Posting to RDA-:
On 1/28/2015 7:00 PM, J. McRee Elrod wrote:
While those reared on Google and other search engines may mostly search by keyword, there are still older folk amongst us who like left anchored exact word searches, whether resulting in a hitlist or a browse list.
To make my point clear, I am saying that it’s OK to keep on with the filing indicators and other traditional methods. But there is a lot more to it, for the following reasons:
- When it is clear–and it is crystal clear–that libraries want to enter the world of “linked data” and/or
- we want to create a “single search box” that searches “everything” (whatever “everything” means)
When we do either, this means that our “separate data silos” (i.e. library catalogs) will no longer be separate and will become part of larger entities. This can happen by:
- libraries will make a “discovery tool” or “discovery layer” that searches our catalog along with dozens of other library and non-library tools, each with quite different types of information and structures, and/or
- we load into our catalogs massive numbers of metadata records created by completely different methods–as many libraries do now when they take perhaps tens or hundreds of thousands of metadata records created by various organizations, converting (shoehorning) them into a simplified form of MARC21 using tools such as MarcEdit, and importing them into the catalog and/or
- our records are exported in other formats (e.g. schema.org, Bibframe) and loaded into some non-library systems, e.g. Google or Bing, or perhaps there will be other special systems that aggregate everything together
(probably other ways too)
As we enter this world–against our will in many cases, but we are compelled to do so–then we know without any doubt that things such as sorting by title cannot work in that environment because many (the overwhelming majority?) of the non-library records do not add non-filing indicators. And there are massive numbers of those non-library records, often dwarfing the number of records in the traditional library catalog. If those non-library records are to work for sorting by title, it will demand a huge amount of editing of hundreds of thousands of records (or more).
Library catalogs are no longer separate, or those that are probably won’t be much longer because separate library catalogs are becoming ignored by society. (Perhaps this is sad but it has been shown conclusively many times and is the reason for the creation of the “discovery tools” and Bibframe–so that people will have a greater chance of seeing our records)
Personally, I like the idea of the single search box, and I know a lot of places consider the idea of a separate catalog as backward–that it must be included with everything else because people just won’t search it separately. All right, but there are serious consequences to doing it.
Only one tiny instance of this is that sorting by titles cannot work in such an environment and this relates back to the utility of the non-filing indicators. This is a relatively insignificant issue as I pointed out, and there are far more important issues, such as our traditional methods of authority control and more importantly, the usefulness of our subject strings.
This is why I keep harping on the point that we must see the catalog through the eyes of our users (who demand things like the single search box) and not through our own eyes that focus on the local library’s catalog. From the user’s point of view, sorting by titles will not work in that environment that they will use, and we must acknowledge that. If we try to implement a title sort anyway, the user will get bizarre results they cannot understand. Therefore, to make title sorting work in the environment of our users (which is by far the most important environment) would demand either a huge outlay in resources from cataloging, or some pretty tricky technical manipulations from IT. What would either cost? Are they worth it? What are the tradeoffs? Who should decide?
I have always been hoping that librarians and catalogers would begin to see this new environment and plan for it. That is part of what I was hoping RDA would do. Issues such as non-filing indicators, that seem so small at first, take on new significance.