On the topic of improving the idea of "Extent," this discussion paper is on the right track:
http://www.rda-jsc.org/docs/6JSC-ALA-Discussion-1.pdf The main problem has its source in cramming too many overlapping ideas into the 300$a field. There are different things being counted. Extent of Expression and Extent of Manifestation are the first distinctions that should be made. Extent of Notated Music is unabashedly an expression level measurement as the terms are pulled from the expression element in RDA 7.20.1.3. Cartographic resources and still images often don't have the same measurement as the number of carrier units (as in "1 atlas (2 volumes)" or "1 print on 24 sheets"). The norm for Extent should be the number of carrier type units, accompanied by carrier subunits as appropriate: Carrier type: audio disc Extent: 3 audio discs Carrier type: filmstrip Extent: 1 filmstrip (28 frames) I do have an issue with Extent of Text, in that this measurement shouldn't be associated just with text. The other problem is that pagination subunits aren't just associated with physical volumes either. Consider the example in RDA 3.4.1.7.1: "1 computer disc (xv pages, 150 maps)" or in RDA 3.4.1.7.4: "3 microfiches (1 score (118 pages))". For those reasons I would treat Pagination as a new independent element under Extent of Manifestation, to be used wherever it is appropriate. To make this work one would have to count out every Extent measurement. To recreate the classic catalog card display as found in 300$a, one would have to follow rules and/or algorithms to collapse some measurements into the original compact displayed form. So for example, a book would be: Carrier Type: volume Extent of Carrier: 1 volume Pagination: xiv, 383 pages Traditional display: xiv, 383 pages But where the units of extent draw in the Carrier Type (from RDA 3.4.5.17), the logic of this arrangement becomes more apparent: Carrier Type: volume Extent of Carrier: 3 volumes Pagination: xx, 300 pages Traditional display: 3 volumes (xx, 800 pages) Such a clean and logical separation would do wonders. Consider atlases in RDA 3.4.2.5 in this way: 1 atlas (1 volume (various pagings)) would be encoded as: Content Type: cartographic image Extent of Cartographic Resource: 1 atlas Carrier Type: volume Extent of Carrier: 1 volume Pagination: various pagings where Extent of Cartographic Resource would be under a new Extent of Expression element. Consider notated music in this way: 1 score (viii, 278 pages) Content Type: notated music Extent of Notated Music: 1 score Carrier Type: volume Extent of Carrier: 1 volume Pagination: viii, 278 pages Another example of multiple things being measured-- here we see Extent of Manifestation, Extent of Expression, and Pagination all together: 3 microfiches (1 score (118 pages)) Content Type: notated music Extent of Notated Music: 1 score Carrier Type: microfiche Extent of Carrier: 3 microfiches Pagination: 118 pages The other benefit to treating Pagination as a separate element is that it's unique in that the measurement isn't usually based on the actual number of pages, but on the recording of the last numbered page. Thomas Brenndorfer Guelph Public Library