04. Multiscript records (under construction)
For records containing the original (non-Latin) script and a transliteration in Latin script, tag 880 Alternate Graphic Representation is not used in the CBS MARC 21 cataloguing format. Tag 880 is not defined nor allowed in a CBS MARC 21 database. The reason is that has been chosen for another "philosophy". Tags need to be “duplicated“, e.g. in the first tag M245 Title Statement, the original script is added and in the second tag M245 the transliteration of this tag. MARC subfield $6 Linkage is used to distinguish both tags. So, in the CBS MARC cataloguing format, tag 880 is not allowed.
During the import of MARC records into a CBS MARC 21 database, the 880 tags in the original records need to be split up, and during the MARC export of records from a CBS MARC 21 database, e.g. to WorldCat, the 880 tags need to be generated.
The actual situation is that the 880 tags in the MARC export contain the non-Latin script; the tags with the Latin script are converted to the non-880 tags. But this is only a choice. Other options are possible in principle. (under construction)
There are two flavors of the MARC 21 export: the OCLC MARC export and the LoC (Library of Congress) MARC export. In general, they are the same, but there are small differences.
So, in a CBS the non-Latin, original scripts are considered the most important, so tags with those scripts are mentioned first. The transliteration is considered additional or does not even exist, so those tags are mentioned second. (under construction)
For the script codes1 we use ISO 15924. On the Library of Congress website, you can find the “Script Code and Term Source Codes“ under “Source Codes for Vocabularies, Rules, and Schemes“. The introduction says: “Script Code and Term Sources contains a list of standard code and/or term sources for specifying the scripts used in a resource and assigns a code to each source. The purpose of this list is to enable the script vocabulary used in metadata records to be identified by a code.” The list that is referred to is ISO 15924: http://www.unicode.org/iso15924/.
Extra subfield for languages? And an extra subfield for transliteration standard?
However, a script can be used in a different way for several languages. E.g. Cyrillic is used for a.o. Russian, Belarusian and Ukrainian, so the (automatic) transliteration into Latin script can differ per language. Maybe, we need to define an optional separate subfield, besides subfield $6, for the language.
Is it useful to add a subfield for a language?There can be several ways of transliterating a script. The English way is different from e.g. the German (DIN), Dutch or French (AFNOR) practice. And per language too, it is possible that there are variants. So, is it useful to add a subfield for the transliteration standard?