Thank you, Kevin and Bob, for those very enlightening answers.  I knew 
of course that the AAP for a work represents all expressions of that 
work, and I can see why it is critical in subject searching to have the 
other-language variants of the work title in the work record, but I 
guess I haven't encountered this before in anything besides musical works.

I was wondering why there were no authority records for language 
expressions of Da xue, but it occurs to me now that if the 
other-language variant access point is already in the work record, maybe 
there's no need to create an authority record for the expression even if 
you use the expression access point in a bib record.

Is that correct? If I get an English-language expression of Zhong yong, 
and the title on the manifestation is "The constant mean," I should add 
<430 _0 Constant mean> to the work record and put <130 _0 Zhong yong. $l 
English. $f 2017> in the bib record, and not create an expression-level 
authority record for the new English-language expression with a SEE 
reference from "Constant mean"?  Or establishing the language expression 
in LC/NAF is optional?

I ask because establishing the language expression seems to have been 
carefully avoided over many years in the case of Da xue.  I don't think 
there's any problem having two records with same 430 as far as the NACO 
file comparison rules are concerned, but is there something somewhere 
else that instructs not to create an expression-level record with a SEE 
reference that matches one in that expression's work record?

Thank you again!

Michael A. Chopey
Catalog Librarian
University of Hawaii at Manoa Libraries
Hamilton 008
Honolulu, HI  96822

phone (808) 956-2753
fax (808) 956-5968

On 12/1/2017 12:52 PM, Robert Maxwell wrote:
> Michael,
> In RDA there isn’t actually any such thing as a variant title of an 
> expression (there’s no title element at the expression level) so 
> technically speaking all variants are variant titles of the /work/ 
> (RDA 6.2.3), and as such could appropriately be included in the 
> authority record representing the work. This makes theoretical sense 
> because the authorized access point for the work, while it is usually 
> based on a form in first language expression of the work, it does not 
> represent any particular expression, it represents /all/ expressions 
> of the work. “130  0 Da xue” does not represent the work in Chinese, 
> it represents all expressions of the work. It is the authorized access 
> point for the /work/, not for any particular expression of the work.
> Note that the authorized access point for the work is based on the 
> preferred title of the work. The preferred title is /usually/ based on 
> a form in the language of the original expression, but this is not 
> always the case. Preferred titles for classical and Byzantine Greek 
> works, for example, use an English or Latin form rather than a Greek 
> form, if possible (; similarly, the preferred title for a 
> liturgical work is “a well-established title … in a language preferred 
> by the agency creating the data”, i.e. English for PCC ( 
> Preferred titles for books of the Bible are English in PCC practice. 
> Preferred titles for musical works are often in English even though 
> the original title of the piece was in another language. Etc.—there 
> are many many examples of this practice. I think you’ll find that 
> authority records representing the work in all of these cases contain 
> a mix of linguistic variants, at least a variant title from the 
> original language to the English preferred title.
> I point out that these instructions for choice of preferred title 
> aren’t different in any major way from the rules for choosing the 
> title in AACR2.
> RDA does have the concept of a “variant title associated with an 
> expression” (, which allows such variants to be included in 
> authority records representing the expression. Again, this was also 
> the practice in AACR2.
> All that said, I generally don’t include variant titles associated 
> with an expression in the authority description of a /work/ unless 
> they’re in the same language as the preferred title, or if the 
> preferred title is not in the language of the original expression (in 
> which case I usually include a variant in the language of the original 
> expression). I think this is a fairly common practice, but it’s only a 
> practice; it’s not required by RDA.
> Bob
> Robert L. Maxwell
> Ancient Languages and Special Collections Librarian
> 6728 Harold B. Lee Library
> Brigham Young University
> Provo, UT 84602
> (801)422-5568
> *From:*Program for Cooperative Cataloging 
> [mailto:[log in to unmask]] *On Behalf Of *Michael Chopey
> *Sent:* Friday, December 1, 2017 3:16 PM
> *To:* [log in to unmask]
> *Subject:* Question about two work records: Da xue (n82031679) and 
> Zhong yong (n79068391)
> Dear PCC colleagues,
> Does anyone know why these two work records have SEE references from 
> various other-language expressions of the work? I can't find the rule 
> that justifies doing this.
> e.g.:
> 130  0 Da xue
> 430  0 Taehak
> 430  0 Daigaku
> 430  0 Velikai︠a︡ nauka
> 430  0 Great learning
> 430  0 Grande étude
> etc., etc.
> 130  0 Zhong yong
> 430  0 Chungyong
> 430  0 Chūyō
> 430  0 Nauk o sredini
> 430  0 Doctrine of the mean
> 430  0 Tchoung ioung
> etc. etc.
> As you might know, these are two of the "Four books" (Si shu; 四書), 
> Chinese classics of Confucian thought.  The other two are attributed 
> to Confucius and to Mencius, and each of them is entered under the 
> creator's name heading. The one by Confucius represents only the work 
> and its original Chinese-language expression, as I would expect, and 
> other-language expressions of that work are entered variously as:
> 100 0  Confucius. ǂt Lun yu. ǂl English
> 100 0  Confucius. ǂt Lun yu. ǂl Japanese
> 100 0  Confucius. ǂt Lun yu. ǂl Indonesian
> etc. etc.
> The one by Mencius is a mixture of the two different approaches. The 
> work record seems to represent the original Chinese-language 
> expression and also an English-language expression and a 
> Japanese-language expression:
> 100 0  Mencius. ǂt Mengzi
> 100 0  Mencius. ǂt Sayings of Mencius
> 100 0  Mencius. ǂt Mōshi
> but there are also other-language expression records, as I would 
> expect to find, e.g.:
> 100 0  Mencius. ǂt Mengzi. ǂl English
> 100 0  Mencius. ǂt Mengzi. ǂl Japanese
> 100 0  Mencius. ǂt Mengzi. ǂl Korean
> 100 0  Mencius. ǂt Mengzi. ǂl Russian
> 100 0  Mencius. ǂt Mengzi. ǂl Latin
> 100 0  Mencius. ǂt Mengzi. ǂl Portuguese
> etc., etc.
> Can someone explain what is going with these?  Three of the four work 
> record are RDA, and the other is coded AACR, but I can't remember any 
> reason why you wouldn't have established the other-language 
> expressions separately in AACR either.
> Thank you in advance if someone can show me what I'm missing!
> Aloha,
> Mike Chopey
> Michael A. Chopey
> Catalog Librarian
> University of Hawaii at Manoa Libraries
> Hamilton 008
> Honolulu, HI  96822
> phone (808) 956-2753
> fax (808) 956-5968