I sense a tension here between desire for "simplification" and "user language" (e.g. "book"), and well thought out properties and relationships.
I'm hoping we can have the latter in BIBFRAME, but be able to derive the former.
The MARC21 leader codes appear to represent the latter. They appear to have definitions that are plainer than fuzzy - that is, something is or isn't language material, although it might have other characteristics.... something is monographic or not (although sometimes publishers surprise us and we have to reassess). From experience, there are always odd instances that may seem fuzzy when we encounter them.... but...
..."book", on the other hand, is subject to interpretation. If it's "physical", is it a book if it doesn't come in a binding of some kind? Is it a book if it's only 6" tall and has 23 pages? (or is it a pamphlet?). Is it a book if it
has no text, only images, but is in a book binding? If it's "e-", what makes it a book vs. an article or essay? The number of "pages"? The fact that it has a title-page? The fact that it has chapters? What if the chapters are by different people and
have different titles, and are presented as, say, separate PDFs?, with links on a web page that has a title, such as "Proceedings of such and such a conference"... ? Is it more or less a book than if all the chapters are presented in the same PDF?
I think if our classes were defined well, a particular application or use of BIBFRAME data could say "if a work is a member of "language material" and its instance is a member of "paged text" and its number of pages is greater than 20, it's a" book"; if the instance is "e-" instead of "paged text", it's an "e-book"....
....and a different application could have different criteria (maybe number of pages doesn't matter, or has to be 30, etc.), and could call their class (or type) of resource (or work), "book", or"Tome" or something else....
There could be a standard "interpretation layer" (set of queries, scripts, even stylesheets for RDFXML) that could generate additional categories based on reasoning - IF we have reasonable "types" or categories.
This would carry forward what happens in many "discovery layers" or online catalogs, which translate things like leader codes in MARC into those little book, CD, etc. symbols (and also facets, drop-down selection lists, etc.). Only we could have much more leeway in creating categories. If some institutions want to use criteria to label something "pamphlet" (and have data to make that meaningful), they can; if others don't, they don't have to.
One more thought - I deliberately used pamphlet as an example because I think extent can be a component of what people think of as "type". I also think extent is something end users want to know about, but our ways of telling them about it for e-stuff aren't great... telling someone how many bytes something is doesn't help when they want to know "is it book-length, article-length, or blog post-length?"
Purpose and context can also be part of a "what is it" typing - a blog post and an article in a peer-reviewed journal might contain almost exactly the same text... but I think users would want the distinction made clear...
I'm not sure I'm using the current or proper terminology for any of these things (concepts) but do you get what I mean?
First, those first two respond with a request for authorization, so most of us can't see what is/was there.
Second, "Book" is also not a "class" in MARC21 (and I'm not sure about RDA). There is:
LDR 6 = language material
LDR 7 = monograph
007 0 = t ("text")
So we don't have a clear "book" designation now -- in spite of the fact that in many cases users enter the library in search of precisely that: a book. An essay published online is monographic language material. In fact, I wonder if an LL Bean catalog isn't monographic language material? And yet, most libraries interpret those leader codes with a book-like icon next to the entry in the catalog, and it's right most of the time, even though the standard never uses the word "book."
p.s. I thought that the LDR 6 had been redefined to "language material" from something else that was more book-like, but it doesn't show in the documentation.
On 2/5/13 6:05 PM, Young,Jeff (OR) wrote:
In the original MARCR mockup, “Book” was defined as a sensible class:
This was carried over into the BIBFRA.ME mockup:
In the BIBFRAME.ORG mockup, though, “Book” appears to be gone:
What’s going on?
Jeffrey A. Young
OCLC Research, Mail Code 410
OCLC Online Computer Library Center, Inc.
6565 Kilgour Place
Dublin, OH 43017-3395
Voice: 800-848-5878, ext. 4342
Email: [log in to unmask]