«Part II Submissions by: Judy Anderson Heather S. Miller Bryan Baldus Allen Mullen Deanna Briggs John J. Riemer Diana Brooking Karina Ricker Lloyd ...»
The Library of Congress also needs to understand that if it gives up what it does, due to cost, our institutions are not going to pick up where they left off. Instead, that will be used as justification for us to give up the same things. They need to understand that when they make policy decisions, those ramifications do extend to the whole country, and indeed the whole world. They are a de facto national library, and even though that is not what they were created to do, I hope they can understand that and realize that the decisions they make have the same impact as if they are that national library.
Whither they go… Those are my thoughts this Tuesday afternoon. Please note that opinions expressed are my own and may not reflect those of my institution. Please feel free to contact me if you would like more elaboration. Thank you for taking on this task!
Sarah Sarah Simpson Technical Services Manager Tulsa City-County Library 400 Civic Center Tulsa OK 74103 918-596-9126 (ph) 918-596-7907 (fax) Submission to the Library of Congress Working Group on the Future of Bibliographic Control Simon Spero (firstname.lastname@example.org) University of North Carolina at Chapel Hill Prologema
1. The purpose of cataloguing is to guide a reader to his book.
2. Any aspect of cataloguing that does not support this task is superﬂuous.
3. Any aspect of cataloguing that inhibits this task is harmful.
4. A catalogue is a biologically enhanced information retrieval system.
5. Readers are familiar with other information retrieval systems, and will use the tools which, based on past experience, they believe to be most eﬀective for their needs.
6. A library may not injure a reader, or through inaction allow a reader to come to harm.
2. No ﬁnancial model predicated on ownership of such records is sustainable.
3. Claiming ownership of bibliographic records removes them from the realm of the social.
4. Providing open access to catalog and authority records will encourage their use on the wider Internet.
5. The Library of Congress is required to charge for its cataloging distribution on a cost recovery basis, even though this cost model is not eﬃcient and harmful to the Library’s larger mission.
6. Congress should permit the Library to distribute bibliographic records free of charge.
1.2 Open cataloging is social cataloging
1. Under current models of shared cataloging, fees are paid to download copies of individual records.
2. When errors are detected in these downloaded records, they are typically corrected in the local system only.
3. Because a fee has been paid to access these records, there is no social obligation felt to contribute corrections back to the source.
4. Locally changed records are much more expensive to maintain than those which are retained unchanged.
5. Just as the open source model allows ordinary users to ﬁx and enhance the software they use, open cataloging allows changes to be contributed back to the community, improving the quality of records for all.
6. Because open cataloging takes place in a social environment, social conventions can be used to solicit improvements in existing records.
1.3 Metadata should be created at the most appropriate point in the life of its data
1. Records for books should be created by publishers.
2. Records for journals should be created by the sponsors.
3. Records for articles should be created by authors and their institutions.
4. Name authority records should be created as early in an author’s career as possible, for example, on publication of their ﬁrst article in a peerreviewed journal.
5. Authors should be able to update their own authority records.
2 Structures and Standards for Bibliographic Data
2.1 Support standards that scale beyond the catalog
1. Move away from library-speciﬁc formats.
2. Move towards general standards, such as Dublin Core, CIDOC, RDF, SKOS, and OWL.
3. Design standards to allow ordinary people to create basic records for web sites, articles, etc.
4. Specify standards in a modular fashion, so that intricacies that only apply to specialized formats such as sheet music do not complicate the description of simpler items, such as novels.
2.2 Develop tools to support metadata creation
1. Automate as much of the task of metadata creation as possible, allowing humans to concentrate on tasks that require deeper semantic understanding.
2. Develop tools to assist naive users to create valid records.
3. Develop tools to assist expert users to create better records faster.
5. Whenever possible, check data against authority ﬁles automatically.
2.3 Support linked and networked bibliographic data
1. Support incremental change records.
2. Maintain references to authorized ﬁelds as links rather than strings.
3. Support multi-leveled compound works, and generalized work clusters.
3 The Future of The Subject Headings
3.1 Treat headings as concepts, not strings
1. Treating the authorized form of a heading as just one kind of label for a concept allows for the use of other kinds of labels without requiring any changes to bibliographic records.
2. Translations of the headings into foreign languages allows non-native English speakers to browse more easily.
3. Bermanization of the headings into regular English allows native speakers to do likewise.
3.2 Improve the syndetic structure of the headings
1. Assign broader terms to all non top-level terms.
2. Convert BT references that do not correspond to genuine broader terms to plain RT. Where such a change would leave a term orphaned, and a broader reference to a true parent.
3. Support more types of semantic relationships between terms.
3.3 Establish name-like headings using name-like procedures
1. It is far more expensive to establish entries as Subjects than it is to add entries that are Names.
2. Names can be established without direct LC intervention, whereas subjects require much more centralized evaluation.
(a) Fictitious characters.
(b) Famous animals.
(c) Models of calculators.
3.4 Establish patterns, rather than instances of patterns
1. Creating explicit patterns in a machine processable form can greatly reduce the number of headings that need to be separately established.
2. Common relationships implied by patterns can be automatically inferred, rather than having to be explicitly enumerated in each casethus avoiding many inconsistencies in the current headings.
3. As the syndetic structure of the headings is repaired, constraints upon the types of concepts to which patterns can be applied can automatically be enforced, and implicit patterns embedded in currently established headings detected.
3.5 Separate out domain speciﬁc vocabularies
1. If domain speciﬁc vocabularies are separated from the general headings, their administration can be distributed to other committees or bodies.
2. By delegating these vocabularies to domain experts, the headings can be made more speciﬁc and better adapted to the needs of communities of practice.
3. As the syndetic structure is repaired, delegated vocabularies can be anchored to speciﬁed broader terms within the general headings, allowing systems that only wish to deal with such general headings to collapse the delegated subtrees to the anchoring term.
3.6 Create a wiki-like system for heading maintenance
1. A social approach to heading maintenance enables a distributed approach to building relationships between terms.
2. By simplifying the creation of new headings, such a system can encourage the use of speciﬁc, narrow headings.
4. If subject headings are stored as links, rather than strings, rejected headings can be converted to those which achieve currency.
5. Any social system for headings must track the provenance of proposed headings, and clearly distinguish between established and nonestablished terms.
Many of my concerns have been expressed eloquently and in detail by David Bade, Martha Yee, and Shawne Miksa among others. I will just add a few thoughts on some assumptions that should be questioned, followed by an observation or two from my perspective as, mostly, a cataloger in public libraries.
The summary of the ﬁrst meeting uses language that often appears in our professional discussions, referring to a “consumer environment” for bibliographic data. I do not want to overgeneralize or to impute to the presenters motives that were not theirs, but I would suggest that there is a danger in thinking too narrowly of the users of our data and services as consumers. They are that; but they are also citizens and learners, among any number of other roles, as well as individuals--simply human beings. When we use exclusively the language of commerce in speaking of what we do, we obscure the fact that our profession’s values are not simply those of commerce--or should not be. The marketplace has its role, but when we start to take the marketplace as the model for everything, we fall into the absurdity expressed in
one of my favorite jokes (a cultural artifact from Milton Friedman’s time):
How many University of Chicago economists does it take to change a light bulb?
None. If the light bulb needs changing, the market will change it.
Martha Yee has referred more than once in her writings to the anti-intellectualism of our culture. Sadly, I have to agree with that judgment; and in particular, when I think about the staff of the Library of Congress and the rethinking of roles they have been undertaking, I remind myself that these librarians--among the most knowledgeable and dedicated in our profession--have for most of the past six years been accountable to a Congress more deeply anti-intellectual perhaps, and more hostile to the public and nonproﬁt sectors and to open access to information, than any other in our history. But in such a context, part of our responsibility is to be countercultural--to push back against assumptions and values in the larger society that are not our assumptions and values. Part of that push is learning how to argue persuasively, and out of informed conviction, that the cost of good cataloging is not in itself a reason to sacriﬁce it, any more than the cost of Shepard’s Citations is a reason for a law ofﬁce to Google citations instead.
One other tendency I would like to see us avoid is the uncritical acceptance of the new and the assumption that the old is necessarily, and only, an impediment to progress. Obviously the future is coming, and coming fast, whether we want it to or not, and obviously we must plan for change. This need not mean discarding existing structures wholesale. An existing system can indeed be in whole or in part an encumbrance; it can also be a valuable resource, the collective result of generations of observation, thought, and experiment. Look at almost R. Stewart, Response to Working Group 2 any land vehicle. No matter how advanced the machine, its contact with the surface of the earth is a device whose basic design predates recorded history. So when we evaluate what we are doing and how we can do it more effectively--not simply more economically--let us take care to look at the strengths as well as the shortcomings of our existing resources, and let us be open to thinking ﬁrst of how they can be adapted, improved, and extended, not abandoned or gutted of structure and content.
I was the audience member in Chicago who raised the issue that institutions other than research libraries, particularly public and school libraries, had been wholly unrepresented at the meeting, and was invited to comment from a publiclibrary perspective. As I remarked at the time, many issues are in fact shared by libraries of all types--economic constraints, for example, and problems of multilingual access.
In fact, the multiple-language issue is increasingly critical in public and school libraries, maybe even more so in some respects than in many academic libraries, both because of the sheer number of languages that may be spoken in a community and because patrons speaking those languages may not have the English-language proﬁciency, educational attainment, or research skills normally expected in an academic environment--not to mention that our public-library patrons come to us with pretty much the same wide variety of age and interests no matter what their native language.
That variability is another challenge peculiar to public libraries. Here I will raise my one minor disagreement with Shawne Miksa’s paper: when she says “that a list of hits on Google is enough to those using libraries that do not serve more scholarly or in-depth researchers,” I answer (smiling when I say that, partner), “Show me those libraries!” To be sure, some of our patrons are satisﬁed with that list of Google hits; others come in looking for the latest Lilian Jackson Braun cat mystery or whatever Junie B. Jones book they haven’t read “and that’s all” (to quote Junie B.). But others are doing school research, sometimes at a challenging level; some are tracking stocks or looking for information that could affect whether their new business succeeds or fails; still others are pursuing personal interests with scholarly dedication. And, of course, some of them are barely able to reach the second shelf, some are in their nineties, and others are every age in between (we even have our teenagers and young adults, though not as many as we’d like to reach). They all deserve quality of access to our resources. That is a real challenge to catalogers--I’ve long maintained that subject cataloging for children is harder than for adults--and much as administrators might wish it, it can’t be done on the cheap.
It is not only all kinds of patrons, but all kinds of materials, that we deal with.