On 6/25/15 4:40 PM, J. McRee Elrod wrote:
> Since a majority of works have only one expression and one
> manifestation, how important is it to assign characteristics to a
> particular level? Could that sorting out occur after there is an
> additional expression and/or manifestation?

I made a mental note to get back to this, because I think it is an 
excellent question.

First, I was curious about this and asked a cataloger what was being 
done in RDA cataloging about work titles - were they being created for 
all cataloged items? The answer was that work titles are getting the 
same treatment that they always did: a work title (aka uniform title) is 
only created if it is different from the title proper. This refers to 
RDA cataloging in MARC. I don't know how direct cataloging in BIBFRAME 
or RDA handles work titles for those works that would not get a uniform 
title in AACR2.

Now, about creating work and instance (or work, expression and 
manifestation) for those items for which these are 1:1....

First, the numbers. We have ratios of works to manifestations from OCLC. 
Of course, OCLC is unique, so that doesn't tell us much about actual 
libraries. We have the studies done by Richard Smiraglia  [1] in the 
late 1990's on a large research library, a theological library, plus 
other studies he cites. We don't have any figures for public libraries, 
medium or small libraries.

These studies tell us about the ratio of (and I'll use Patrick Wilson's 
[2] term here) "bibliographic families" vs. singlets in collections, but 
they don't tell us about the ratio of singlets vs. new additions to 
bibliographic families in the items that cross a cataloger's desk in a 
given time. So we don't know what the ratio will be in the cataloging 
workflow, nor how it will affect that workflow in different libraries.

Mac asks (rightly), does it make sense to create separate work-level 
information for singlets, even though some of them will never become 
family members? My question is: How can we actually answer this 
question? Are there efficiencies for catalogers in not creating 
work-level information, or will that be taken care of by systems? Are 
there efficiencies for systems, either by having all bibliographic 
descriptions using the separately defined work graph, or by avoiding 
that "extra step" in search and display for the majority of resources?

I suspect that "sameness" will win out for technical reasons, but I 
still wonder what the best solution is for catalogers. While it makes 
sense to me that one would think in terms of: "have I provided all of 
the necessary work information?" does that add a burden to the cataloger 
for those singlets? Are the systems that are being designed alleviating 
that burden where they can?

[1] Smiraglia, Richard P. /The Nature of “a Work”; Implications for the 
Organization of Knowledge./ Lanham: Scarecrow Press, 2001.
(Chapter 5 of the book summarizes other studies, including Vellucci on a 
music collection, which is an interesting special case. Of course, 
archives are another special case in the opposite direction.)
[2] Wilson, Patrick. /Two Kinds of Power : An Essay on Bibliographical 
Control/. University of California Publications: Librarianship. 
Berkeley, Los Angeles, London: University of California Press, 1978.

Karen Coyle
[log in to unmask]
m: +1-510-435-8234
skype: kcoylenet/+1-510-984-3600