Brenndorfer, Thomas wrote:
Perhaps it would have been better to use an example from Codex Alimentarius that resembled the textual properties displayed on bibliographic resources which catalogers must take into account in assisting people in identifying those resources. The General Standard for the Labelling of Prepackaged Foods (http://www.codexalimentarius.net/download/standards/32/CXS_001e.pdf) prescribes a series of instructions for recording the the name of the food that is no less onerous than the rules for bibliographic description in libraries:
4.1 The name of the food
4.1.1 The name shall indicate the true nature of the food and normally be specific and not generic:
18.104.22.168 Where a name or names have been established for a food in a Codex standard, at least one of these names shall be used.
22.214.171.124 In other cases, the name prescribed by national legislation shall be used.
126.96.36.199 In the absence of any such name, either a common or usual name existing by common usage as an appropriate descriptive term which was not misleading or confusing to the consumer shall be used.…
Thanks for pointing that out. This is a much better example of what I have in mind. For example, I can imagine that determining a *precise form* of a named entity may become less important as URIs begin to be implemented and displays of names become more fluid. Still, I can imagine a highly predictable form that would, in a sense, "guarantee" access to the name for librarians; in other words, an "expert form" of the name and that could continue current AACR2-type practices more or less.
Of course, the same methods could work for subjects as well, and perhaps better. So, if we have a form of subject that really no one would ever think of, e.g. "Byron, George Gordon Byron, Baron, 1788-1824--Homes and haunts--England--London", this would not necessarily be the first thing displayed and it could be something more like "Lord Byron and British pubs"! ☺