Dear PCC members,

I would like to make an announcement of a change in the CJK NACO Best Practices that LC approved recently.  This new practice might be of interest to any member who is creating and using the authority record as well as bibliographic records for user services.  This change is most relevant for members who catalog CJK language materials. 

Catalogers now can input non-MARC 8 characters using an automatic machine convertible method with the Numeric Character Reference (NCR) in the authority records until the full Unicode character set is supported in NACO records. 

Catalogers have been able to input any Unicode characters in the bibliographic records in OCLC and most library systems have had a capacity to handle them for many years.  However, inputting non-Latin characters in the authority records are still limited to MARC 8 characters only.  CJK Catalogers have been inputting non- MARC 8 characters in NACO records in various ways; using a geta symbol, inputting romanization in brackets, or optionally, inserting Unicode information of the non-MARC 8 character.

Using the NCR notation is a lossless technique referenced in MARC 21 documentation for character sets in MARC bibliographic and authority records.  The MARC 21 standard allows/recommends this lossless technique from 2006. The notation does look cryptic, but it will make automatic conversion possible with the UTF16 value in the future.  A detailed description is available at:

The new CJK best practice on this change is being prepared and will be posted at CJK NACO Best Practices site sometime in February.

When it is posted, I will notify you in case you want to read more in detail. 


Erica Chang

University of Hawaii Library