Subject determination during the cataloging process: An intensive study of five catalogers

Alenka Sauperl, Jerry D. Saye


Subject headings and call numbers for the subject description of documents in library catalogs are provided by catalogers, who bridge the gap between producers (e.g., authors) and users of documents by developing representations to support information retrieval. While catalogers prepare document representations according to a set of standards and requirements and using subject heading lists and classification schedules, they also are in a position to select headings and classes that bring together authors' and readers' interests. In spite of the long tradition of subject cataloging, it is still not completely understood how this process works. The question that the research reported here investigates is how do catalogers decide about the topic of the document and appropriate subject description? Five experienced catalogers were observed and interviewed about their work and experience with subject cataloging in libraries. The think-aloud method was used for the observations, and unstructured interview was used for the follow-up discussion. A follow-up discussion clarified any uncertainties and explored some issues in more detail. The results were compared to the results of other similar research. The observed catalogers consistently performed subject cataloging in five stages: (1) identification of the topic of the book; (2) identification ofthe authors' intent; (3) inference of the possible uses; (4) relation of the topic to the existing collection; and (5) relation of the topic to the classification scheme and subject headings list. The sequence of stages was not necessarily linear, but was flexible in such a way that the catalogers returned to any of the previous stages whenever they realized they were on the wrong track. Their common strategy was searching for existing patterns in sets of call numbers and subject headings that had been assigned to the items in the existing collection. This study, limited in sample of five catalogers and thirteen books, does not allow for generalization over all catalogers. It is, however, useful in comparison with other studies and valuable in presenting different research questions that are still open in subject cataloging.

Full Text: