 |
 |
The relationship between RDA and BIBFRAME needs a closer look by the PCC. PCC adopted RDA for authorities in 2013 and for bibliographic data in 2015. PCC has also promulgated many RDA guidelines, as well as the BIBCO RDA and CONSER RDA standard records. BIBFRAME is a new way of representing and exchanging bibliographic data. One of its principles is that it "generally aims to be independent of any particular set of cataloging rules" (http://www.loc.gov/bibframe/faqs/#q04). In other words, BIBFRAME includes many but not all RDA elements, and by design will not include them all. As an example, consider support for relationship designators in RDA and BIBFRAME: a mapping of Work designators from RDA to BIBFRAME (http://www.lib.washington.edu/msd/pubcat/ld/resource-work) shows that a great deal of specificity is lost in BIBFRAME. PCC support for both RDA and BIBFRAME creates a tension that must be resolved. If, on one hand, RDA is the PCC content standard, why should RDA data be encoded in a scheme that irretrievably loses many intellectual distinctions made by the cataloging code? If, on the other hand, BIBFRAME is the PCC encoding standard, why should members pay catalogers to make distinctions that are immediately discarded? To my mind, our content standard is of primary importance. Its rules embody the intellectual added value that catalogers bring as they make sense of a complex bibliographic world. Once a content standard is chosen, then a linked-data encoding scheme should be selected that is capable of handling the full range of elements in the content model. It is the content model that is chosen first, not the encoding scheme. In my view, PCC needs to make a commitment to a content standard and to a linked-data encoding scheme that support completely loss-less data transfer. Note that the new environment does not require the use of a single encoding scheme. Different schemes support different use cases and BIBFRAME has its place. But for the use case of library-to-library exchange of RDA cataloging, PCC needs a loss-less encoding scheme.