On Wed, 14 Sep 2011 11:21:25 -0700, Karen Coyle <[log in to unmask]> wrote:
>Peter, I was hoping we'd hear from some catalogers, but here's my take:
>
>The transcribed parts are intended to be a surrogate for the title
>page (primarily). That way, if you have metadata for two books (or one
>book and metadata for a similar book) you should be able to compare
>the transcribed fields and see if they are the same.
OK, I'll bite (hoping that a retired cataloguer counts)!
Transcription is intended to match what's on the title page (or other
principal identifying matter). As others go on to explore, it's an
imperfect match. However it does matter a great deal, because we need also
to match what writers do when they create a citation for what the give as
sources, additional reading, reference lists, systematic bibliographies, and
it needs to be able to respond to searches which depend on informal
citations too -- comments in academic lectures or conference presentations,
word of mouth information, passing citations in radio and television talk.
>
>HOWEVER...
>
>Not all of the transcribed data is transcribed "as is." For example,
>the case of titles is changed to sentence case, regardless of the case
>of the title on the title page. This is because the title, like so
>many data elements in cataloging, performs more than one function in
>the catalog record. It is both a surrogate (which in fact it isn't
>quite) and an entry element.
Sentence case, title case or all caps, and font, are a style choice of the
publisher's graphic designer, and don't affect the information content of
the title page -- if they do (as in interlaced parallel titles with words
common between languages) special account may have to be taken of them.
>The concept of providing a surrogate for the title page is a useful
>one, IMO, but the "sorta/kinda/almost" way that it is done in
>cataloging makes it less useful.
Not much less -- for textual works, anyway.
Ed Jones <[log in to unmask]> points to a notion that I too have considered:
>Would transcription still be necessary if a title page (or analogous source
for other types of resource) image were routinely included along with the
metadata as part of a catalog record "package"?
Not all conveyance of bibliographic information is web-based, nor is it
carried out by experts. Therefore I think the transcription-based rendering
of the most essential recognition elements (author/creator/etc., title,
place, publisher, date, edition if not first, ISBN -- and equivalent for
other than textual resources) remains necessary -- we can't count on having
a linked t.p./cover image always associated with these essential recognition
elements.
Note, as an aside, that "place" (of publication) was omitted from RDA "bare"
core elements; I disagree with this omission, on the grounds that I continue
to read recently-released, substantial books which persist in giving place
but not publiher's name in the bibliography.
Hal Cain (actively cataloguing till 2 months ago)
Melbourne, Australia [who says major jurisdiction isn't needed for
recognition of place?]
[log in to unmask]
|