Re: update on lcsh??

> -----Original Message-----
> From: Next generation catalogs for libraries 
> [mailto:NGC4LIB_at_LISTSERV.ND.EDU] On Behalf Of Karen Coyle
> Sent: Wednesday, June 25, 2008 4:27 PM
> To: NGC4LIB_at_LISTSERV.ND.EDU
> Subject: Re: [NGC4LIB] update on lcsh??
> 
> James Weinheimer wrote:
> >
> > I've been labelled a dreamer, and this may be true, but so 
> long as I 
> > am dreaming, I want to point out that we do have coworkers 
> in this field: i.e.
> > the book indexers, who analyse much father down than we do. *If* we 
> > could get other communities to use our (reworked) subjects, 
> using URIs 
> > as it was suggested in the Working Group report and has 
> been accepted 
> > by LC (they exist now at: http://lcsh.info/) why couldn't book 
> > indexers fit into this scheme as well?
> >
> >   
> The problem is that the LCSH as published is a partial 
> vocabulary with complex instructions on how to create a 
> heading. We do not have a list of actual headings, and we do 
> not have a set of cross references that link to headings -- 
> the xrefs only link to the entries in the instruction book. 
> This makes it quite hard to algorithmically link the xrefs to 
> the actual headings in library catalogs.
> The online list represents the Red Books, but doesn't seem to 
> include the detailed instructions. So a heading like:
> 
>      Language policy -- Italy -- Venice -- History -- 16th century.
> 
> Is represented only by the term "Language policy" 
> (http://lcsh.info/sh85074564#concept) in the online list. 
> Somewhere there are other instructions about dividing by 
> place, topic, and date.
> 
> The discussion in the working group report centered around 
> the difficulty of using the "Red Books" to actually create an 
> LC subject heading. I don't think this comes through in the 
> report, but the discussion (at which I was present) seemed to 
> be leaning toward a more uniform way of adding "facets" to 
> the listed headings rather than having separate rules for 
> each heading. 

Thanks for the clarification. I was considering this more from the point of
view of the user: how easy or difficult is it to actually find the headings?
Obviously, I think that it is--and always has been--extremely difficult,
while the arrival of keyword searching and OPAC displays has made it even
more difficult than it was in the card catalog.

If we take "close, but not 100%" as an indicator of success however, I think
the task becomes more manageable. As Bernhard pointed out in his examples
from his *excellent* LCSH browser, there are headings for each of the
heading/subdivisions in the authority file now, and there could allow very
interesting and useful displays.

Once we decouple the headings (from your example),
Venice (Italy) -- Civilization -- 16th century.  
Language policy -- Italy -- Venice -- History -- 16th century.  
Italian language -- Italy -- Venice -- Political aspects.  
Venice (Italy) -- Social life and customs -- 16th century. 

Replaced with: 
Venice (Italy); Civilization; 16th century; Language policy; History;
Italian language; Political aspects; Social life and customs.
(deleting the duplicate entries).

I personally find the traditional array far more comprehensible and it gives
a better idea of the item. When everything is decoupled, things get very
fuzzy, e.g. "Social life and customs" decoupled from "Venice" and "16th
century." Perhaps this is just my brain as a library cataloger working
overtime, though.

While I don't know if the arrangement of subdivisions is so important as in
the card catalogs because people rarely do browses anymore, I believe the
coupling (no pun intended!) of the various parts is still important. In this
way, we know that "History" describes what happened in a special place at a
special time and is not about the discipline of history.

I know it would be considered sacrilege, and the reptilian part of my brain
rebels against this idea ;-) but perhaps we could simply abandon the rules
for ordering the subdivisions and consider them all as "Tinkertoys," and
only distinguish between primary use of a concept, e.g. 650 $aHistory (as a
discipline) vs. 6xx $xHistory (everything else), also allow for duplicate
entries (in this case, reuse 16th century and Venice (Italy) more than once
for clarity's sake), and finally, get rid of indirect geographic subdivision
when all terms are duplicated (i.e. abandon the use of --Italy--Venice and
always use Venice (Italy)). 

As a result, we could have something like:
Venice (Italy) Civilization 16th century; Language policy Venice (Italy)
History 16th century; Italian language Venice (Italy) Political aspects;
Venice (Italy) Social life and customs 16th century.

In a correctly configured catalog, these could display in all kinds of ways,
e.g.
Venice (Italy) Civilization 16th century
Venice (Italy) Language policy History 16th century
Venice (Italy) Italian language Political aspects
Venice (Italy) Social life and customs 16th century

Or:
Civilization 16th century Venice (Italy) 
Language policy 16th century History Venice (Italy) 
Italian language Political aspects Venice (Italy) 
Social life and customs 16th century Venice (Italy)

Or many other ways. This could retain at least some of the good points of
current subject practice and be a lot easier to create. Maybe this is how
FAST works, I haven't looked that much at it. This could work very well in a
system of concept URIs and maybe even book indexers could follow it.

A somewhat similar idea is in a very interesting article that Melissa
Hofmann was kind enough to send me:
Anderson, James D. and Melissa A. Hofmann. "A Fully Faceted Syntax for
Library of Congress Subject Headings." Cataloging and Classification
Quarterly 43.1 (2006):  7-37.

Jim

James Weinheimer  j.weinheimer_at_aur.edu
Director of Library and Information Services
The American University of Rome
via Pietro Roselli, 4
00153 Rome, Italy
voice- 011 39 06 58330919 ext. 327
fax-011 39 06 58330992