I do the original cataloging for Kresge’s theses and dissertations. I get perturbed when assigning subject headings because SACO just isn’t keeping up-to-date with subject headings.
In library school over six decades ago, we were taught to use the headings from Wilson periodical indexes when more current headings were needed than in Sears or LCSH.
Why not 650 7 $a<Current term>$2wilson? That would create more uniformity amongst us than each of us using our favourite term in 650 4, 650 7 $2local, or 653 (an example of redundancy in MARC).
Perhaps we should also consider 650 7 $a<Current term>$2wiki for Wilipedia.
The codes “wilson” and “wiki” are not in the MARC21 source code list, but there is a *long* list of other codes representing sources which might be more current in their respective fields than LCSH. We find the great number of possible lists daunting, so don’t use any of them, apart from LCSH and the new LCGFT.
If we *really* wanted to improve matters, we would adopt classed subject catalogues with indexes having current terms. We are too locked into our Anglo practices.
And they accuse *me* of opposing change!
One of the strengths (or weaknesses?) of LCSH is that it is based on literary warrant, i.e. on the materials (primarily books) when they are received at the library. See: LCSH: structure & application. http://www.itsmarc.com/crs/mergedProjects/subjhead/subjhead/3_2_lcsh.htm
“3.2 Literary warrant
The Library of Congress collections serve as the literary warrant (i.e., the literature on which the controlled vocabulary is based) for the Library of Congress subject headings system. The number and specificity of subject headings included in the Subject Authority File (the machine-readable database containing the master file of Library of Congress subject headings from which the printed list, the microform list, the CDMARC SUBJECTS, etc. are generated), are determined by the nature and scope of the Library of Congress collections.”
Of course, now this policy goes beyond LC collections, but still applies. “Literary warrant” relies on the literature in the collection, which means in practice primarily that of terms found in titles of printed books received and being cataloged. This has certain consequences. Literary warrant is not the only method for generating these terms, e.g. there is “scientific warrant” which uses terms in current use within the field. These are normally generated by the real experts in the field, which includes articles and non-published materials–even conversations can be used here. Obviously, terms get included much more quickly since you don’t have to wait for the literature to be created. There are other methods too however, as outlined in this nice article here: http://www.iva.dk/bh/lifeboat_ko/CONCEPTS/literary_warrant.htm
The reliance on literary warrant for physical materials in the library’s collection is obviously evidence of how the library is living in another epoch. Back in the days when all information was printed and distributed, the problem was less pressing but the method is showing its age since it is perceived as too slow. The question is: what is the best way to change? One possibility would be to move from “literary warrant” toward “scientific warrant” but I want to emphasize I do not like that at all. It too easily turns into chaos. Nevertheless, once librarians decide to include web materials, such as scholarly blogs, into the mix somehow (which must happen sooner or later) something will have to change.
Also, Jerri is completely correct with the observation: “Let’s get real no one is going to do a search for Future, The, in motion pictures.” But, while they will never look for it that way, they do want the materials collected under that topic. http://www.worldcat.org/search?q=su%3Afuture+the+in+motion+pictures
A lot of people look at something like this and conclude: “The subjects clearly do not work. Just dump them.” I do not agree, but everybody can see that the current situation is not useful for the public.
What is a practical solution? The first task would be to get the syndetic structures to work in a keyword environment (which should have been done 20 years ago), and then I think we could use logfile analysis, refer to other thesauri and even folksonomies to add terms used, figuring this is a type of “user warrant” and perhaps we could even open it up to the public so that they could add cross-references for these terms.
Should the public be able to create headings? Perhaps on a preliminary basis, to allow for a type of “scientific warrant” but this could be potentially difficult since it could easily lead to chaos.
A few ideas.