![[RSS]](http://www.mySiteFeed.com/images/rss.gif "Add the RSS feed to your RSS reader by dragging and dropping this icon onto the address box of your RSS reader"){kind=small}
![[Google]](http://www.mySiteFeed.com/images/add-to-google-plus.gif "Click to add the RSS feed to your Google account"){kind=small}
Library Philosophy and Practice Vol. 2, No. 1 (Fall 1999)ISSN 1522-0222Retrospective Application of Subject Headings, Part 1: a Case Study at the Central Washington University LibraryDaniel CannCasciatoHead of Cataloging Central Washington University Library Ellensburg WA 98926-7548 | While many libraries update their bibliographic records when existing headings change (e.g., fromLabor and laboring classes to any of three headings), it is not known how many libraries attempt to retrospectively apply new headings to existing bibliographic records in the way that LC does. The lack of publications on this topic indicates little activity in this area. Additionally, the static nature of bibliographic records in many catalogs (as illustrated by theSustainable development case) also indicates very little activity in this regard. A procedure for the retrospective application of subject headings is needed to fulfill the catalog's objectives regarding subject access. Without it, patrons are misled in various ways. In the case where the heading is simply not included in an appropriate bibliographic record, patrons are denied access after they have found terminology that suits their research needs. Additionally, the syndetic structure of authority records will actively mislead patrons. In catalogs with authority modules in place, patrons are led from the heading Economic development-Environmental aspectsto the heading Sustainable development. In catalogs where bibliographic records are static, the patrons who take that guidance (frequently just a mouse-click's effort) will then be diverted from their research needs, since some bibliographic records for that item will not have been updated with the newer and more appropriate subject heading. Thus, either on their own or with guidance, patrons find less than a library has. The intent of this article is to describe a methodology used to implement a procedure for retrospective application of subject headings new to the catalog. The focus is on the implementation of a practical procedure for this process and includes a discussion of preliminary results. Part 2 of this article will provide results from a larger set of data and also investigate secondary implications. The author believes that if libraries are committed to fulfilling the objectives of the catalog as related to subject access-that is, to allow a patron to find what a library holds by subject-then a procedure for the retrospective application of subject terminology is required. Background EnvironmentThe catalog at Central Washington University Library (CWU Library), called CATTRAX, is the one used for this study. The author became the head of cataloging at CWU Library in October 1995. The CWU Library migrated to the Innovative Interfaces system, INNOPAC, in early 1996. At that time, bibliographic and authority records were loaded into the catalog, and new records have continued to be loaded. While authority control has been ongoing, it has been very sporadic. Automated authority control was performed on the original tape load of records and also on a subsequent retrospective conversion project in late 1997. No ongoing, systematic, subject authority control is yet implemented. The number of bibliographic records in the catalog is approximately 700,000. MethodologyFor the purpose of this study, a list of subject headings proposed in calendar year 1995 was extracted from the CATTRAX. Proposals are assigned a number with an SP prefix when the heading is keyed into the LC system. The heading, when accepted, has its prefix changed to an SH. The number assigned however, does not change. The headings from 1995 were selected based on the SH number beginning with a 95. Five hundred fifty-nine records were compiled into this list on June 25, 1999. TheAppendix contains a list of headings reviewed. The records proposed in 1995 were selected because they denote an easily culled subset of records that, by and large, had not been reviewed by the author already for a retrospective application process. While the author undertakes the retrospective process on an ad hoc basis in general, and comprehensively for all new headings proposed via SACO participation, as yet there is no systematic procedure in place at the CWU Library for this process. If a systematic procedure were already in place (or becomes employed) then headings added based on calendar year of addition to the catalog (or some other periodic time frame) would be used. The first fifty valid headings in this file were reviewed for retrospective application. Headings created by the CWU Library's participation in the SACO Program were omitted from the study, as were headings that had been previously reviewed by the author5. The headings, after being compiled, were not sorted further. They are generally in SH number sequence. Due to some vagaries in the MARC load table at the time records were loaded, some records are out of the SH sequence. The procedure followed was meant to allow for timely processing of records. In the description of the procedure, the subject headingLearning strategies (sh95-636) will be used as an illustrative example. Procedure:
Limiting the search by title was selected from the many limiting options available in the INNOPAC system, because of section 12 of instruction sheet H 180, in LC'sSubject Cataloging Manual: Subject Headings (SCM:SH).
While there are caveats for this practice, in general it is a principle that catalogers follow and which patrons should expect. Each broader term from the subject authority record was
then searched as a subject string The number of records found includes the heading or the heading-subdivision string. In the CWU Library catalog, word searches search on names (personal, corporate, and conference) titles (including uniform, alternate, and series) subjects, summary notes, and contents notes. When suitable bibliographic records were discerned as viewed, no further review of materials was carried out. That is, not all the physical items were pulled and reviewed for appropriateness. When probable titles were discerned, those titles were searched in various online catalogs. If that did not resolve the question, then the item was paged from the stacks and reviewed. If appropriate, the new heading was added to the bibliographic record. Total new headings were then tallied. ResultsFor the fifty SARs reviewed, fifty-one new headings were added to the catalog's bibliographic records. Table 1 gives the breakdown by number of changes. Table 1
No more than seven new headings in bibliographic records were added for any individual subject heading in this process. The results in Table 1 account for those headings added when strictly following the procedure as outlined. In the following discussion, methods in which additional headings were added to the catalog are described. Those headings are not included in Table 1. An adjusted count as well as a list of all headings in this study are included in the appendix. By using extra-procedural strategies, an additional twenty-one headings were added, bringing the total number of changes to seventy-two. DiscussionThe appropriateness of application of the newer terminology to existing bibliographic records is of concern. Existing resources do help alleviate doubt in some situations. A number of items had updated records in OCLC's WorldCat and were available for consultation quickly and cheaply. These same records are also freely available via LC's catalog. WorldCat, with member records, is a larger resource than is LC's catalog. In one instance, the information in a 653 field from a bibliographic record in WorldCat made it clear that the newer term was appropriate for the existing bibliographic record. This information is not in the bibliographic record in LC's catalog. Occasionally, reviewing the physical items was the only method available for determining of appropriateness of the newer headings. There are a number of other problem areas involved in this process which will lead to procedural changes for the second part of the project. Briefly, they are: TruncationThe use of truncation can be employed to facilitate retrieval and cut down repetitive searching. This happened in the areas of singular versus plural forms and headings that are qualified. Singular/Plural TerminologyThe problem of singular/plural terminology is of interest. When word searching the plural form, recall is limited. During the word search process in the INNOPAC system, truncation can be used. That is, rather than do a word search onAdversaries (Persons) (a reference toEnemies (Persons)), a better search would be (word)adversar*. (The omission of qualifiers from the search strategy will be discussed next.) Thus, recall is improved. Another example from the sample is:
(word)serial murder* When the word would result in an overwhelming number of records being recalled (enem* for example) then performing a double search of each term, the singular and the plural, proved more manageable. QualifiersFor headings that use qualifiers as a means of clarification, omission of the qualifier terms improved results.Fetishism (Sexual behavior) was better searched as the termfetishism and resulted in the correction of one incorrectly formulated heading. Likewise,Enemies (Persons) was better searched asenemies. Doing so resulted in additional headings being added to the catalog. The same was true for the "see" references forEnemies (Persons). A word search onFoes (Persons), provided nothing, while a word search onfoes did, resulting in additions not only for persons, but also forInsect pests andBird pests. Broader TermsFor some headings, the SAR included broader terms that were further subdivided, usually by place, but they can also be topically subdivided. In those cases, searching also by the topical heading alone seems like a useful strategy, although in this sample it did not result any additional headings being added to the catalog. In the case ofLearning strategies, the broader termLearning, Psychology of did yield useful results. However, the earlier lack of the SAR for this record required catalogers to approach this topic as best they could for the intervening years before the subject heading was established. On one occasion, the headings assigned wereLearning andCognitive styles (among others). In this case, those terms ended up serving as ad hoc broader terms. Thus, some cataloging dispersal, rather than collocation, during the lag time between use of the term in the literature and SAR creation must be taken into account when searching the older terminology. Alternate TerminologyAlternate terminology, not yet incorporated into the initial procedure, is also a valuable retrieval point. For example,American bison hunting has a broader termBig game hunting. This broader term did not net any significant results. A word search onbuffal* hunt* did, which resulted in four additional headings being added to the catalog (two for the headingAmerican bison hunting, two for the headingAmerican bison). It also worked as a means of identifying the misapplied heading,Buffaloes (for Eastern Hemisphere animals) to works about the American buffalo.7 Another example of alternative terminology is demonstrated with the headingEthnic costume. A word search on the heading itself or its broader terms yielded no changes. A word search ontraditional* costum* yielded better results and resulted in five additional headings being added to the catalog (four of those being on one bibliographic record). Geographic HeadingsGeographic name headings provided few changes. Of the nine SARs reviewed, only two resulted in additions to the catalog, one each. Literature and Music HeadingsLiterature headings and one music heading (Music theater) in this sample all proved to be too vague to be used. Each would probably require the review of physical items, yet even identifying appropriate items to review was not feasible. Revised ProcedureThe experience of this small sample has resulted in the following modifications to the review procedure. This new process will be followed in the next part of the project. Procedure [revised]During this process, limitation by title word or consultation with reference sources other than the catalog will happen only as needed. Additions of the subject term will again be tracked, as will secondary changes noted in the previous discussion, e.g., corrected headings, additional headings for other topics, etc. Also, any potential SACO applications will be identified. In this sample, a likely SACO proposal would suggest that Traditional costume be added as a "see" reference toEthnic costume. ConclusionsThe procedure followed produced usable enhancements to many catalog records. The process allows for the application of newer terminology to appropriate items. It secondarily leads to the application of additional headings to provide appropriate access to already cataloged items. Fulfillment of the catalog's objective for subject access is improved by this process. In an age of shrinking acquisitions budgets, the provision of effective access to materials held by the library is vital. Coupled with the stated objective of the catalog to allow a patron to find what a library has on any subject, and the great potential for an effective authority module to divert patrons away from previously cataloged materials, this procedure of retrospective subject application appears to be beneficial and practical. References1. Cutter, Charles Ammi.Rules for a Dictionary Catalog, 1876. 2.Subject Cataloging Manual : Subject Headings, 5th ed., 1996. Instruction sheet H 200, section 9: "Bibliographic file maintenance. Before submitting a completed proposal to the Subject Headings Editorial team, search the MARC database to locate existing bibliographic records that should be updated as a result of the new proposal. Enter the total number of records to be changed in a 952 field." 3. The revision date is inferred from the LCCN for this title's 4. Searches conducted by the author in the following catalogs: Orbis (orbis.uoregon.edu); OhioLink (http://olc1.ohiolink.edu/search/); Melvyl (telnet://melvyl.ucop.edu), OCLC's WorldCat (in cataloging mode, via internet); LC's online catalog (http://lcweb.loc.gov/catalog/); University of Washington catalog (http://catalog.lib.washington.edu/search~/) 5. Some of the headings included in this study were potentially previously reviewed. Only those which were certainly previously reviewed were omitted from the sample. Thus, the results could have fewer changes than would otherwise have occurred, but not more results. 6.SCM:SH (1996), Instruction sheet H 187, section 1. 7. In one of the catalogs searched, the misapplied heading had been updated, but worsened. That is, it had been changed not toAmerican bison, but toBuffalo (N.Y.). AppendixHeadings Reviewed in this StudyThe category ofChanges+ includes changes made due to extra-procedural strategies as well as the application of heading not on this list to bibliographic records. For example, the headingBird pests was added as a result of the word search on the termfoes.Foes (Persons) is a "see" reference toEnemies (Persons). Also, the termEnemies (Persons), when used after a personal name heading, has the form[name]-Adversaries. This form of access was also added to a number of bibliographic records. Headings are listed in the order in which they were reviewed.
Sorted by number of changes made (column 2), then alphabetically by heading
|
