 | |  |
Subject Assigning
Subject assigning constitutes, in actuality, a subset of authority work. The same need to maintain consistency over the forms of access points makes highly controlled subject coding crucial in an on-line environment.
Subject indexing differs from name authority work, however. Subject assigning must be, by its very nature, less exact than name authority work. A part of the lore of many editing projects concerns the time when editors and processors passed around a single document and no one assigned the same subjects to it. The story is probably apocryphal; editors grow so accustomed
Although the style of an index may vary, its quality depends in large part upon internal consistency. If indexing something with great relevance to the social history of the period, such as Samuel Sewall's diaries, one cannot create an entry for "beer" on one page and index "beer" under "alcoholic beverages" on another (if this is the only such beverage appearing on the page).[16] Readers should feel confident that when they look up an entry, they will find there all the relevant portions of the text, unless cross-referenced to another subject heading.
Electronic information storage and retrieval demands not only internal consistency within an index, but that the subject headings used conform in their form, meaning, and application to widely accepted standards. Thesauri of subject terms present editors and archivists with a means of achieving this external level of consistency.[17]
The Library of Congress's Subject Headings (LCSH) have become for the library and archival world the de facto thesaurus, though in reality those two large red books hardly deserve the term. A true thesaurus would allow users to be able to follow, through cross-references, natural language forms of an expression to the most technical. For example, if someone tries to use "street-walkers" they will not be referred to "prostitution." In other words, one must in general know the proper authorized term before using LCSH. The form of LCSH also differs from a true thesaurus. A thesaurus like that of the Educational Resources Information Center gives broader, narrower, and related terms as well as extensive scope notes; LCSH only renders "see also" references and has very little advice on the way the terms should be used.[18]
Tomes can be written about the inadequacies of LCSH for the age of electronic information storage and retrieval. LCSH began as (and to a large extent remains) a manual thesaurus; as of mid-1987 it had still not gone on-line. One must thumb through thousands of pages of two unwieldy volumes to use the work rather than being able to pose simple queries to a data base. To the horror of experienced data base managers, the LCSH has been issued since 1908 in ten editions; no variations—and they have been considerable—between the editions have been cross-referenced or even formally tracked by the Library of Congress. To make matters worse, the headings used by LCSH bewilder specialists in several fields. Why would film historians ever look under LCSH's "moving-pictures" when they probably never used the term in their entire scholarly careers? And LCSH manages to offend, in various ways, several special interest groups, especially minorities. The Subject Headings have been so insensitive to women's issues that "counter-thesauri" have been posed as correctives. Finally, LCSH does not provide the descriptive depth required by the type of document-by-document indexing
Given all LCSH's problems, should editors avoid it completely? Probably not, since the Subject Headings remains the closest thing to a true, national thesaurus. If the federal government ever does realize the importance of such a thesaurus for the electronic information age and supplies funds for bringing together specialists in various fields, lexicographers, and librarians to develop a definitive thesaurus, undoubtedly the basis will be the system developed by the Library of Congress.
Editors can make use of LCSH if they use it only as a basis for subject assigning. For example, the Goldman Papers addressed the limitations of the Library of Congress system in two ways: by reshaping the original scope of the subject headings to include women's experiences and by creating new headings which are then cross-referenced back to the nearest Library of Congress term. The project has avoided working with the large volumes by pre-selecting the terms likely to be used and entering them into a thesaurus data base. Subject terms have been grouped beneath ten major headings that range from "Arts and Letters" to "Economics" and in over a hundred subgroups. For example, "Booktrade" appears beneath "Arts and Letters." Subject headings displayed under their assigned group and subgroup are frequently printed out to be used by staff members who process documents. Subject assigning follows a system similar to that for name authorities in which four-letter mnemonic codes stand for headings (i.e., "aupu" for "Authors and Publishers"). The data entry program ensures that codes will not be duplicated and allows for the entry of new terms in the thesaurus data base with proper group and subgroup headings.[20]
With the new generation of microcomputers based upon the Intel 80386 microprocessor chip and the new operating system being jointly developed by Microsoft and IBM (OS/2) it will become feasible for projects to consult a subject thesaurus directly on the CRT screen in a manner similar to the various thesauri currently available for word processing programs. A pop-up menu would present various authorized options to a natural language subject term which can be entered electronically by merely hitting a number key. However, on today's IBM-compatible microcomputers based upon the 8088 and 80286 chips, such thesaurus authority work would be difficult to manage and would be prohibitively slow given the random access memory limits of these computers.
| | |