Some librarians have said that they will not adopt the MARC II format because it doesn't accommodate all of the information they wish to record. While we've usually heard the complaint that MARC includes too much, we believe there is a solution for those who want more or different information.
The concept of the MARC II communications format is the inclusion of specified information on the tape leader and the tagging of each data element with a specific tag that will match the same data elements in other files maintained on other systems. Should an organization wish to include data elements not provided for in the MARC II format, it is possible to use "vacant' numeric tags or to add alphabetic tags which are distinguishable from those used in MARC. The important thing is to have the elements which are commonly used for searching and matching records-author, title, series, subject-properly tagged so that the records will be compatible with any others created using the MARC "standard." In the data processing context a standard is a set of protocols which have been adopted by several institutions so they can exchange data and programs-it is not just internal consistency. An organization which creates a non-MARC data base closes off not only avenues of future software and computer service support, but also handicaps future data sharing among libraries.
To minimize this limitation, libraries which opt to "extend" the MARC format to meet their individual needs must invest a degree of effort in developing, implementing and documenting the use of the non-standard fields. Consistency and documentation are essential prerequisites for future processing of the records as might be required for the creation of a shared data base or the output of a COM catalog.
The importance of documenting local, nonstandard expansions of the format has been stressed in several recent forums exemplified in Dennis Reynolds' article on the entry of local data on OCLC in the March 1982 issue of Information Technology and Libraries and the activities of groups such as the DataPhase Users' Group's establishment of a Bibliographic Standards Committee to advise on inconsistencies among the MARC records output by the different utilities.
