Stephen Hearn wrote [in part] :
[deletia]
>
> The more complicated question is whether any fixed "legible" form for
> identifying an entity is needed once a logically unique identifier like a
> record number or URI has been provided. In principle, the latter is all
> that is needed to ensure unique identification of an entity or concept. The
> data associated with that identifier could be arranged in any number of
> ways, with our without being "legibly" distinct from similarly named
> entities, without compromising its logical uniqueness. The works of five
> persons identified only as "John Smith" could be correctly sorted under
> five identical "John Smith" headings, as long as a unique identifier lies
> behind each one. Likewise, the presentation of one person's name could be
> algorithmically constructed any number of ways from a base set of data for
> presentation (John Smith; Smith, John; Smith, John, 1950- ; J. Smith; etc.)
> without compromising the uniqueness of the entity. "Authorization" then
> becomes less formal/nominal and more ontological (if I'm using those terms
> appropriately).
>
An interesting idea, and technically correct. But it also appears at
least by inference to suggest that the sole purpose of constructing
headings is so that we as librarians can differentiate among the various
John Smiths in our catalogues. But isn't the user served by this
approach? If I as a user enter a search -- especially, I might suggest,
a browse search -- for John Smith and am then presented with five
distinct sequences of sub-ordered works by these otherwise
undifferentiated John Smiths (since that's all they say), am I expected
to hunt among them until I find the one in whom I am interested?
Collocation without some visual context for the user not only unhelpful,
it is confusing.
My point is that while we might talk about whether we need to store a
constructed name heading in an authority record (or simply store all the
component bits that ensure presentational uniqueness in a uniquely
numbered record), the heading *does* need to be reassembled at the
presentation stage, since at least some users will be ill-served by
doing otherwise.
- mt
--
*************************************************************************
Marc Truitt
Assistant Dean for Systems Voice : 713-743-8979
University of Houston Libraries e-mail : [log in to unmask]
114 University Libraries cell : 713-443-6421
Houston, TX 77204-2000 fax : 713-743-9811
"I keep hearing you're concerned about my happiness
But all that thought you're giving me is conscience, I guess
If I were walking in your shoes I wouldn't worry none
While you and your friends are worrying 'bout me,
I'm having lots of fun..."
-- The Statler Brothers
*************************************************************************
|