Jim, you completely missed my point.
We have an authority record for "Southhampton Insurrection, 1831":
http://lcsubjects.org/subjects/sh85125604#concept [1]
Which includes alternate terms:
"Nat Turner's Insurrection"
and
"Turner's Negro Insurrection, 1831"
Now, if the user happened to match "Nat Turner's Insurrection", why
bother to "correct" them that the "real" subject they're looking for
is "Southhampton Insurrection, 1831" when the computer knows damn well
that they're the same thing?
I'm not asking that this be translated into every language worldwide,
I'm just asking that authorities do what they're supposed to and not
require more (useless) navigation from the user.
-Ross.
1. I'm using lcsubjects.org because id.loc.gov doesn't show alternate
labels well.
On Tue, May 5, 2009 at 9:51 AM, Weinheimer Jim <j.weinheimer_at_aur.edu> wrote:
> Ross Singer wrote:
>
>> No, he's saying we have an authority record which knows all of the
>> terms for Nat Turner's Rebellion.
>>
>> Why should the user know or care what the LCSH preferred term is? The
>> records should be indiscriminate because they are referring to an
>> authority.
>>
>> What Tim is saying is that if the search is restricted to what appears
>> in the 650 tag then we have have completely forgone the advantages of
>> moving it from an index card to a computer.
>
> "An authority record which knows *all of the terms* for Nat Turner's Rebellion."
> Wow! That would be one h*** of a record! How many languages would we limit it to? And besides, I'm sure there were lots of contemporary ways of relating to his rebellion that we would think are not very nice at all!
>
> It seems that making that kind of an authority record would be absolutely impossible and only an exercise in futility.
>
> While I agree that the idea of a "preferred term" has definite problems, the underlying function of the heading: to collocate variant forms, still holds. Choosing a single display (preferred form) for the headings is the problem, however, but I suspect it may be as simple as allowing all "4xx See" references to display equally with the heading, and have a display (as I have mentioned somewhere) such as Thomas Hyde had in his catalog of the Bodleian Library back in the 1500s or 1600s, which showed all variants simultaneously with the inclusion of "seu" (or).
> "Abelard, Peter, seu, Abelardus, Petrus, [followed by the others]".
>
> Today, we could allow the users to decide which each wants. (I have a suspicion nobody will care that much)
>
> Now that the authority file is being made available for experimentation, I hope someone tries something like this.
>
> Jim Weinheimer
>
Received on Tue May 05 2009 - 10:24:59 EDT