Posting to Autocat
Serials are a very delicate matter and I think, should be reconsidered in their entirety. The major/minor change rules come from the latest ISBDs (available through the CCR Wiki at:
http://sites.google.com/site/opencatalogingrules/isbd-areas, all under 0.2 Treatment of Resources and 0.2.4 Changes requiring a new description: for continuing resources).
There is a real problem here balancing the needs of maintaining coherent library inventories with the needs of the users. In my own opinion, which is certainly not shared by everyone, latest entry was easier for non-specialists to grasp, since they had a nice overview of the entirety of the serial, seeing how it changed over the years, while successive entry broke everything up, at times making it difficult even for an expert cataloger to get an idea of the whole of a serial.
Yet, in a networked environment it becomes extremely difficult to share records based on latest entry, because each library will have vastly different holdings. If the idea is to share serial records, a library must have an option other than only taking the latest entry, since it would mean, in effect, major editing not only of each library's holdings, but of each bibliographic record, thereby defeating the purpose of sharing records. Thus, we have successive entry, which is designed to provide libraries this option, although as I said, I feel this situation becomes far more difficult for our users.
Are there any options today? I think there are. New formats, such as XML, allow brand new displays never available before, which can perhaps give the best of both worlds, allowing complete nested displays that the user could interact with, such as drop-down menus showing descriptions of earlier titles. But a more radical approach would question even more: how many *library users* want an entire serial? I would bet that 99% or more want individual articles from the serials, or they may be interested in knowing what is in the latest issue of a serial. It is very rare that anyone would have a practical need to know the entire history of the Atlantic Magazine, other than a cataloger or selector for inventory purposes. Lots of people want individual articles from the Atlantic, or they may want a thematic issue if they know about it.
In the printed world, to get an individual article, people need to know the physical location where a journal is shelved; therefore, they need to know something about how serials work in a catalog and how they are placed on the shelves. With virtual materials, more and more systems are being built with a link from the article metadata straight to the journal article and the user needs to know practically nothing about the serial as a whole. I agree there is a loss and it has profound implications in several ways: think of Google News and its impact on newspapers. Newspaper publishers are anguished about how their newpapers are being losing their identity. But this is exactly how Lexis-Nexis works, and most other online databases: the individual journals and individual issues disappear almost completely. Academic journal publishers are saying similar things as the newspaper publishers (although I can't find any examples at the moment, I have read them) Nevertheless, we all know that this is happening, and the direction seems to be toward more disintegration, not less.
When we are discussing the *needs of the users* and not the *needs of library inventory* (both equally important), in a world where information is disintegrating into smaller and smaller chunks, I think we have to seriously consider what *we* need as librarians. Serial records will more and more be there for librarians and inventory purposes, not for users.