How do you choose a subject heading?

K.4.3 Subject Heading List

A subject heading list is “a standard list of terms to be used as subject headings, either for the whole field of knowledge or for a limited subject area, including references made to and from each term, notes explaining the scope and usage of certain headings, and occasionally corresponding class numbers” [5]. Such a list is normally arranged alphabetically. Both preferred and rejected terms are listed in the same sequence. The terms are linked by “See” and “See also” references. The most well known subject heading lists for the whole field of knowledge are Library of Congress Subject Headings and Sears List of Subject Headings, while Medical Subject Headings [MeSH] is an example of subject headings list on a limited subject area. However, most of the subject headings lists have now adopted thesaural structure. More discussions on subject headings lists may be found in any book on library cataloguing or resource description.

10.6 Summary

Subject headings are controlled lists of terms or phrases used to describe the subjects of items. They facilitate access to resources by subject, as they define a set of additional topics included in an item, in contrast to classification, which deals only with the main subject. The best-known list is the LCSH. Other subject lists, like the Children’s Subject Headings and the Sears List of Subject Headings, are derived from LCSH. Subject headings may be subdivided to provide more tailored representation of the topics in the resource. References within the lists make possible the location of narrower, broader and related topics. This allows the identification of the most appropriate terms, no matter what the starting point may be.

Subject headings

Subject headings go back to the days of the card catalogue. The subject of a book was described in one or a few words which were typed on top of a catalogue card. This practice had some consequences. It was important to choose a subject heading that expressed the contents of the document as exactly as possible. If you would choose only a few vague or general terms you would have to retype the catalogue card as many times as you had terms, which was of course a lot of work in the days before library automation. For the same reason a system was developed to formulate complex subjects in compact form. In the LCSH one can find combinations like:

Europe–Civilization–Nineteenth century

Automobile Association [Great Britain]

Agriculture and state–Denmark–History–Nineteenth century–Statistics

The downsides of such precoordinatve indexing systems are that they tend to keep on growing rapidly and that it is labour intensive to add even obvious variations of existing terms to the list. Suppose the system has these subject headings on libraries in Jordan:

Libraries–Jordan

Libraries–Jordan–Catalogs

Libraries–Jordan–Congresses

Libraries–Jordan–Directories

Libraries–Jordan–Periodicals

The day the library gets a book on the history of libraries in Jordan a new term will be added to the list:

Libraries–Jordan–History

In this case an indexer has to suggest this as a new term; the term has to be approved and formally added to the list of allowed terms. In the early days the already long list of terms had to be retyped over and over again. In LCSH not all combinations are listed anymore; LCSH now have what is called ‘free-floating subdivisions’: words or combinations of words that can be combined with a vast range of other terms, e.g.

Atlases

History, military–Twentieth century

Climate

The syntax rules of the subject headings also determined which part of a multi-word term should come first, because this was important for the alphabetical classification of catalogue cards in the subject catalogue. The reader who was looking for documents on a certain subject in a card catalogue had to browse the paper cards in one of the little drawers of the catalogue. The reader would find all catalogue cards dealing with

Agriculture–Africa–History

Agriculture–America–History

under ‘Agriculture’ and not some under ‘Africa’ and others such as ‘History’, etc. In automated systems it would not matter which of these terms came first:

Computers–Law and legislation–France

A combined search, e.g.

France AND Computers AND Law

would retrieve all the records with these terms, no matter how they are arranged. But in a manual system you have to be rigid about the order of terms.

National and de facto standards give different rules for the construction of subject headings. A combination like

Agriculture and state–Denmark–History–Nineteenth century–Statistics

may be possible in the LCSH, but would be considered wrong in many other systems, because their rules would not allow terms to be built out of five parts and they would also forbid using a broad and a specific temporal attribute [‘History’ and ‘Nineteenth century’] at the same time.

A subject headings system may take one of the following forms:

A simple, although probably long list of allowed terms.

A list of allowed and disapproved terms, with a mechanism to guide the user from disapproved to the preferred terms. Mostly also references to related terms are included.

A list in which relations between terms are documented in the way of a thesaurus.

In many cases a subject heading system uses simple relations between terms:

Reference from a disapproved term to an approved one:

References from a term to related terms:

Although the ‘see also’ relations are not subject of strict rules, one of Cutter’s laws is still valid, i.e. that a ‘see also’ relation may not refer to a broader term such that:

is not allowed. This may be one of the reasons why some people prefer thesauri.

Real World Example 3

M.2.4 Cross-References

From the above specific subject headings it can be easily visualized that the document which deals mainly with “heart disease” or at least “medicine” is indexed under a place “India.” This is due to backward rendering process. Therefore, a very effective system of cross-references is necessary here to satisfy the users from all other angles of the subject. Ranganathan has suggested a series of cross-references from the first sought link to the last for this purpose. The following cross-references can be derived for the CC number.

MEDICINE

See also

INDIA, TREATMENT, DISEASE, HEART, MEDICINE

CIRCULATORY SYSTEM, MEDICINE

See also

INDIA, TREATMENT, DISEASE, HEART, MEDICINE

HEART, MEDICINE

See also

INDIA, TREATMENT, DISEASE, HEART, MEDICINE

DISEASE, HEART

See also

INDIA, TREATMENT, DISEASE, HEART, MEDICINE

TREATMENT, DISEASE, HEART

See also

INDIA, TREATMENT, DISEASE, HEART, MEDICINE.

The above cross-references are made from general to specific; therefore, “see also” cross-references are adopted. In the cross-references from no. 2 onward, the necessary qualifying terms are added to the headings to specify the particular context of the subject. The procedure for deriving cross-reference entries for DDC number will also be the same.

Using metadata to catalogue individual issues of nickel weeklies

The digitized nickel weekly collection at the BPCL is slowly becoming part of the OhioLINK Digital Resource Commons [DRC], Ohio’s comprehensive digital repository. However, before any issues were entered, there was much discussion about which fields to use to describe each issue. The DRC uses a modified Dublin Core metadata schema [see Example 5.10b for many of the fields], which made decisions a bit easier. Discussions included: how much information should be included in these records, what information would be the most useful, which stories should be assigned specific metadata, and which could be mentioned in the descriptive notes.

The digitization has been completed, but assigning of metadata continues. As the metadata is completed the issues are batch uploaded into the DRC [//drc.ohiolink.edu/]. The most pertinent fields are outlined in Example 5.10a.

Challenges

When cataloguing these materials began, the first challenge was determining how to catalogue them – as serials [i.e. continuing resources in OCLC] or as books. It was determined that the serials format would be more appropriate, as most zines are typically published as serials, albeit irregularly. A few zines have been catalogued as books if they appear to be unnumbered and only single issues.

A second challenge is the lack of publication information and the irregularity of publication. Oftentimes no place or date of publication is available on the zines and there is no way to research them. The author’s address, if available, is often used as the place of publication. Numbering

sequences are sometimes non-existent or inconsistent, and publication frequency is usually given as irregular, since most zine publishers print zines around their own schedules, not standard publication periods.

A third challenge is the assignment of subject headings. All zines in the BPCL and MLSRA were given an LC subject heading 650 0 Fanzines [before Zines became available as a heading]. In addition to that, other topical LC subject headings were added as appropriate. Local subject headings were also created for different types of zine. The following are some examples of local 690 subject fields:

Musiczines [for all zines held in the MLSRA relating to music]

Reviewzines [all zines that included reviews]

Perzines [for personal zines]

Artzines [for art-related topics]

D.I.Y.zines [for do-it-yourself zines]

Dipzines [for Diplomacy zines]

Slashzines [for underground/erotic zines].

Many zines lack even one particular subject, while others cover too many subject areas to apply all the appropriate subject headings. While the lack of subjects/topics is prominent in many zines, it is still important to provide some subject access rather than none.

Databases and Search Strategy

The following databases were systematically searched: MEDLINE, EMBASE, and Cochrane. Databases were searched from their inception to August 31, 2015.

For the search strategy, the Medical Subject Headings of “warfarin” and “cost benefit” were used to combine with the following keywords “incremental cost-effectiveness ratio,” “Quality-Adjusted Life Years,” “cost utility,” “economic,” “cost effectiveness,” “pharmacy service,” “pharmacist-provided,” or “pharmacist-managed,” “anticoagulant.” Appendix A shows the search strategies.

There was no language and study design restriction. References of initially identified articles were examined to identify additional studies that met the selection criteria.

7.3.2 Enumerative

An enumerative structure is an alphabetical listing of subject headings with numbers assigned to each heading. Enumerative classification attempts to list [enumerate] all the required single and composite subject concepts. Figure 7.2 is an excerpt from class N of LCC showing a few of the subclasses and the divisions of the class. Enumerative structures are also referred to as top-down classifications because they start from the whole universe, and through repetitive divisions reach more specific concepts. The difference between an enumerative structure and a hierarchical structure is that the enumerative approach contains composite concepts which disrupt the arrangement from ‘general to specific’.

For classifying a resource in a purely enumerative system, the classifier has to discover the predefined subclass that best matches the ‘aboutness’ of the item. For compound-subject items this is a concern because there is a high possibility that no perfectly matching division exists. Another problem is that the subclasses are lengthy, containing composite and complex subjects with certain terms recurring often in different divisions. LCC is a distinctive paradigm of enumerative system with limited characteristics of a hierarchical arrangement. DCC is also an enumerative system, but with more hierarchical features.

3.4.2 Keywords

Some journals will require authors to provide keywords or subject headings for their articles. As an informationist with a keen knowledge of the importance of keywords for findability, we are uniquely suited to help in this regard. Let authors know that this is how their peers will find their article in a literature search, and that a wider audience generally leads to more citations. Warn about keywords that are too broad, and to be sure to use the maximum number of keywords allowed by the publisher.

Some publishers specify the vocabulary to be used, such as MeSH. If they are having difficulty in choosing terms, search PubMed for articles on a similar topic, and look at the terms that were assigned to that article.