Sorry, I wasn't reading e-mail yesterday, so I'll respond a bit late to some of this.

"The series title applies to the aggregate work, not the literary work". First of all, this is an assumption that I am not willing to make. In this case, "The Chronicles of Narnia" is a literary work. It is also an aggregate work. They are one and the same. The series aspect of it applies to both. Or not--whether or not an aggregate work is treated as a series is in many ways an arbitrary cataloging decision. I could choose to catalog all the volumes in a single record, that is, not treat it as a series. Or I could choose to catalog each title separately, that is, treat the aggregate as a series. Or the publisher could choose to publish it all in one volume (that is, not as a series)--as in fact at least one of the NARs in question represents. Choosing to treat an aggregate work as a series or not, I repeat, is an arbitrary and somewhat artificial cataloger imposition. It has no effect whatsoever on the basic nature of the work or expression. Nor does it affect whether the work is an aggregate work or not.

I am indeed recommending we think about a significant shift of how we regard series. RDA and FRBR invite us to look at the bibliographical universe in a different way and I think we should embrace that invitation, putting aside our paradigms and assumptions. Do we need to continue doing things in the old way? In the case of the Chronicles of Narnia are we really doing our users a service by insisting on continuing to line the series up as we did under AACR2 (a practice which, as I pointed out, wasn't even required by AACR2, but by LC's series policies)? Do our users really come in and say "I want to read The Lion, the Witch and the Wardrobe" and I will only accept it if it's in the HarperCollins series, not the MacMillan series? No, they don't. Do they in fact think of these different publisher editions as different series? No, they don't. And rightly so, in my opinion. They quite logically say "I want to read the books in the Chronicles of Narnia", perhaps specifying in English. In other words, they without knowing it are in fact sensing the FRBR organization of the bibliographic universe.

Stephen makes an unquestioned assumption when he says "The series title needs to be qualified to distinguish one series from another and from the literary work itself." Why is that? First of all, the series in this case is the literary work so I don't see the distinction. And why don't we feel the need to distinguish access points for other works by adding their publisher? Users seem to get along quite well without that level of specificity (which as I pointed out is manifestation information anyway--I'm not against having manifestation-level authorized access points, but there's no provision for them and in my opinion no need for them in this instance).

I am far from saying we should abandon series. We do need to take the series nature of some aggregate works into account. What I am suggesting, however, is that the overall structure of the FRBR model is a better, more logical way of organizing these things and more in line with what library users actually need.

Bob