Dear Heidi,

I was the chair of Access Points for Expressions Task Group, and I agree 
that this is a complicated and confusing area. And while I can't comment 
on what actions PCC will ultimately take in response to our 
recommendations, I hope I can at least provide some context behind our 
report and continue the discussion.

Your question about conventional collective titles beginning with 
"Works," did come up in the course of our task group's work, but we 
didn't address it in a conclusive way in our final report. This is in 
part because at the outset of our work, the section of LC-PCCPS 6.27.3 
which required LC catalogers to distinguish "Works" in this way (as was 
the case under LCRI 25.8) was removed in the August 2012 release at the 
request of the PCC 
( about midway 
through our task group's timeline.

As I understand it, this change was in response to the The RDA Policy 
Task Group, which said the following about it in its final report (April 
2012, p. 81): TG recognizes that this has been the practice for years, 
but sees no good reason to continue it. Objections: Adding the date 
anticipates a conflict that may never happen, a practice that LCPS disallows in the case of access points representing works. Date 
is not a particularly useful way to distinguish collected works; users 
will often know the edition but not the date, so that seems a more 
useful addition. In practice the instruction causes catalogers not to 
add the date of the expression to the authorized access point, but the 
date of the manifestation. Finally, it is bizarre that same LCPS insists 
upon adding a differentiating qualifier in all cases, even when in most 
of those cases there is only one expression (Works) but disallows 
qualifying in cases where there are more than one conflicting 
expressions (multiple French translations of Hamlet).

And while I generally agreed with this, based on feedback and questions 
about our report, it was clear to me that many catalogers and users 
continue to find the practice of adding dates useful in the current 
environment and wanted additional guidance about how proceed in the 
absence of the LCRIs and the deleted section of LC-PCCPS 6.27.3.

I understood the removal of this requirement to mean that LC catalogers 
would no longer be required to add a date to the access point "Works" in 
all cases. The LC training modules further instruct not to add a date to 
the access point in the absence of conflict, but elsewhere say that 
dates and other attributes may optionally still be recorded separately 
if readily available. I take this to mean that they WOULD continue to 
add a date in the presence of conflict. Thus, the first-cataloged 
instance of a creator's collected works would be represented by "Works" 
without a date in the access point (though it could still be recorded as 
a separate element), but further instances would include a date 
(presuming later editions of the same author's complete works 
constitutes a conflict).

I don't know that this accurately summarizes LC's practice, and of 
course what this all means for PCC catalogers is the question at hand.

The DCM Z1 intro states: "PCC practice:NACO participants may contribute 
name authority records for works or expressions as needed for 
cataloging." And what remains of LC-PCCPS 6.27.3 also acknowledges the 
possibility of NARs for expressions existing with elements LC would not 
have included, instructing LC catalogers to use these records when 
found. Within these statements, we generally saw room for PCC catalogers 
to use cataloger's judgment to include dates (or other attributes) to 
the access point to resolve conflict, even if LC would not have done so, 
I don't think our task group necessarily wanted to encourage adding 
unnecessary attributes in the absence of conflict (or omitting necessary 
ones in the presence of conflict), and that any optional or additional 
elements should of course be formulated following LC-PCCPS guidelines 
for use of controlled vocabularies, codes, etc. Thus, if choosing to 
include dates, we did recommend following the established practices of 
using Gregorian dates in the access points, and following appropriate 
ISO or EDTF standards to represent dates of works/expressions in 046 $k 
and $l.

I also take this to mean "[Creator]. Works. [Date]" is acceptable for 
the purposes of recoding and using records previously established as 
such under AACR2 (provided the creator's name is also RDA compatible), 
just as we accept "RDA compatible" elements such as fuller forms of 
names etc., as found in other access points established under AACR2 even 
when current practice would not have required them in the absence of 
conflict. I believe the NACO Phase 2 revisions flipped many records 
beginning with "Selections" to "Works. Selections" as well, some of 
which include dates and some of which don't. So it seems likely that we 
will see "Works..." with and without dates in NAF and bibliographic 
records. No matter what, access points for works and expressions raise 
authority and bibliographic file maintenance problems much thornier than 
those for names, and what will emerge as "best practice" is still 
somewhat of an open question in my mind.



Matthew C. Haugen
Rare Book Cataloger
102 Butler Library
Columbia University Libraries
E-mail: [log in to unmask]
Phone: 212-851-2451

On 6/7/2013 2:22 PM, Heidi G Lerner wrote:
> I would like to know how PCC libraries should handle the creation of 
> access points for expressions in the absence of a final report from 
> the “ PCC Access Points for Expressions Task Group”. We have been 
> advised by PCC that these reports contain the recommendations of the 
> Task Group alone and should not be used until such time as the final 
> document is posted as an official PCC document or a LC/PCCPS.
> The guidelines in the LC-PCC PS 6.27.3 reflect LC policy only. I am 
> currently a BIBCO RDA devisor and also create PCC records and need 
> some help figuring out how to crate access points for expressions that 
> are not covered by the guidelines in 6.27.3.
> An example is the creation of an access point “[Name of 
> author].$tWorks.$f2013. According to LC-PCC PS 6.27.3 we would not add 
> an expression attribute since these are only applied in the cases of 
> music resources, sacred scriptures, translations, and language 
> editions. I would like some help in deciding if this expression is 
> permissible for PCC libraries or if we should not be adding the date.
> Another issue that I am not sure what to do with is how to handle the 
> existence of other NARs for the same “work” and have the addition of a 
> date that are coded AACR2.
> Thank you, Heidi Lerner
> -- 
> Heidi G. Lerner
> Hebraica/Judaica Cataloger
> Metadata Development Unit
> Stanford University Libraries
> Stanford, CA 94305-6004
> e-mail: [log in to unmask]
> ph: 650-725-9953
> fax: 650-725-1120