Libraries and Standards

By Karen Coyle

Preprint. Published in The Journal of Academic Librarianship, Volume 31, Number 4, pages 373-376

Introduction

Libraries are in a unique position to take advantage of standards as compared to many other institutions. Unlike banks, or manufacturers, or retail businesses, libraries are not in competition with each other. And unlike elementary schools, or city governments, or non-profits in general, libraries have a strong professional connection that promulgates standards.

Libraries also have the motivation for standards. Like old-age pensioners, libraries exist on fixed incomes that notoriously do not keep up with inflation. Standards create efficiencies both for libraries and for the vendors who serve them. They make it possible for all libraries to be customers for the same library system design. They also make it possible for libraries to share data. When I want to impress non-librarians with the incredible efficiencies in libraries, I tell them about the MARC21 record, which is created once (often by the Library of Congress) for each book published in this country, is stored centrally by library service providers, then is downloaded to the database of every library that purchases that book. This an impressive savings of time across the libraries in the U.S.

Standards not only create efficiencies in terms of time and costs, they also provide a uniformity of product that is important to customers. Many standards in the manufacturing area assure customers that indeed a new light bulb will fit into the old lamp, and that, at least within a single country, all electrical items purchased will have plugs that match the wall sockets of the buildings. These are issues of interoperability between parts of the same system. In libraries, standards are also about interoperability, not only for the exchange of data but also the ease with which library users can move from one library to another without having to learn entirely new skills in bibliographic research.

The Landscape - Who's Who In Library Standards

The original library standards were set by the fledgling American Library Association in the late 19th century. ALA created standards relating to cataloging and the creation of catalogs. Today ALA is still involved in the development of cataloging rules, but the development of library standards has been taken up by the National Information Standards Organization (NISO). NISO is a formal standards development organization that is accredited by the American National Standards Institute (ANSI). NISO is responsible for all standards in the Z39 range. These represent standards for libraries, information systems, and publishers. In addition, NISO is designated by ANSI as the U.S. representative to the International Standards Organization (ISO)Technical Committee 46 on Information and Documentation. NISO is a membership organization and its activities are funded for the most part by member dues. The membership, primarily professional organizations in the library field and vendors who serve libraries, both reflects and directs NISO's mission. As a formal standards organization and a member of the ANSI family of standards organizations, NISO operates under certain "rules of engagement," such as the consensus process of standards development, and the periodic review and re-certification of standards in its care.

NISO now "owns" the original MARC record standard, originally ANSI Z39.2 and now ANSI/NISO Z39.2, and was the conduit to getting that standard certified at the international level through ISO as ISO 2709. The organization has about two dozen active standards ranging from the management of libraries (ANSI/NISO Z39.73 - Single-Tier Steel Bracket Library Shelving; ANSI/NISO Z39.7 - Information Services and Use: Metrics & statistics for libraries and information providers), to publishing (ANSI/NISO Z39.9 - International Standard Serial Numbering (ISSN) , Z39.18 - Scientific and Technical Reports - Preparation, Presentation and Preservation), to information retrieval (Z39.50 - Information Retrieval : Application Service Definition & Protocol Specification, OpenURL). Yet the technology standard that is the most used by libraries, the MARC21 standard for library cataloging, is not a NISO standard. This standard is instead managed by the Library of Congress.

Library of Congress was the force behind the development of the ANSI standard that defined the structure for the Machine Readable Cataloging record (MARC) in the 1960's, which was needed to create computer-driven print-on-demand service for the Library of Congress card program. Using that structure, the Library of Congress developed the fields and subfields that would be used to encode the content of a library catalog record. While the record structure of the MARC record has not changed, and is still defined by ANSI/NISO Z39.2, the content of the record has been under constant evolution under Library of Congress's care.

In addition to the MARC21 standard, Library of Congress is the maintenance agency for some other standards. As maintenance agency, the Library is the central information point for the standards and any documentation related to the standards. The degree of formality of these standards and the amount of maintenance varies, however. The Library of Congress is the maintenance agency for Z39.50, one of the best known of the NISO standards, and for the web-based version of Z39.50, SRU/SRW. It is also the maintenance agency for METS, an independently created standard that is used by digital libraries to structure resources for storage, and for a number of standards created by the Library of Congress itself, such as the XML transformation of the MARC21 record, and MODS, a bibliographic description record based loosely on MARC21. Although these ad hoc standards gain authority through their association with the Library of Congress, they generally serve a smaller community than the standards that have been treated to the formal process that NISO provides. Another notable library organization that has engaged in the development of standards is OCLC, and its sponsorship of the Dublin Core Metadata Initiative (DCMI). Dublin Core is especially interesting in that it was expressly developed as a non-library standard. Although Dublin Core could be used by libraries, the organizers of the effort wanted to create a standard would be a light-weight resource description language that could be used by organizations that do not have the history or experience of libraries. DC is not associated with any particular set of cataloging rules so that any community that has a need to describe documents can make use of it. In fact, Dublin Core is used widely today both within and outside of the library community.

Some standards do not have an actual standards organization behind them, but are curated by a group or committee of interested parties. The METS standard, mentioned above, is such a standard with its own governing board, as is the Text Encoding Initiative (TEI) , a standard for the markup of books and other texts. This type of standard is viable as long as the community is relatively small and cohesive; expansion of the technology's influence usually requires more structure and more coordination than a small group can provide.

Another committee effort, but one with wide adoption, is that of the Anglo-American Cataloging Rules. Although not strictly a technology standard, AACR has a profound effect on the technology of libraries. The Joint Steering Committee for Revision of the Anglo-American Cataloguing Rules has six member organizations representing the Anglo-American library world.

Non-Library Standards In Libraries

The highly networked world in which we live has increased the flow of information in all aspects of our lives. Libraries have taken advantage of these new communication technologies since their very inception: some of the very first computerized library catalogs were also the first open access databases on the Internet in the early 1980's. Today's libraries are fully integrated with the Internet and its technologies, providing services and delivering documents to users over the network. In terms of standards, this has meant that the library has become an active user of the standards underlying the Internet, and of the application standards that make it possible to deliver library services over the Net. The user interface of library catalogs today is written in HTML, the language of Web interfaces, and these travel over the Net using the Hyper-Text Transport Protocol (HTTP), both managed by the World Wide Web Consortium (W3C). The basic Internet protocols that allow all this to work are the standards of the Internet Engineering Task Force (IETF). The MARC standard now allows the creation of MARC21 records using the Unicode character set, an ISO standard (ISO/IEC 10646), in place of the ANSEL character set that had been the library standard for many decades. And digital library applications are increasingly being written using the eXtensible Markup Language (XML) , also managed by W3C. Some future applications, especially in academic libraries, will make use of the Shibboleth standard to authenticate library users, and perhaps XACML, an XML language that defines authentication criteria. Shibboleth is being developed by the Internet 2 project, and XACML has been developed by OASIS, an organization that fosters standards in XML. These are just a few examples of standards that libraries are using.

Once libraries had their own separate technology, based on standards that were not shared with other communities. Now libraries are often developing their applications from technology that originated in the broader information community. There are a number of positive aspects to this, not the least that library technologies developed on non-library standards will have the capability to interact with products in the broader information technology world. When Google introduced "Google Scholar" services early in 2005, within days librarians were announcing experimental links that they had developed between the Google service and their library catalogs. These links used the OpenURL, a library standard that makes use of HTTP to connect metadata to services. The OpenURL service has now been added to Google Scholar for some academic institutions. This is a good example of interoperability between libraries and non-library information services, and how these are made possible through standards.

Libraries and Information Standards

In the previous column I talked about changes in the standards landscape, from a traditional process that put long-lived standards in place to a modern, fast-paced process that allows standards to be part of rapid-development. This new standards environment affects libraries as it does any other institution, but there is another aspect of today's standards development that is having even more of an effect on libraries and library technology, and that is the near ubiquity of information standards in all environments. There are information standards that affect individual industries and most industry standards groups list information standards as part of their charge. But information is itself an industry today, encompassing computing and networks, online and offline publishing, scientific data, and web services, among others.

The intense philosophical questions on the role of libraries in this new information age, often couched in terms like "the library vs. Google," are played out in the standards arena as well. There, the question is whether library standards are - or should be - information standards, and what role the library profession should play in the development of standards in the general area of information technology. Regardless of what one thinks should be the case, the fact is that standards developed in the library community are rarely adopted by others, even when the standard is highly flexible and could serve a variety of purposes. As an example, the data structure of the MARC record that was developed in the mid-1960's was very advanced for its day, allowing the creation of variable-length fields in a time when most computer systems were limited to fields with short fixed lengths. Surely this could have been of use to others who were working with variable length data, but such use did not arise. Z39.50 can serve as another example: the basic protocol allows the retrieval of records from a database, not just a bibliographic database. Although adopted by some members of the geospatial community for database searching, the great majority of users of Z39.50 are libraries and library systems. There is no analysis of these standards that would answer the question of why they didn't gain greater acceptance. One possible answer is that although these standards appear to be generalizable, in fact the need that they fulfill for libraries is not one that is felt in other communities. Another answer could be that libraries did not to reach out beyond their own community with these technologies and therefore they did not come to the attention of others who could have used them.

The example above of Google Scholar and its use of the OpenURL shows that library standards that are based on standards of the broader information community, as OpenURL is based on HTTP, can be used to create interoperability between libraries and non-library services. Google is positioning itself to bridge the gap between the web and libraries with its Scholar and Print services, but the OpenURL was developed as a NISO standard with library and library vendor representatives. If the OpenURL standard were being developed today, it is possible that Google would have been a member of that standards committee.

There are standards that are not library standards but in which libraries have had a formidable role. Dublin Core, a general metadata standard, developed out of the research division of OCLC. Those working on Dublin Core made an effort to create a community that was not limited to libraries, and in fact Dublin Core has become the standard for descriptive metadata in the web community. Libraries also had a role in the creation of the Unicode character set because they were able to provide language expertise to the Unicode committee. These two experiences tell us that libraries are able to influence standards outside of their own community, but to do so they must join broader standards efforts. Unfortunately, those broader standards efforts are expensive, both in terms of the time that expert staff must spend on the standards, and on the costs to join the standards effort and attend meetings.

Problems and Issues

There are many things that are right with the world of library standards: libraries have a long history of standardization of their operations, the result of which has been great efficiencies in the management of libraries; libraries have begun to make more use of mainstream standards and therefore are gaining interoperability with larger world of information services; and libraries have broken through their own boundaries and have contributed to standards used by others in the information arena.

But there are also some areas of the library standards environment that are problematic. As noted above, although NISO is the primary organization developing and maintaining library standards, there are mature standards in the library environment that are not included in NISO's standards purview. This creates a need for coordination among standards developers which currently is not taking place. A notable lack in the library standards world is the role of professional leadership that would logically be coordinated by our primary professional organization, American Library Association (ALA) . ALA, however, has divested itself of its standards role over the last decades. NISO could conceivably play this coordination role, but although many professional organizations are members of NISO, NISO's role is broader than just libraries, as the "Information Standards" portion of its name implies. In essence, we are a profession without a professional standards focus.

The fact of having multiple standards organizations requires that specific roles in the standards sphere be clarified, something that has not occurred in the library standards world. Many standards efforts will naturally begin as isolated development projects, and some will stay small and ad hoc for their entire lifetimes. But where a standard could benefit a larger segment of the community there needs to be a way to channel the efforts into a standards assembly line. What happens today, however, is that developers of applications are faced with a set of unattractive choices: the most formal standards body, and the one that has the support of the largest community, is NISO. However, the NISO standards process is known to be slow and tedious, not in keeping with the fast pace of information technology today. NISO does, however, provide a process and structure for standards development which are funded by the NISO members. Developers who choose to remain independent have to set up a governing body and have to fund the activities themselves. The risk is that activities like maintenance of the standard will be slighted. But the greatest risk is that the application will not reach a wide enough audience to become integrated into general library systems, and therefore most libraries will not be able to make use of the technology. NISO has initiated a light-weight standards registration process that allows the NISO community to put its seal of approval on ad hoc standards that have not gone through the full, formal process that NISO standards require. It remains to be seen if this process helps those standards find greater distribution in the library community.

Conclusion

At the end of the previous article I asked the question: Does it even make sense to talk about a "library" standard? There are some standards that we can call library standards, such as the NISO standard for library statistics (Z39.7) or the standard for library shelving (Z39.73). The goal of technology standards, however, is primarily the exchange of information, and what we think of as library standards are often technologies in which the library is one of the parties in that exchange. Libraries are broadening their potential partnerships by making use of mainstream technology in their standards, such as XML and the technologies of the World Wide Web. These interchange standards still need to serve the library's mission and goals, however, and a vital role of our standards efforts is to define technology that supports the values of this unique institution.



The copyright in this article is NOT held by the author. For copyright-related permissions, contact Elsevier Inc.