Posting to NGC4LIB
Ed Jones wrote:
This assumes that the “place of publication” element fulfills these purposes. It only does so if ALL places of publication are recorded (which can run to dozens for publishers like Oxford University Press that like to have lots of places of publication) in an unambiguous manner. Since the place of publication is a transcribed element, I’m doubtful on these points (especially purposes dependent on reliable machine manipulation). For example, how many people know that “Harmondsworth, Middx” (long the place of publication for Penguin Books) refers to a place in the London borough of Hillington? Taylor & Francis, on the other hand, is located in Abingdon, Oxfordshire, but you would never know it from their title pages.
At any rate, I didn’t come here for an argument, just to register my amusement–at my age, this is probably healthier–that the code-makers consistently try to make this element optional and are just as consistently thwarted. If evidence were needed that there was a vocal cataloging constituency for retaining place of publication, that would suffice. My suspicion remains, however, that this support is more a matter of inertia: “We’ve always done it this way, and we must have had our reasons”. The arguments put forward now seem more ex post facto and fatally assume this element is formally controlled in some way when it’s merely transcribed.
I would like to emphasize: “If evidence were needed that there was a vocal *cataloging* constituency for retaining place of publication”
Well, to be honest, *as a cataloger* I couldn’t care less about place of publication, so I really don’t care if this matters to *catalogers* or not. I would just as soon leave it out because there would be less for me to do. This statement might make some on the list angry, but we should never lose the fact that the records we make are for the *patrons*, including not only the public who use our collection, but just as importantly, library selectors who need more specific information than the public so that they can determine if certain texts already exist in the collection.
Rather, the question should be framed as: exactly to whom is it that adding “place of publication” matters? Does having place of publication matter to anyone at all? And if it does matter, how important is it? Also, this information may be possible to be reused for other purposes, as I mentioned with determining copyright. In any case, in former times it was seen as important, and I think that importance can be demonstrated beyond any real doubt for those materials, in that it did matter for those who were interested in highly specific editions. Today, it is important to determine if it still matters to others.
I have mentioned some instances where as a *user*, place of publication could matter to me; others have mentioned some other points, and I am sure others can bring up still more. *If* we have any hope to cooperate with other metadata communities, we must *all* change our practices, not only the library community, or any other single community. Whenever one community must bend completely to the wishes of another community, it is not called cooperation, but subservience. The library world of metadata should not be subservient to ONIX and vice versa, while there are many other metadata worlds out there. We need cooperation.
Perhaps it would be decided that adding place of publication is not worth the effort. OK, I can deal with that and we could dump it as essentially useless. In that case, good riddance! But the interested constituencies should not at all be limited to catalogers, whose responsibilities are limited merely to entering (or not entering) information that is one of the simplest parts of the record.
I still believe that the Wikipedia principle of “disambiguation” should be considered in cataloging matters, and not only to subjects: this principle is to enter additional information only when necessary to “disambiguate” something. This implies that work is done retrospectively as well. Place of publication could be one example: add place of publication only if it is necessary to disambiguate, and apply this information retrospectively.
If the OCLC statistics are correct, fewer than 20% of records would need to be disambiguated/reworked at a later time (and perhaps far less than that in practice), and this seems as if it could be a genuine savings.