Ian, I'm still confused about why you want both a work AAP and an expression AAP in your bib record.  I notice you have used second indicator 2 for the expression AAP and not for the work AAP, but that isn't helping me understand.

Are you suggesting this so that the records can simultaneously and separately be collocated by work but differentiated by expression?
That is the only thing I can think of.

A book doesn't contain a work, it contains an expression of a work.  If there is only one such expression, we currently put the title part of its AAP in a 240, assuming the title proper won't do double duty.  We could move the AAP to a 7xx field, as you suggest, though since right now a name/title 7xx field is understood to be for a related work (often analytical), it would be something of a conceptual shift to put the work/expression that composes the *entirety* of the manifestation into a 7xx name/title access point.   (We'd be edging into RDA chapter 17, maybe?)   But when you suggest making a separate access point for the basic *work*, you're doing something brand new, not just using different fields, and I'm not sure why.

"Poems. Selections" has the same qualities as "Works. Selections."  If qualified with a year or something else, it refers to one particular collection; if not qualified, it really represents ALL aggregate works containing selections from a person, as Robert Maxwell said.  "Poems. Selections. 2014" is a work and an expression, but an expression only of its work-self. "Poems. Selections," unless there's only been one such collection, is a sort of chimerical super-work, though we can *pretend* it's a work to simplify cataloging.  In any case the former is not an expression of the latter.

Actually, translations of works with normal title-like titles are the same way.  LC doesn't make expression AAPs to distinguish one translation from another in the same language, but this also results in AAPs that correspond to chimerical super-works.  This:

Dobbs, Bob. $t Gospel according to Slack. $l Spanish (Jones)

is not an expression of the work

Dobbs, Bob. $t Gospel according to Slack. $l Spanish.

The latter may or may not represent all Spanish translations of the book.  It depends on whether there have been more than one.  It would certainly be possibly to pre-emptively qualify such a heading even if there are currently no other Spanish translations, but LC definitely doesn't do that.

Pete Wilson

From: Program for Cooperative Cataloging [mailto:[log in to unmask]] On Behalf Of Ian Fairclough
Sent: Friday, October 10, 2014 8:19 AM
To: [log in to unmask]
Subject: Re: [PCCLIST] use of field 240

PCCLIST readers,

Thanks again to those who've contributed to this discussion.  The most recent post is Robert Maxwell's.  His book, Maxwell's Handbook for RDA, contains the phrase "format agnostic" (page viii), about which I asked a while ago (post to RDA-L, June 2, 2014).  The phrase is germane to the discussion, which I'd like to refocus back onto the use of MARC encoding, and in more general terms rather than specifically as to whether "Works. Selections" represents a work or expression, with or without various qualifiers.  My concern applies also to cases like "Poems. Selections" and "Preferred title. Language into which it is translated".  Setting such concerns aside for the moment (if possible!) and generalizing, I wonder whether:

100  AAP for author, |e author.
240  Preferred title for work, qualified by data specific to the expression
245  Title proper (as transcribed from preferred source).

is a practice that should be replaced by:

100  AAP for author, |e author.
245  Title transcribed from preferred source.
700 1_ AAP for author. |t Preferred title for work.
700 12  AAP for author. |t Preferred title for work, qualified by data specific to the expression.

Responding to Ted Gemberling's earlier question "Why would that be better than using the 240?":

If you concatenate the data in 100/240 as presented above, you get
AAP, author.  Preferred title for work
The relationship designator is obviously out of place in a field that represents the work or expression.  And the data is performing double duty, in these respects:
1. Field 100 is an access point qualified by the relationship term in subfield e, as well as a linking term for the preferred title;
2. Field 240 is a preferred title for work, however, it is also serving as preferred title for an expression of that work.  Particularly if a language into which the work has been translated is given.

Although use of MARC 700 duplicates the AAP for author data, it avoids having the 100 and 240 fields do double duty.  Despite the duplication, perhaps our records would be better if each field has just one function.

If I'm mistaken I'd appreciate correction.  In practical terms: is there any reason why not to prepare MARC records according to the second way?  (In OCLC Connexion, it has the added advantage that 700 fields with subfield t can be controlled, whereas field 240 cannot.)

Sincerely - Ian

P.S. Recording for reference purposes: Applicable guidelines are
100: RDA 9.19 & I.1, plus LC-PCC PS link to the document PCC Guidelines for the Application of Relationship Designators in Bibliographic Records
240: RDA 5.5
245: RDA 2.3.2

Ian Fairclough
Cataloging and Metadata Services Librarian
George Mason University
[log in to unmask]<mailto:[log in to unmask]>