Nancy and others, Of course I work at LC and you're not going to hear anything but party line from me--I strongly agree that variants of the name of a body should appear in the NAR for the body, variants for the name of a sub-body should appear in the NAR for the sub-body, and so on. And this does result in mixing roman and nonroman in the reference.
This situation is (in my opinion) an inevitable development of the rule Nancy cites, that in references containing such subfields, the $a subfield should be in catalog-entry form--and all catalog-entry forms are roman. What I don't understand is the attraction of doing otherwise. Why should these references be constructed so that all, or anyway more than just the last, subfields are not in catalog entry form? The nonroman references for all the bodies will appear in the authority file either way. Is it an esthetic desire that roman and nonroman shouldn't appear in the same reference? Joan >>> Nancy Sack <s...@hawaii.edu> 06/26/09 5:48 PM >>> Aloha, safranim ve-safraniyot, Could I ask you to weigh in on a debate that has been raging in our department for the past few days? Apparently there are differences of opinion regarding the nonroman exception in RI 26.1: Until practices related to the form and style of nonroman script references have been established, it is not necessary to construct nonroman references in the same form as the heading. Some catalogers believe that this exception applies to the restriction on making references only at the level of the heading established, and are giving complete original-script references for corporate body-subordinate body strings and author-title strings. Do you understand the exception to exempt nonroman references from the guidelines that appear, for example, on the NACO corporate body FAQ ( http://www.loc.gov/catdir/pcc/naco/corpfaq.html ): This LC/PCC policy is a parameter for contribution of authority records to which all NACO participants agree. Once a heading is established it is a violation of this policy to make cross references from unauthorized forms of the elements in a hierarchical string. The foreign language or variant form of a parent body is included as a cross reference only on the NAR for the parent. Each additional subordinate unit must then use the established forms of the heading even though the subordinate unit variant can be given in the foreign language. LCRI 26.1 Forms of References statement: "... construct a reference in the same form in which it would be constructed if chosen as the heading ..." is interpreted by LC/PCC policy to reinforce the principle of not creating a cross-reference for a subordinate unit with unauthorized forms of the parent units. FWIW, I'm convinced that it's better not to mix romanizations and original scripts but I've assumed that the current guidelines demand that we do just that. Thanks. Nancy
