On Thu, 17 Sep 2009 11:09:26 -0400, Jonathan Rochkind <rochkind_at_JHU.EDU> wrote:
>Doing it probably requires use of authority files, which only VERY
>recently became available (and still not the COMPLETE authority files)
>via lcsh.info. Figuring out the right way to use the authority files
>(which may or may not have the info you'd really want encoded in a
>machine-retrievable way in them) is yet another load of work. Again, I
>think Google just decided "too complicated to figure out, not worth it."
>
>Until WE LIBRARIES who 'own' LCSH have figured out some 'best practices'
>for displaying LCSH in the user environment in a way that is both easy
>to understand and powerful -- and provided a freely available
>infrastructure to make this as _easy as possible_ for everyone else
>(lcsh.info is a big step in that direction, but not close to the end of
>the road).... it is unrealistic to expect someone else, who does not
>consider themselves the custodian of LCSH and doesn't understand it's
>potential power (because we haven't figured out how to show it to
>them!) to do this for us.
I agree with this and would like to emphasize how critical these issues are.
I would like to see us experiment like crazy with the authority files, now
that they are out, at least one form of them. Bernhard Eversberg has made
his LCSH browser, which is much better in many ways that anything LC has.
One of my problems with some of the current suggestions is that LCSH must be
recoded in some way, and that is just not going to happen. We should rather
be concentrating using what we have right now in much better ways than
browsing like in a card catalog. One suggestion I have made in Autocat,
which would make finding headings and creating headings much easier is ...
[gasp!] to simply eliminate the rules for order, and let the user sort the
order however he or she would want. An extract from that message:
"Since these things are very nicely discretely coded now, not only with
600,610,611,650,etc. plus the subfield coding $x $y $z $v (this last much
less consistent), interesting displays can be generated through computer
displays. So, when someone does a search for “france politics” they should
not only retrieve “France—Politics and government” but also be aware of the
subdivision “political aspects” under specific headings to get this very
nice list:
http://authorities.loc.gov/cgi-bin/Pwebrecon.cgi?DB=local&Search_Arg=political
aspects france&Search_Code=SHED%40&CNT=20+records+per+page
and it would be even better if people could work with it in some other ways
instead of being forced to browse it through alphabetical arrangement of the
entry elements (e.g. by subarranging by chronological subdivision).
The display itself could change depending on the search, e.g. search for
“political aspects france” could return (among other headings)
Political aspects France Communication
Political aspects France Constitutional law
Political aspects France Corporate governance
Political aspects France Communicable diseases History 19th century
Political aspects France Dance History 20th century.
The user could determine the display. Someone could click to subarrange the
above display by chronological subdivision to find political aspects of
different topics in France in the 19th century."
{I would like to add to this, that users could also click to see political
aspects of communication in other countries. All the coding should be
available now.)
For this example, I purposely chose the subdivision "political aspects"
which is more or less meaningless without the topic it is related to, but I
think that when seen as a whole, a string such as "Political aspects France
Dance History 20th century" that is all mixed up (to us), is still readily
understood by a patron. Such sorting is quite easy to do today for a computer.
So, I believe that it is still possible to browse headings, so long as we
update the functionalities from their 19th century roots. And one of the
nice things is, to implement this requires no changes from catalogers, and
the rules of LCSH order could be abandoned, thus helping productivity--a
nice thing to point out today.
I understand that this is a radical suggestion, but I think these are the
kinds of suggestions we need today.
Jim Weinheimer
Received on Fri Sep 18 2009 - 05:50:33 EDT