NGC4LIB what is a “next-generation” library catalog?

Posting to NGC4LIB

On 5/3/2016 10:29 AM, Eric Lease Morgan wrote:
So, what is a “next-generation” library catalog?

About ten years ago a mailing list called NGC4Lib was created, and via a number of bullet points the list more or less posed the question, “What is a ‘next-generation’ library catalog?”: [1]

  • Who are the primary intended audiences for a library’s “card catalog”?
  • Considering the changing nature of information access in an Internet environment, how is an electronic “card catalog” of today different from the one designed ten or fifteen years ago?
  • What kind of content should these “card catalogs” contain?
  • To what degree are these things “catalogs” (as in inventory lists), and to what degree are they finding aids?
  • To what degree should traditional cataloging practices be used in such a thing, or to what degree should new and upcoming practices such as FRBR be exploited?
  • How would such a thing get created and by whom?
  • What are some of the functionalities of “next generation” catalog?

Now that NGC4Lib is coming to an end, I hope we — the Library Community — have learned something. If so, then what have we learned? To what degree are we able to answer the primary question? What has happened/developed in the past ten years to inform our answers? Are we further along? Have we advanced knowledge and understanding? Has library service been improved? Was it worth it?
I invite you, — the community — to reflect, articulate, and share your thoughts. Maybe, sometime in the future, our ideas will form the basis of a historical description of librarianship near the beginnings o the 21st Century.

Personally, I have enjoyed this list mainly because it seemed one place where IT and cataloging could meet and exchange ideas–even though it could sometimes get bumpy!–but that seems to be happening less and less. The general acceptance seems to be that Bibframe and Linked Data will solve all of our problems–that is, after the 30-40 or so years that it will take for it all to be implemented, and of course, there will be no problem during all this time getting the scads of money we will need since the funding will be raining down on us from the Heavens. These ideas seem to have taken root, so everything seems to be solved.

There is also the library “Discovery Service” which many would probably say IS the next-generation catalog. It is mainly nothing much more than either a federated search (where one search box actually searches multiple sites and databases) or everything is bunged together into a single database (normally MARC) or both. The belief that this is the solution comfortably avoids the problems of authority control (since not all databases use the same forms, if they use any forms at all) and different records based on different rules get mixed together, along with unpredictable results from full-text searches. To be fair, since our authority records have been so well-hidden from the public for such a long time, it doesn’t surprise me that lack of authority control is not seen as a problem.

The traditional idea of a library catalog was always to provide a more-or-less complete and reliable listing of the materials in the collection that people could consult relatively quickly and easily; the same tool had to serve both the public and internal inventory control. Finally, there was a healthy dose of reality: the catalog could not be too expensive to create and maintain. When the catalog got too complicated or proved inadequate, the public complained loudly. When it got too expensive, heads in technical services began to roll!

The main words in that definition of a catalog were: “complete” “reliable” “quick and easy to consult” and this implied consistent cataloging rules plus authority control of various types. The authority methods used in traditional catalogs have worked fairly well, but they were always clunky, got clunkier when computerized, and many have always thought they are too expensive anyway.

The goals of linked data and the discovery services are quite different and far more abstract, but no matter what, it seems to me that implementing either one will result in something far more complicated than any catalogs today. Sites searched through the discovery services will never follow library methods, and linked data, while it can in theory provide authority control, the reality will be very complex and probably won’t work out in practice. That will all be very expensive too.

So I wonder: will there continue to be a need for a “more-or-less complete and reliable listing of the materials in the collection that people can consult relatively quickly and easily” without the bells and whistles of (probably) tons of additional information dragged in through linked data: the images, the Wikipedia info, charts, graphs, maps etc. etc., or the extras found through the discovery services? Again, this still implies consistent cataloging rules and authority control.

One last question, perhaps the key one: will the catalog be seen as a tool for information discovery, or will it be seen as a tool to get an item that someone has already found elsewhere, on Google, Amazon, Linked-in, a shared citation, or something similar? In that case, although there will always be a need for a complete inventory tool there would be little or no need for authority control. For instance, I can imagine a simple tool that works this way: You would see a citation on a webpage or in an email but it would actually be marked-up in, and in the background before you even noticed the citation, the computer would have searched it in a series of databases (which would include a library’s catalog), and if it found a convenient copy for you, it would say “Buy it!” or “Borrow it!” or something. If there were no copy, you would see only the bare citation. In this case, the “catalog” would run completely behind the scenes and the user wouldn’t even know it existed. It wouldn’t even enter your mind to search it for related materials.

From what I have read in the library literature, I think lots of people want exactly that.

If the answers to these questions is that yes, we want a tool that lets people discover what is in a collection and it should be complete, reliable, quick and easy to consult, then I would say there is a need for what we think of as a catalog; if no, then the linked data/discovery services will be adequate.