"42" |
Introduction: Don’t Panic, It’s a Common Disaster
In their 1996 Lay It Down album, the Cowboy Junkies’ lead vocalist Margot Timmins asks the question, "Oh, won’t you share a common disaster?" (from the song, "A common disaster," written by Cowboy Junkies’ guitarist and Margot’s brother, Michael Timmins). This is in fact just what we do share.
While "disaster" may seem on initial appraisal a harsh characterization of the current tumultuous, shared, electronic information landscape, one does not have to stray far into either the recent library professional literature or skim many current issues of computing science journals, or even the unevenly researched and informed work of the digeratti in Wired and similar publications to learn two things: 1. the electronic information "commons" is a rich, vast, but unevenly accessible global resource; 2. everyone and their neighbors are busy trying to develop information access technologies and standards to make the commons easier to organize and navigate. With the rapid ramp-up of the Internet and especially the World Wide Web, it’s as though a tornado (our "disaster"?) has ripped through the existing information landscape disrupting all the various more-or-less independent meta-data (i.e. 90’s speak for "cataloging") villages in its path and has left everyone (librarian or not) with a lot of the same "how do we build a new infrastructure?" headaches. It’s a radical concept, but what if we found out that museums and libraries and archives and newspapers and etc., etc. have similar information-about-information needs? Could we build the solutions in concert?
So if we think of the here-and-now electronic resource landscape as a "common disaster" (either a land laid to waste, or a new, unconquered frontier -- take your pick) this presents us the opportunity to freshly examine and form opinions on
In other words, what should the infrastructure(s) be and if we’re not already on course, how do we get there?
It is exactly these issues taken from both a library perspective and beyond that "42" will selectively and thoughtfully examine. Three writers, each with different perspectives and a penchant for long lunches (when gathered as group anyway), will collaborate in selecting topics, assigning a lead writer to draft a piece and then as a collective, we’ll refine and eventually submit a "42" column for each issue of the Journal. Who knows, we might even occasionally get it right?! Who are the writers? (See below).
Why the title, "42?" For fans of Douglas Adams’ very amusing novel, The Hitch Hiker’s Guide to the Galaxy, based on the original radioplay, the title was a dead giveaway. As for the rest of you: "42" is the answer Deep Thought, a gargantuan computer built especially to answer the question, "What’s the meaning of life, the universe, and everything?" provides. It seems sometimes the questions may be more enlightening than the answers. We hope JIC readers will be challenged, perhaps even provoked, and hopefully left the wiser by each "42" column. Suggestions, feedback, and comments will always be welcome -- please direct them to the JIC editors.
And lastly, when all hope seems to be lost, remember the words emblazoned on the cover of the portable, Hitch Hiker’s Guide to the Galaxy:
Don’t Panic!"42" Now that We Know the Answer, What Are the Questions?
Time to take stock.
Cataloging is one appropriate and valuable method of providing improved description and access for electronic resources. Just a few short years ago this assertion was in doubt, and, for some, it remains so. Until recently, libraries had little or no experience cataloging Internet resources (computer files, yes; Internet resources, no). The suggestion that libraries catalog Internet resources in traditional ways evokes a range of understandable reservations and outright objections, which I have boiled down to three: (1) there is nothing on the Internet worth cataloging, (2) everything on the Internet is here-today-gone-tomorrow, and (3) MARC and AACR2, which were developed primarily to control print materials and to facilitate the production of printed catalog cards, would not work anyway. So why even consider it?
I call these complaints the Three Stoppers. To this day, these beliefs continue to prevent some libraries from cataloging Internet resources. Because the consequences of the Three Stoppers are significant––the unexamined dismissal of library standards, practices, and systems, for example––I propose that we explore their underlying tenets. Let’s take them one at a time.
There is nothing on the Internet worth cataloging. Generalized assertions are easily dismissed. Of course not everything on the Internet is worthless. This belief finds its strength, however, in the realization that the Internet is populated with millions of freely accessible resources that fall outside most libraries’ collection policies. Things like the Squashed Bug Zoo come to mind, although I shudder to offer specific examples lest I be liable for too quickly diminishing the value of pictures of squashed bugs.
Libraries have never cataloged the entire universe of print materials, and for many good reasons: cost, time, value, warrant. This is important to remember. And libraries should not envision extending cataloging treatment to the universe of Internet resources. Same reasons, squared.
But not everything on the Web is a squashed bug. To dismiss all Internet resources as "not worth cataloging" would be to ignore an important and growing body of data, information, and knowledge for which bibliographic control is not only appropriate but vital. Libraries have identified and selected a small but growing number of such resources, 15,000 of which are cataloged in WorldCat (the OCLC Online Union Catalog) and are publicly available through InterCat, an experimental database comprising only bibliographic records for Internet resources (http://purl.org/net intercat). As of this writing, the number Interner resources cataloged, the number of cataloging libraries, and the rate of cataloging continue to grow, albeit modestly.
Even these 15,000 resources, however, are not of equal interest to all libraries. What material is? But they are of interest to the libraries that chose to expend the resources to identify, select, and catalog them presumably because of their expressed or anticipated interest to library users. Indeed, it is likely that these resources would be of interest and value to other libraries and their users who share common collection goals or information needs. Shared cataloging remains viable for Internet resources.
Everything is here-today-gone-tomorrow. Well, not everything, to be sure, although reliable, long-term access to Internet resources is a problem that has not yet been satisfactorily solved. Who has not encountered "Error 404, Object Not Found"? You’ve just experienced a dead link. Understandably, some resources may be temporary by nature, not intended for long-term access. Many other resources, however, become unreachable when URLs change, and from a user’s point of view, unreachable resources share this in common with resources that do not exist: they are useless.
Cataloging Internet resources does not address this problem directly, but the resources that are typically selected for cataloging tend to be rather more stable compared to the Internet at large. This is easy to understand. Producers of significant resources want their users to find and access their resources with ease. Reliably. Over time. For those who care about the information they publish, there is no advantage to impermanence or transience.
Selecting stable resources is a good initial strategy, and any library approaching the cataloging of Internet resources would do well to spend as much time revising their selection and collection development policies as they do their cataloging procedures. But selection alone is insufficient. Internet resources do move, and hoping that they won’t is a losing game. Alternative solutions include Uniform Resource Names (URNs) and Persistent URLs (PURLs). The former enables globally unique and location-independent naming, which names can be "resolved" to reveal location or other related information. PURLs create a one-to-one relationship between an unchanging, persistent URL and the actual URL of a resource, a relationship that can be maintained easily over time.
MARC and AACR2 won’t work. Fifteen, twenty, fifty, or a hundred thousand MARC records later, some will still assert this point of view. It’s a point of view that could be applied equally to materials in other media, so the charge is not new. Nor is it to be summarily dismissed.
Librarians may wish that the USMARC format were more flexible, able to express hierarchical relationships more easily, for example, or able to inherit information from related records. But, as a standard communications format, USMARC is serviceable, if not perfect, and remains the most widely accepted format for machine-to-machine communication and exchange of bibliographic information. The benefits of adopting MARC format records are numerous and immediate: leveraging existing MARC-based systems, exploiting the economic benefits of shared cataloging, preserving and extending the useful life of deployed systems, facilitating the machine-to-machine exchange of bibliographic data, and enabling the integration of records for Internet resources into local catalogs. Because these benefits occur at the margin, nearly all libraries can realize some benefits with modest training and effort.
MARC needn’t be perfect to work. It’s not, and it does. So it is with the Anglo-American Cataloguing Rules (AACR2). Libraries are making them work, but the gap between the current rules and what may be desired seems to be larger with AACR2 than with MARC. This fact is not lost on the Joint Steering Committee for the Revision of AACR, which hosts this October in Toronto, Canada, an international invitational meeting to discuss the principals and future development of AACR. Bold committee action will seize the moment to rethink and refashion AACR in ways that break inhibitive ties to nineteenth century notions of bibliography. Less than that may yield distinctly unsatisfactory tinkering that will accomplish little and take a long time to do it. Time is in short supply for libraries that want to stay relevant in the rapidly transformational world of the Internet.
Are Library Methods Stop-Gap?
Using traditional cataloging methods provides benefits that libraries and their users can realize immediately. The MARC format and AACR2 cataloging rules have been widely adopted and implemented in thousands of library systems worldwide. Moreover, an increasingly efficient and highly skilled workforce, distributed among libraries and other information organizations, has created an unparalleled exercise in cooperation: shared cataloging. Such sharing is the basis for a large community of libraries that participate in cooperative cataloging ventures or union catalogs. Recorded in a MARC record, information describing an Internet resource can be created by one library and used by another with little or no additional work. Records thus created and shared can be integrated into library catalogs along side those for information in other media. Thus integrated, library users can benefit from the OPAC’s familiar searching, retrieval display, and navigational capabilities to explore an information storehouse than includes Internet resources, whether local or remote, print or digital. In Web-based OPACs, access to the Internet resource is just a click away.
Simply put, the OPAC can be a centralized resource that library users turn to for concise searching of a well selected collection of resources in all formats.
Summary
Information resources, regardless of their medium, exist on a value continuum; so, too, the methods of organizing, storing, describing, and providing access to information range greatly. For any given information resource, the "cataloging" methods used are generally associated with the value ascribed to the information: the more highly valued the information, the more rigorous the methods. Libraries would do well to apply this simple logic to the cataloging of Internet resources. A good rule of thumb might be this: If an Internet resource were published in any other media, say, paper, and it met your library’s selection and collection development criteria, would you catalog it? If the answer is Yes, then select the Internet resource and catalog it. Libraries that take this approach find themselves offering improved patron services, staying abreast of technological change, and positioning themselves for next-generation cataloging alternatives even if their actual level of cataloging is low. Some cataloging of Internet resources is far to be preferred than no cataloging.
Cataloging is relatively expensive, however. We must continue to drive down the unit cost of cataloging, and Internet resources are likely to yield to cost reductions more rapidly than other media. Standard methods of encoding metadata will soon enable resource creators to associate cataloging-type information in ways that can be universally accessed by software applications. With information supplied and captured "upstream" from the cataloger, libraries can begin to imagine entirely new cataloging systems that may, for instance, create a basic bibliographic record from encoded metadata. At a minimum, such approaches will drastically reduce the amount of transcription catalogers perform, reducing time and entry errors. Selectively applied, librarians would add value to basic records through subject analysis, classification, authority control, and uniform headings. And we can imagine computers assisting those functions as well.
Regardless of the level of cataloging applied or the methods whereby metadata or cataloging are created, libraries must also begin to address problems related to long-term access to digital resources. Cataloging is a long-term investment, and best applied to resources whose expected value exceeds the immediate and may extend to decades or centuries. What role will libraries play in long-term digital archiving? That question is better answered from a position of experience gained through providing bibliographic control for selected resources today. Again, the balance tips in favor of experience gained through cataloging.
So, time to take stock.
Let’s dump the Three Stoppers. Inaction gains nothing.
The library that identifies, selects, and catalogs even one Internet resource benefits in ways that far exceed the incremental costs. And the library that systematically integrates the identification, selection, and cataloging of Internet resources into its routine practice will find itself both shaping the future and all the more prepared to meet it.
If you haven’t cataloged one Internet resource, do it for yourself. If you already catalog Internet resources, accelerate the pace. Do it for others: your users, other libraries, and the future well being of your profession.