After doing some searching of the OCLC WorldCat database, I can say that my
experience matches Millie's.  Most of the records with roh are for items in
Romansh with fewer than 20 that appear to be Ladin.  So, I agree that
Peter's suggestion that rm/roh be treated as the identifier for Romansh
seems reasonable.

--Glenn

-----Original Message-----
From: Milicent K Wewerka [mailto:[log in to unmask]]
Sent: Monday, March 22, 2004 9:44 AM
To: [log in to unmask]
Subject: [ISOJAC] Rhaeto-Romance


To: ISO-JAC
From: Milicent Wewerka, Library of Congress

Peter Constable asked for comment on the usage of the language code for
Rhaeto-Romance.

I checked past usage for the MARC Code List for Languages.  It is clear
from past editions that the MARC usage for [roh] was intended to cover
both Ladin and Romansh as well as Friulian before it had its own
identifier.

I checked records in the Library of Congress database and find that
most of the record with [roh] are actually Romansh, although there are
also some Ladin texts.

If we consider past practice only, then it would be better from the
library standpoint to keep [roh] as a more comprehensive identifier.
However, it is possible that we may at some future time wish to
implement a separate identifier for Ladin in ISO639-2.  In that case
[roh] would be a collective designation containing only one language, a
situation that seems undesirable.

I think Peter's recommendation that [rm/roh] be treated as the
identifier for Romansh is an acceptable solution.  For MARC usage, this
would mean that Ladin would fall into the category of "Romance
(Other)."

I would like to hear the views of others on this, especially those who
are active in the library community.

By the way, I found that the term "ladin" was also used to refer to
Romansh and Friulian.