<html>
<body>
I will also throw my hat in with those who see $5 as potentially useful
for all 6xx not just 655. <br><br>
The question of whether to allow $5 in situations that are not strictly
copy-specific is a little trickier but since the current MARC21
definition for the subfield does allow it and there do seem to be
situations when it could be useful, I think I would also lean towards
allowing it. Seems like we can break down 6xx/7xx headings in special
collections that could make use of $5 as follows:<br><br>
Headings that are clearly "copy-specific":<br>
-- 6xx (e.g. what is "depicted" in a fore-edge painting or what
a tipped-in ms. letter is "about")<br>
-- 655 (e.g. the fact that a particular copy includes something that
"is" a fore-edge painting or "is" a tipped-in ms.
letter)<br>
-- 7xx (e.g. who executed the fore-edge painting or who wrote the
tipped-in letter)<br><br>
Headings that are of primarily "institutional interest":<br>
-- 6xx (a level of subject indexing that is deeper than what most
institutions would find acceptable or desirable, e.g.
"African-Americans$zVirginia$vFiction" to describe a novel
featuring a single African-American character who appears in only 2 out
of 200 pages - because that library has a "research collection"
of African-American materials)<br>
-- 655 (a level of genre indexing that is deeper than what most
institutions would find acceptable or desirable, e.g.
"Limericks" added to a book of 200 pages that contains only a
single limerick - because that library collects books with
limericks)<br>
-- 7xx (e.g., a name-title analytic for a single poem by John Betjeman in
an anthology containing poems by 100 other authors - because that library
has a comprehensive Betjeman collection)<br><br>
The decision on how exactly to tag such headings would be influenced by
such things as:<br>
1. Whether the headings are constructed in accordance with a standard
source or non-standard local conventions or some combination of the
two.<br>
2. For 6xx that use a standard source, whether that source is identified
using an indicator value or a combination of indicator plus $2<br>
3. How your local system handles such things as indexing, display,
hyperlinking, and heading validation (e.g. Yale uses 69x but they do NOT
validate and we had to create a separate index for them)<br>
4. Whether your records must "play well" with other records in
your local system (e.g. non-special collections records or special
collections records from other units)<br>
5. Whether you use an authorities vendor (or might in the future) and how
the vendor does its heading matching and processing (e.g. Yale had to
customize 69x processing).<br>
6. Whether you contribute records to a bibliographic utility (or
utilities) or might in the future and how the utilities handle the
data<br>
7. Past practice at your institution<br><br>
None of this seems directly relevant to the MARBI Discussion Paper that
John brought to our attention, since that DP is proposing using 6xx $5 in
a different context. I think our reasons for possibly adding 6xx $5 are
actually better than theirs (though, like Bob, I'm not 100% sure I
understand theirs)!<br><br>
-Manon<br><br>
<br>
<blockquote type=cite class=cite cite=""><font size=3>However, in this
case, I am going to voice an opinion and agree with Nina and Richard on
the 650 score, largely because MARC21, as Nina points out, directly says
using a subfield 5 is permissible under the defined circumstances. For
those of us who load records into OCLC, this is in fact why the subfield
5 is so important, as it is a way to strip local headings out of exported
records.</blockquote>
<x-sigsep><p></x-sigsep>
____________________________________________<br><br>
Manon Théroux<br>
Authority Control Librarian<br>
Catalog Department<br>
Yale University Library <br>
P.O. Box 208240<br>
New Haven, CT 06520-8240
<br><br>
203-432-8376 (tel)<br>
203-432-7231 (fax)<br>
manon.theroux@yale.edu</font></body>
</html>