11

The documentation of bibtex states that 'Publisher' is a mandatory field for a book entry. The bibliography manager jabref and bibdesk follow this rule.

In the case of biblatex the documentation says that it isn't.

I am not an expert in official citation policies, however, IMHO publisher should be mandatory. Why does biblatex differ from bibtex/jabref/bibdesk?

Related questions:

Community
  • 1
Hotschke
  • 5,300
  • 5
  • 33
  • 63
  • 2
    Why should the publisher be mandatory? You can easily find a book with its title/author/year... – samcarter_is_at_topanswers.xyz Feb 02 '17 at 13:03
  • 2
    There may be books for which the "publisher" is either unknown or irrelevant, e.g., self-published books. Just out of curiosity: Why do you care whether the publisher field is declared to be mandatory or optional? It's not as if biblatex will discard the publisher field if it's non-empty... – Mico Feb 02 '17 at 13:05
  • Well, apparently bibtex suggests that these should be mandatory and warns the user about it. Journal citation style guides show only examples with the publisher. My intention is to use biber with validate-datamodel to check if I have incomplete bib entries. This does happen. – Hotschke Feb 02 '17 at 13:23
  • 1
    Regarding unknown publisher or if a book is only published by yourself: I would highly appreciate it if a reference list would tell me about it. This is a clear indication that the book might be hard to get . This is valuable information. – Hotschke Feb 02 '17 at 13:31
  • Regarding identifying a book by author, title and year: this also applies to journal articles (except people submit the same work more than once), but the journal name is mandatory for articles. – Hotschke Feb 02 '17 at 13:37
  • 2
    The standard biblatex styles do not need publisher to print a citation or the bibliography. So it's technical optional. You can easily define your own style that adds tests for empty field that should be mandatory in your opinion. – Schweinebacke Feb 02 '17 at 13:38

1 Answers1

17

It's important to remember that 'required' for a classical BibTeX style means broadly 'The output might look odd without this field': BibTeX is not really trying to do data validation!

When Philipp Lehman wrote biblatex he looked carefully at what was and wasn't common in bibliographies and tried to make a sensible set of standard styles. He also set up biblatex so that in the main missing fields (even 'required' ones) don't give badly formatted output, although from the point of view of looking up references it may be sub-optimal. The 'required' fields are therefore best viewed as a suggestion for the minimum set that are needed to have any kind of 'reasonable' output, and are deliberately kept to as small a set as possible.

If you wish to do data validation, you really need to look at a range of issues not limited to the simple question of 'which fields have some content'.

Joseph Wright
  • 259,911
  • 34
  • 706
  • 1,036
  • I hope this covers some of the issues: it's as much a question of the ideas as anything technical. – Joseph Wright Feb 02 '17 at 13:49
  • Thanks for your answer. Also the comment by @Schweinebacke was helpful to understand the motivation by biblatex how to choose mandatory fields. – Hotschke Feb 02 '17 at 14:25
  • 3
    I think it was interesting that Biblatex earlier let "incollection", "inproceedings" and "proceedings" have a required field "editor", even though that was optional for the same types in BibTeX. Of course the standard classes handles works without this fine, and it was made optional after I reported it. Having the set being kept as small as possible is good! – pst Feb 02 '17 at 15:56