How to get authoritative bibliographic information about journals

Question

When I collect bibliographic material, I seem to get grossly inconsistent information. Papers from the same journal are often linked to different publishers, the same journal or publisher is often stylized differently (e.g. Wiley, Wiley-Liss, Wiley & Sons), location information about the same publisher is different (see What address to put in bibliography for Springer (and others)?).

Maybe I'm a bit obsessive, but I would like to be consistent when citing papers from the same journal/publisher, especially in my thesis.

Is there any source for authoritative information about journals and publishers? Also: Is there a good guide for the semantics of the different fields in a bib entry? If I knew, for example, that what goes into the publisher field should be the registered trademark or the trade name, I could work out some of these on my own.

(BTW: I'm using Citeulike's mechanism for getting bibliographic info for most of the papers I cite.)

Answer 1

I always like to go back to first principles when considering questions like this. The purpose of a citation is to allow another researcher to find the reference in question.

The conventions for citation were developed many decades ago, when documents were always looked up in print, having information like the publisher was critical. These days, we live in an era of search enginers and highly international publishers. That means that:

1. Often there is no such thing as a definitive publisher name or address, because it depends on what perspective you take.
2. There is no unified database for anything, anywhere. If it is valuable, there are many versions of it, and they will contain conflicting information.
3. Authors and title are typically all that's actually needed to uniquely identify a document.

What's typically really important to get right is: author, title, and publication venue (journal, conference, [chapter in] book). Many citation formats won't even show the publisher or address. Likewise, online-only publications like the PLoS journals don't even have a meaningful address or page numbers. For the rest of the fields: don't give incorrect information, but any of the reasonable variants of correct information is fine.