SIGIA-L Mail Archives: Re: SIGIA-L: Insight moves sight to the
Re: SIGIA-L: Insight moves sight to the site
From: Chris Chandler (chrischandler67_at_earthlink.net)
Date: Sat Feb 09 2002 - 01:56:26 EST
Ziya Oz:
> You seem to suggest that given any subject, there is one
canonical way to
> organize it. (If not, you can ignore the rest.)
Are you illiterate or just obstinate? <smiley>
Seriously though, James said you could use "any
classification you want to come up with" to categorize
Hemingway titles. There is no CANONICAL classification
scheme. However, having once chosen a classification scheme,
it is only worth something if it is executed consistently.
(Sometimes... ok, often... in the messy real world the
requirement of consistency means bending, blurring or even
breaking the original boundaries between classification
terms.)
To your point, and at the risk of blurring the distinction
between cataloging and access that James worked so hard to
establish, if the owner of a site chooses to "generate
multi-dimensional classifications" as you say -- they still
need to apply that (arbitrary even in n-dimensions)
classification consistently, because searchers tend to
expect consistency, and, let me repeat, since there is no
canonical classification system you can't expect the
searchers to know what classification scheme you've used
ahead of time.
In an earlier post in this thread, you made some good points
about why you almost always want to use human brain power
over pure classification-by-algorithm. Even after you've
decided on a scheme and made the tough decisions about where
stuff goes, you still need human catalogers to decide where
to put new stuff, because the choice doesn't just depend on
the scheme and rules that a computer can understand, it is
(path) dependent on the choices that were made about
"similar" previous material by previous catalogers.
Also, could you expand on what you see as the "confines of
library science?" --because it sounds like you think
librarians only know about books and periodicals and have
never heard of computers, databases or SGML.
-cc
This archive was generated by hypermail 2.1.2
: Sun Nov 23 2003 - 22:55:02 EST
|
