Posting to RDA-L
On 27/01/2015 16.07, Ehlert, Mark K. wrote:
<a bf:Instance, bf:Monograph>
<bf:title “A Connecticut Yankee in King Arthur’s Court (pbk. : acid-free paper)”,
“Connecticut Yankee in King Arthur’s Court (pbk. : acid-free paper)”@x-bf-sort>
See also lines 2087-2090 here:
What is the purpose of doing it today?
Sorting search results after the fact, for one. There’s more than just “by date” order.
Thanks for pointing that out. The sorted title is not handled with a separate element tag, but with an attribute, e.g. adventures of Huckleberry Finn, which can be used for sorting but it does not have to be used for display.
I don’t know how Voyager catalogs work, and they certainly don’t use Bibframe, but perhaps this helps me to understand why titles vary in the LC catalog, e.g. I can browse for “adventures of Huckleberry Finn.” In the browse display, I see “adventures of Huckleberry Finn” (without the article) but in the display of the record I see “The adventures of Huckleberry Finn” e.g. http://lccn.loc.gov/2002555884
Perhaps a similar process is going on inside Voyager. Probably only a cataloger would notice this and it is unimportant.
And yes, it is correct that it will be possible to sort search results by title, but my main question about how really useful such sorting is now for the public, and how useful it will be in the future, remains to be demonstrated. I suspect other types of sorting could potentially be much more meaningful than alphabetical order by whatever happens to be the first word of the title, which in its essence is nothing other than a type of random order. For instance, sorting by classification number could be made into something that may be useful for the public. While it can be done in several systems now, I don’t think most people understand what is happening and it would take some doing to get it to be comprehensible.
And especially now, with the development of tools such as the “single search box” or “the discovery layer” when all kinds of records that differ in a host of ways (including schemas that do not delineate initial articles at all–but that is the least important of those differences!), when all of those are mixed together, what then? Title sorts–and lots of other methods–can’t work then, at least not without a ton of work. These sorts of issues are being dealt with right now by lots of libraries.
It’s a huge world of metadata, and we are only one of its smaller members. Everything is becoming topsy-turvey. That’s the world we must keep in mind and we need to fit into that world. While I think we are among the most important parts for the public, it must be accepted that we are some of the smaller “fish in the metadata ocean”.
But thanks again for pointing this out!