Twenty-five years ago preservation was largely a neglected area. This historical review focuses on major events, activities, and publications that have contributed to the emergence of preservation as a vital specialty within librarianship.
A survey of the cataloging departments of 166 OCLC-member academic libraries showed that libraries did not rely exclusively on OCLC for card production and that a large majority did not accept non-Library of Congress OCLC records without substantial checking. The survey indicated also that libraries that used or planned to use their OCLC-MARC tapes generally were more concerned with the completeness and accuracy of their OCLC-MARC tape records than libraries that did not plan to use their tapes.
This study is designed to provide librarians with a practical guide for reaching an informed policy decision on the question of submitting error reports to OCLC for the purpose of data base quality control. It addresses three questions: (1) What types of errors, changes, or additions should be reported? (2) Once reported, will errors be corrected promptly? (3) What is the cost of error reporting? It also reports on ENHANCE, a new approach to quality control under development by OCLC.
Late in December 1979 a questionnaire was sent to 446 members of OCLC, Inc., to ascertain their views on the development of a special minimal standard to be used only for retrospective conversion projects. It was found that 74.5 percent would oppose the creation of a new lower standard. Despite the possibility of extra requirements to upgrade an earlier catalog record to meet current standards, librarians, in general, showed a strong commitment toward quality and compliance with current standards. Their attitude generally was altruistic, often reflecting a greater weighting toward future uses of the data base, a desire not to repeat the past mistakes of minimal records, and a desire to see standards raised.
Research libraries face difficult choices in establishing technical services pro¬cessing priorities, especially in the selection of items to receive original cataloging. The author designed a study to examine the disposition, one year after input, of a sample of original cataloging contributed to the OCLC data base by the Indiana University, Bloomington, Libraries. Findings indicate that most original cataloging contributed by this library was not superseded or duplicated by Library of Congress cataloging and that many items were not cataloged by any other OCLC member library. Systematic replication of methods similar to those employed in this study should provide information useful both for library management review of cataloging activity and for decisions to be made at the national and/or regional level about designation of centers of cataloging responsibility.
This investigation examined the characteristics of subject headings found in OCLC cataloging records. The study analyzed a sample of 33,455 mono-graphic records taken from the OCLC data base. The sample contained a total of 50,213 subject headings, 94 percent of which were Library of Congress subject headings. Each record had an average of 1.4 Library of Congress subject headings. However, 18.6 percent of the records had no LC subject headings assigned. Topical subject headings accounted for 71 percent of all LC subject headings, and 62 percent of all records contained at least one LC topical subject heading. Each LC subject heading had an average of 0.78 subdivisions associated with it. Form subdivisions were the most common type found, followed closely by place and topical subdivisions. Period subdivisions were used relatively infrequently.
As a small to medium-sized library, Carnegie-Mellon University Libraries' approach to library automation was to aim for an integrated system. It was decided to label all items with machine-readable identifiers and enter the unique identifier into the data base at the time each record was created. The rationale for these decisions is discussed. Different types of machine-readable identification systems are reviewed. The type of labels chosen and the procedures used for labeling items and for entering the barcode information into the machine-readable record are discussed.
This paper speculates an directions for research in the field of bibliographical control, where bibliographical control is taken to include indexing, classification, and cataloging. The approach takers is to consider questions in the field that need answering. The position taken is that while concerns of a how-to-do-it nature drive this field's research, which is of an evaluative or developmental nature, there is a strong need for this research to be hacked by basic theoretical research.
This paper is an unsavory mixture of library automation literature and science fiction. The author attempts to fill the ecological vacuum to be caused by the disappearance of the card catalog by the invention of the ultimate amiable user interface. For lagniappe, the author embeds this dream console in a framework of underground bibliotaphic libraries connected by broadband communication networks and bicycle messengers.