would it work to view what we hardened catalogers
call "name-authority files" in this way?
a. users of a catalog of metadata may call a given person or
agency whatever they want, AS LONG AS it's
one of the cross references in the name-authority
file for that person. OR it's a combination of
words in one of the cross references, just rearranged.
Furthermore, the users don't need to know
what the authority form of the name is - they
just need to get a list of all works by a given
person or agency.
b. only the metaloger/cataloger has to use exactly
the same form of the name every time. this is
essential, so that metaloging/cataloging is consistent.
I'll use myself for an example. Here's the LC
name-authority record for my name:
010: : |a n 77019511
040: : |a DLC |c DLC
100:10: |a Larsgaard, Mary Lynette, |d 1946-
400/1:10: |a Larsgaard, Mary, |d 1946-
670/1: : |a Her Map librarianship, 1978: |b t.p. (Mary Larsgaard, Central
ash. St. Col.) CIP data sheet (b. 8/4/46)
670/2: : |a Her Topographic mapping of the Americas, Australia, and New
ealand, 1984: |b CIP t.p. (Mary Lynette Larsgaard)
A user should be able to find my name by inputting
any arrangement of the words in the name:
- larsgaard mary lynette
- mary larsgaard
- larsgaard mary
- mary lynette larsgaard
and so on.
It does mean that a metaloger has to find an
authority file/list to accept. The same thing
is true with subject headings.
Mary Larsgaard
(alias 77019511)
At 09:18 AM 7/22/99 -0400, James Weinheimer wrote:
>I guess I have to wade into this one! Bernhard, where are you??
>
>One of the problems here is what we mean by "right" and "wrong". Simply put,
>in a catalog,
>what is "right" is based on precedent--what has been done in the past. For
>example, Mr.
>Fausto di Biase may not like the form of his name, but if all we do is put
>the form: "di
>Biase, Fausto" on all the new records, it is--by definition--"wrong". In
>order to make the
>situation "right", we must update all the former "Biase, Fausto di" records
>to the new form.
>We must think of the catalog as a whole and not only of individual records.
>
>In short, the catalog is not there to make people happy--it's there for
>people to find
>things, and the best way for people to do this is to have a single,
>consistent, form of the
>name, be it personal, corporate, geographic, title, or subject. Some of
>these forms may seem
>ridiculous, e.g. "United States--Politics and government--To 1775" before
>there was a United
>States. But for the sake of consistency, we don't worry about such logical
>conundrums.
>
>This is why I have problems with the "quick and dirty" solution. Let's say
>that the form is
>"Biase, Fausto di" but somebody doesn't like it and thinks "I'll make it 'di
>Biase,
>Fausto'". Somebody else might say "Fausto di Biase", and who knows what
>somebody else will
>want.
>What if somebody prefers to make it: "Great Britain--Colonies--North
>America--Politics and
>government--To 1775"? Sure--someone may find any of these forms, but how
>will they know of
>the others?
>
>If metadata creators can do whatever they want without having to tell
>anybody else, or do any
>sort of maintenance, there wil be nothing but chaos in the database, which
>will be only a
>slight step above what exists today in Alta Vista.
>In this way, "quick and dirty" to me, means making everything dirty--a
>synonym for
>"impossibility of finding works in the catalog".
>
>I believe--if we are going to promote DC as a genuine solution--we had
>better make sure it's
>a solution that we truly believe will work, because when it doesn't work,
>our credibility
>will go down the toilet. Our training lets us know that quick and dirty
>can't work in the
>long run, so we must speak out.
>
>I think there are solutions to this problem, so all is not lost.
> Jim Weinheimer
> Princeton University
> [log in to unmask]
>
>
>> > I am simply worried that without a strong set of rules DC will lead
>nowhere...
>>
>> I would guess that James Weinheimer and Bernhard Eversberg have something
>to say
>> here ;-)
>>
>> Best regards,
>> - Michael
>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|