In reply to Marilyn Geller:

A few thoughts in response to one question in your inquiry to the
Serials Discussion Group.
--Mark
-------------------------------------------------------------------------------
Mark Braden                             Internet: marker@oxy.edu
Systems Librarian                       (213) 259-2668 Occidental
College Library         Facsimile: (213) 341-4991 1600 Campus Road Los
Angeles, California 90041-3392

        "Nel mezzo del cammin di nostra vita
        Mi ritrovai per una selva oscura ... "
                                --Dante
This note reflects my own thoughts, no one else's
-------------------------------------------------------------------------------

> 1.  Keyword searching in OPACS vs. Controlled vocabulary searching:  I
> would appreciate comments on the theoretical and practical issues (pro
> and con); documented searches would be helpful too!

Theoretically and practically, I think that both sorts of searching are
important in an OPAC.  Occidental's OASys (an INNOPAC) has a keyword index
containing words from title, LC subject, contents, and a few other fields.
Further, we have a separate LC Subject index.

The keyword index which contains subject-heading words along with title words
is very important for the searcher beginning their work.  I find that title
keyword indexes only help the user who has a "semi-known" item in mind.
A richer keyword index helps the beginning searcher learn what's in our
collection using the terms already known, *and* helps educate them about
standardized words and subject phrases which focus better on the subject of
interest.

So once a person has reviewed a few keyword searches, they've had an
opportunity to "interact" with the database and learn the extent of their
conceptual/vocabulary knowledge of the subject.  Then they can use the
standardized thesaurus of the LC subject headings (and in our system,
the LIMIT function for specific title words/authors/subject words) and
pinpoint materials having a higher probability of relevance.

Implicit in this is adequate user education to equip users for this approach.
Individuals need to learn various aspects of catalogs, databases, boolean
logic, and the peculiarities of specific computer systems.

Using only one or the other index achieves only half the job.  Many systems
have the capacity for both, although sometimes at added expense.