Procedures for Bypassing Initial Articles

Department: Technical Services

Bypassing Initial Articles

Your Innovative system automatically detects the English language initial articles "a", "an", and "the". It updates the SKIP fixed-length field and, for full MARC records, the second indicator of the 245 field (a 2, 3, or 4 for "a ", "an ", and "the " respectively) to ensure that the article is stripped from the title for indexing.

The system does not automatically detect initial articles in languages other than English. When you enter a non-English title, enter the correct indicators for proper indexing and retrieval.

You can set the system to remove contracted articles from the beginning of words before adding the words to the index. For example, the system can index the word "l'enfant" as "enfant".

Note that if "a" is not an initial article, as in the title A to Z, you must correct the indicator appropriately. To facilitate retrieval by user searches in the OPAC, you can make an added entry with the indicator set as if the title had an initial article. The following note examines this situation in more detail.

Coordinating Title Indexing and User Searches

You can set the system to ignore specific articles when they appear at the beginning of OPAC search statements. The list of articles to ignore is, technically, a list of character strings and can include non-English articles in addition to English words such as "a", "an", and "the". (For information on making this setting, see Characters removed from beginning of user-typed search.)

If the list includes the article 'a', users can retrieve the title "A Book to Remember" by searching for "a book to remember" or "book to remember". Without this setting, users must learn to omit initial articles from their searches.

The system, however, cannot determine when a character such as 'a' is not an initial article, as in the title A to Z. Consequently, it reads a search for "A to Z" as a search for "to Z" and fails to return the correct title.

To compensate, you can make an additional title field entry in the bibliographic record (MARC tag 740, Innovative field tag 'u') with the first (non-filing) indicator set to "2", as if the first word (and space) is an article. The system indexes this additional field as "to Z". A search for "A to Z", read by the system as "to Z", would then return the appropriate record. Alternatively, you can instruct users to double the initial letter on searches of this type. If the user enters the search string "A A to Z", the system removes the first "A" and searches for "A to Z".

Though not recommended, you can set the second indicator in the 245 field to indicate that the "A " in the title A to Z is an article.

Related problems can occur with non-English articles. For example, if you correctly set the second indicator to '3' in the 245 field for the title "Le Monde", the system does not index "le ". Thus, if "le " does not appear in the optional list of initial articles (characters) to ignore in OPAC searches, a search for "Le Monde" fails. In this case, the user must search for "Monde". To solve the problem, you can:

Note that adding non-English articles to the list of initial articles ignored in OPAC searches can cause the omission of some English titles from the search results. For example, adding "los ", "la ", and "el " to the list causes searches for names such as "Los Angeles Times", "LA Times", and "El Paso Times" to fail.

Likewise, adding the German article "die" causes the omission of titles such as "Die and Tool" or "Die Hard". You can use the methods in the preceding list to address these omissions.

Libraries with a primarily English catalog and relatively few non-English titles might prefer not to strip non-English articles from OPAC searches.

________________________________________
Last updated April 2008