be unable to move forward so long
as MARC was being used as the primary descriptive metadata schema, and
he thus provoked an ongoing debate about whether libraries should use
MARC or XML. However, different schema serve different purposes, so it
is normally not productive to think in terms of using only MARC or only
XML. Much of Tennant’s dismissal of MARC relates to the general rules that
govern MARC record creation.
In part, it was this question that spurred the development of RDA. For
years, libraries labored under AACR2, a set of rules created during a period
when libraries largely dealt with analog materials and still printed and
maintained card catalogs. While these rules were updated, and improved
to provide support for the burgeoning web environment, AACR2’s inability
to provide the flexibility needed to rapidly shift and support non-analog
description and discovery required libraries to develop an alternative: RDA
(Resource Description and Access).
The development of RDA was ambitious. Given the need to create
a standard set of best practices for bibliographic description beyond the
library community, RDA was developed as a first attempt to harmonize
the various bibliographic description data models and practices and bring
them together. The ultimate hope was that RDA would be used not just to

116