On 8/6/2015 3:57 PM, McDonald, Stephen wrote:
> The system that James Weinheimer described does not confuse > different people with the same name and does recognize individuals > with multiple or changing forms of names. There will still be unique > identifiers, but we won’t have to agonize over using_names_ as > unique identifiers.
> What James was saying is that, with such a system, you do not need a > single authorized form of the name as the unique identifier for that > person. There will still be a unique identifier, but it need not be > in the form of a unique name. The authority record will contain some > information about the person, and a unique identifier. It will also > contain links to other sources of information. Those links will > basically say “The person with identifier XYZ in this catalog is same > as the person with the unique identifier 123 in wikipedia, unique > identifier ABC in ORCID, the unique identifier 99999 in VIAF, the > unique identifier QWERTY in Who’s Who in America, and the unique > identifier 1A9Z in the University of Nowhere faculty database”. The > unique identifiers will probably never display; there is no need to > display them. You never see the unique identifier for Wikipedia > entries, and you don’t need to. The system handles it, and displays > disambiguation links and identifying information as needed. The > unique identifier will probably be a fairly arbitrary alphanumeric > string.
Thanks for clarifying what I was trying to say. Headings have always served a double duty in that they fulfill two very different, and very important functions. The first function is to serve as a collating device, bringing all “instances” of a concept together. This is how it worked in a card catalog, such as all cards describing items by Peter Tchaikowski, or in a printed catalog, which just brings the printed records together. In early computer catalogs, the textual string did the same thing.
The second purpose of a heading is to provide a label by which something may be found. In physical catalogs, it was absolutely necessary to choose a single “authorized form” where all the cards or printed records could be found. Of course, it would have been too much work–and too crazy–to duplicate the entire card file found under the “authorized form” so that all records could appear under each variant, and the final product would have been too unwieldy. Instead of doing that, catalogers would insert a single card, or reference, that said See: Tchaikovsky, Peter Ilich, 1840-1893.
It’s easy to merge the two different functions in our minds so that the collating function = the label.
But with the introduction of relational databases, these two functions *can* be separated although they often haven’t been. The collating function is performed by an internal primary key (almost always an arbitrary number) that takes the place of the text string. Therefore, in a relational database if you search for a specific heading, you are *really* searching the internal database number, but you aren’t aware of it. To see it in action, here is a record from the Deutsche National Bibliothek, http://bit.ly/1KSU0Qq and when you see the editors’ names, don’t click, just run the mouse over them, and you will see that what looks like e.g. Messent, Peter is actually a link to a local number “idn=D187641625”.
The label itself can be displayed from another table and it can be done in various ways, including any of the other information in the table. Therefore, it could include the 4xxs or 670’s, although in the library catalogs I have seen, they display only the traditional “authorized heading” and nothing more.
With linked data, this basic method continues except the computer is not limited to importing only from an internal table in the database, but it can come from any other open RDF resource, from dbpedia, wikidata, or whatever.
Once again, this is not 100% good and there are definitely bad sides to it, but it seems as if we are headed in the direction no matter what. It also seems to me that it would be wise to prepare for various outcomes.
James Weinheimer email@example.com
First Thus https://blog.jweinheimer.net
First Thus Facebook Page https://www.facebook.com/FirstThus
Personal Facebook Page https://www.facebook.com/james.weinheimer.35 Google+ https://plus.google.com/u/0/+JamesWeinheimer
Cooperative Cataloging Rules
Cataloging Matters Podcasts
The Library Herald http://libnews.jweinheimer.net/