The Distributions of MARC Fields in Bibliographic Records

Abstract

Library catalog systems worldwide are based on collections of MARC records. New kinds of Functional Requirements for Bibliographic Records (FRBR)–based catalog retrieval systems, displays, and cataloging rules will build on ever-growing MARC record collections. Characterizing the kinds of information held in MARC records is thus an important step in developing new systems and rules. This study examined the incidence and prevalence rates of MARC fields in two different sets of library catalog records: a random selection of bibliographic records from the Library of Congress online catalog and a selection of records for two specific works, Lord of the Flies and Plato’s Republic. Analysis showed that most fields were used in only a small percentage of records, while a small number of fields were used in almost all records. Power law functions proved to be a good model for the observed distribution of MARC fields. The results of this study have implications for the design of new cataloging procedures as well as for the design of catalog interfaces that are based on the FRBR entity-relationship model.