BibLaTeX translator and (in)collection entry type

pprw
I use a lot (in)collection entry type of biblatex.

------
According to biblatex doc :

collection

A single-volume collection with multiple, self-contained contributions by distinct authors which have their own title. The work as a whole has no overall author but it will usually have an editor.

incollection

A contribution to a collection which forms a self-contained unit with a distinct author and title. The author refers to the title, the editor to the booktitle, i. e., the title of the collection.
-----------

These two kind of publication are very (very) common in the academic word (in particular, in human sciences).

According to the biblatex translator (https://github.com/zotero/translators/blob/master/BibLaTeX.js) the zotero entry "booksection" gives "incollection' ("bookSection": "incollection",). However, in my (French) Zotero, I have no booksection entry (just a book chapter which is convert by inbook in biblatex.

So my question is : how can I obtain a "(in)collection" entry by using zotero and biblatex translator? If there is no entry which is actually converted in (in)collection, should we create one?
«1
  • adamsmith
    book section in English and book chapter in French are the same, so this should work. Try updating your translators from the general tab of the Zotero preferences, make sure are not using an old custom version of the BibLaTeX translator (this was changed in December) and try again.
  • pprw
    Thanks, I just updated the translators and, indeed, book section (or chapitre de livre) creates an incollection entry.

    However, I have two questions :

    1. How can I refer to a collection (i.e. a book which is only a collection of articles and then have no author but an editor). This question is, in fact, about what kind of zotero entry is translated into the "collection" entry of biblatex? According to the biblatex translator, there is no zotero entry which will be translated into "collection". But it's a pity because such citations are popular (for example, see [collection of articles in a book] for an introduction to this field ; in this example, we want to refer to the collection as a whole).

    2. There is a difference in French between book chapter (chapitre d'un livre) and book section. A chapter of a book is written by the same author(s) of the entire book and is consistent with a global argumentation. At the opposite, an collection contains articles which are on the same subject but from different authors and not always consistent between them. And there are no author of the book but editor(s) in a collection. So, I think "chapitre de livre" is a very bad translation of "book section" or "collection". In French, we use the world "collectif" for a book which assembled articles from differents authors on the same subject (i.e. a collection). Therefore, I would suggest to replace "chapitre de livre" by " article dans un collectif". And we can create another entry called "chapitre de livre" which will really refer to a chapter of a book (not an article in collection).
  • adamsmith
    edited January 17, 2014
    1. In Zotero this would just be a regular book with creators set to "editors" http://www.zotero.org/support/kb/edited_volumes_and_book_chapters (which is also, FWIW, how other bibliographic data formats such as MARC treat it)
    biblatex (edit: i.e. Zotero's biblatex translator) doesn't turn these into "collection" automatically. Is the difference important for citations? It could be done, just strikes me as odd that it needs to be.

    2. are you sure about this? If I google
    'citer "chapitre de livre"' or 'cite "chapitre d'un livre"' I get a lot of examples for chapters in edited books, both from France and from Quebec. We're not going to add a separate item type for book chapters by the same author - I don't see the need. FWIW, in English a "chapter" or "book section" can also be both by the same author or in an edited volume, so the generic translation seems apt to me.
  • pprw
    Thanks for your quick answer.

    1. I tried to see if there is some differences in the rendering when you use @book instead of @collection with the same informations. There is no difference with standard style (but I didn't try some specific field). So it not a big deal to merge both entry in zotero translator. However, it is still useful when you want to sort your bibliography by type of entries (books, collection, article, ...). If someone want to separate @collection and @book in the biblatex translator a good rule could be : if there is no author, then it is a @collection ; if there is one author, it just a book.

    2. Ok but "chapitre" (at least in French, I don't know if there is the same subtlety for chapter) is a logical separation inside a argumentation (for example the chapters of a thesis) when "book section" is just informal and without special meaning. An article in a collection is not a chapter of this collection. And, in lots of collections, there are thematic chapters with several articles of different authors inside. But, again, it changes nothing in most cases in the rendering with LaTeX.

    The subject of both points is the following question : should we create entries for almost every kind of references or should we be pragmatical and say "an @collection is printed like a @book" so let's just create one entry @book for both?

    BibLaTeX choose to create a lot of entries, which is printed often in the same way but which also indicate precisely what kind of document we are citing. For example, if a reference is cited with @inbook entry, we know it is a book with a proper argumentation (so not a collection) but that the author cites only a part of this book which has some coherence by itself. If the reference is cited with @incollection, we know that the book is not important because the reference is a article which is not part of a bigger argumentation.

    Zotero, at the opposite, choose to be pragmatical and to merge entries which are similar (and get ride of subtle differences).

    In both approaches, I prefer the one of BibLaTeX which is more precise.
  • adamsmith
    In both approaches, I prefer the one of BibLaTeX which is more precise.
    but yet you are here... the fact that Zotero is relatively(!) minimalistic with item types has good reasons - it conforms much better with other bibliographic data, so that importing in Zotero is a much more pleasant experience than into BibLaTeX, it's a better experience in terms of the GUI. It allows for a much broader range of citation styles with reasonable effort etc. So there is a trade-off. If you prefer BibLaTeX's solution to that trade-off, by all means use BibLaTeX. But don't expect Zotero to go that route.

    As for the meaning of "chapitre" in French - do the google search. Look at some of the examples. The distinction you make is clearly not something that French native speakers agree on (the translation of the term in Zotero, btw., was also done by French academics).
  • pprw
    Good points in favor of zotero minimalistic entry types.

    About the meaning of "chapitre" I'm afraid to disagree. I did the google search and have found "only" chapitre in the meaning I describe (but my google is in French). I checked in a good dictionary and a chapitre is "an division of a book which aims to clarify and facilitate the reading and the argumentation" (my translation). This, for me, does not correspond to the definition of @incollection. And, in my academic field (human science), when someone says "I just publish a chapter of this book", it means he participates to the overall project and his chapter is important for the general book thesis. It is something different to say "I just publish an article" (in a journal, in a collection).

    But it is a point of detail and I understand perfectly that this subtle distinction is no good reason to change the translation. No worries.

    Cheers
  • adamsmith
    Quebec:
    http://www.bib.umontreal.ca/infosphere/sciences_humaines/module7/livre-chapitre.html
    http://www.bibliotheques.uqam.ca/infosphere/fichiers_communs/module7/chapitre.html

    Belgium:
    http://www.bsv.ulg.ac.be/santePublique/agps/1-books.php#chaptersEdited (see French labels on the side)

    Suisse francophone:
    http://www.hesge.ch/heg/infotheque/boite-a-outils/guides/doc/guide_ref_v4.pdf (PDF - p.15)

    France:
    http://www2.ujf-grenoble.fr/pharmacie/www/etudiants/pages/these/guide41.htm

    All of these are from university pages. Might be a disciplinary thing, but it's quite clear to me that chapitre de livre is used in both meanings at least by a good number of native and academic speakers.
  • pprw
    Sorry but I don't get your point.

    Your examples show only what is the rules to cite a chapter of a book.

    I suppose your point is to show me that these rules are the same than the ones to cite an article in a collection.

    I don't deny it (even if there could be subtle differences), my point is that an @incollection and an @inbook is not the same thing (even if we cite them in the same way).

    A good example to see that is the case of a book which has chapters and, inside a chapter several articles from different authors.

    In the example :

    Chapter 3 : the light
    Article of X : the sun is beautiful
    Article of Y : the light moon is dangerous

    We should cite the chapter 3 by using @inbook because this chapter is an attempt of the author/editor of the book to structure the argumentation. And we should cite article X/Y with @incollection. In this example, it is technically wrong to cite the article of Y as a chapter of the book (because there are real chapter in this book).

    But like I say before, it a question of principle with no (in fact very few) practical consequences
  • adamsmith
    My point is that all of these list contributions in an edited volume under the label "chapitre de/d'un livre".
  • pprw
    If this edited book is only a collection of articles, these articles should be @incollection entries. These edited books are often published in order to give to students some popular article about a notion or a scientific field.

    But there are also edited books which has a internal coherence and are not just aggregation of articles : @inbook

    It is a philosophical difference.
  • adamsmith
    I understand the difference — although if you check the BibLaTeX manual, @inbook has a kind of questionable status:
    If you want to refer to a chapter or section of a book, simply use the book type and add a chapter and/or pages field. Whether a bibliography should at all include references to chapters or sections is controversial because a chapter is not a bibliographic entity.
    I was just pointing out that "chapitre de livre" is not a bad translation as that is how many citation guides in French refer to both a contribution in an edited volume and a chapter in a mongraph (if you really are going to cite it, which I agree with BibLaTeX is a dubious practice), which is exactly what the book section item type is supposed to reflect.
  • nickbart
    Leaving "philosophical" issues aside, I feel the solution to the question of how Zotero items "Book" and "Book Chapter" should be exported to biblatex is straightforward:

    From a pragmatic point of view, the difference between biblatex "book" and "collection" entries is simply that books have one or more authors, and collections do not (though usually they have one or more editors).

    Similarly, a biblatex "inbook" entry has a "bookauthor" (not necessarily the same as the "author", imagine a piece by Marx/Engels in Marx’s Collected Works), while "incollection" does not.

    Thus, if a Zotero "Book" entry has an "Author", it should be exported as a biblatex "book" entry, if it does not have an "Author", it should be exported as "collection" (without an author, it could also be "proceedings" or "reference", but Zotero will not be able to distinguish these, and since "collection" is most common, this should be the default).

    Analogously, if a Zotero item has a non-emtpy "Book Author", it should be exported as a biblatex "inbook" entry, while if "Book Author" is empty, it should be exported as "incollection".
  • adamsmith
    right, as I say above:
    biblatex (edit: i.e. Zotero's biblatex translator) doesn't turn these into "collection" automatically. Is the difference important for citations? It could be done, just strikes me as odd that it needs to be.
    writing more and more such conditional statements into export translators makes it harder to understand - certainly for an end-user - what is actually happening. That's where why I'd be somewhat reluctant to just keep packing them on.
  • pprw
    A nice and general solution to this problem could be to keep the automatic translator the way it is now *and* add a Bib(La)TeX field into each reference. This field would allow the user to indicate explicitly in which entry type he want his document to be converted. For our example, if we have a book and if we know it is a collection, we can override the automatic behavior by saying explicitly we want this reference to be converted in @collection.

    Of course, this field should be available only for users who check the corresponding option in settings (we can call it "Allow manual definition of Bib(La)TeX translator" and add it in the export tab). If this field is blank, then the automatic rules are apply.

    This could be a way to combine the possibilities of BibLaTeX in terms of entry types and the exigence of simplicity of Zotero.

    What do you think?
  • adamsmith
    no, adding a preference and special customized BibLaTeX behavior is not going to happen - core devs would definitely veto that (and it's not really simpler ;)).

    If this is going to happen it will be along the lines nickbart suggests. The question there is if it's worth it and I'm not clear on that.
  • pprw
    ok I understand.

    The arguments in favor of the nickbarts suggestion (at least the ones I am thinking of) are :

    1. Some customized/not-standard BibLaTeX styles may use the distinction between @inbook and @collection (for example) in order to render differently references. It's a theoretical point because I'm using only standard styles and I have no actual example. However, there should be some examples (don't have the time too look right now) ; in particular, scientific journals have often their own style which could do this kind of difference.

    2. Having @collection and @inbook entries allows you to classify references by types. It is useful, for example, in resume where you cite in sections such as "articles in journals", "articles in edited books", "books", "edited books", ...

    Perhaps there are other arguments.

    The decision is up to you.
  • adamsmith
    2. OK, but at least for a CV, I assume if you had written a book you'd cite that and not individual chapters. That goes back to the fact that BibLaTeX seems ambivalent about @inbook.
    But how about if you cite an introduction witten by someone other than the book's author? What would that be in BibLaTeX (in Zotero that would be a book section with an author and book author)? If we follow what nickbart says, that would be classified as @inbook. Correctly or not?
  • nickbart
    But how about if you cite an introduction witten by someone other than the book's author? What would that be in BibLaTeX (in Zotero that would be a book section with an author and book author)? If we follow what nickbart says, that would be classified as @inbook. Correctly or not?
    Correct.

    As to biblatex’s supposed ambivalence about @inbook, I don't think that's supported by a closer reading of "2.3.1 The Entry Type @inbook" in the latest biblatex manual. This starts with
    Use the @inbook entry type for a self-contained part of a book with its own title only.
    Including these in a bibliography is perfectly legitimate, and seems rather clearly what the Chicago Manual of Style, 16e, is referring to in "14.111 Chapter in a single-author book".

    What the biblatex manual calls "controversial" in the passage that follows,
    If you want to refer to a chapter or section of a book, simply use the book type and add a chapter and/or pages field. Whether a bibliography should at all include references to chapters or sections is controversial because a chapter is not a bibliographic entity.
    only refers to parts that are not "self-contained" - and here I'd agree that references to, say, a single chapter from a novel, or some section that does not have its own title should probably not be included in a bibliography.

    As to the question whether an improved biblatex translator is worth the effort, my answer is yes, the main argument being that some of the common biblatex styles make a distinction here in that they render the same entry in different formats depending on entry type: standard biblatex prints "bookauthor" only in inbook, not in incollection, as does biblatex-dw, and biblatex-apa output differs between book and collection. So I do think the - hopefully small - effort would be justified.
  • pprw
    @adamsmith
    2. OK, but at least for a CV, I assume if you had written a book you'd cite that and not individual chapters. That goes back to the fact that BibLaTeX seems ambivalent about @inbook.
    Yes but if I wrote books and edited books, I would like to have two different entry types in my CV (LaTeX @book and @collection) because it's a very different job to edit a book (write introduction and gather texts from several authors than to write the entire book just yourself.

    In the same spirit, if I wrote a chapter for (to take an example), a handbook for students, it is something different from an scientific article. I probably tried to be clear and impartial in the presentation of the subject ; perhaps I didn't defend my particular position because it not the place to do so in an introduction to the field. Therefore, I would like to emphasize that this chapter is a part of a more global introduction in an handbook and I will use @inbook (because my chapter has an internal coherence). So, in my resume, I would separate chapters of books where I force myself to stick to the general purpose of the edited book (@inbook) and articles published in edited books (but which could be also published by journals) which aim not to be coherent : a lot of famous articles (previously published in journal) are re-publish in anthology : it is what I have in mind.

    @adamsmith
    But how about if you cite an introduction witten by someone other than the book's author? What would that be in BibLaTeX (in Zotero that would be a book section with an author and book author)? If we follow what nickbart says, that would be classified as @inbook. Correctly or not?
    Correctly.

    I agree with nickbart previous post.
  • andersljohansson
    It has been interesting to follow this discussion. I (as the creator of the BibLaTeX translator) excluded inbook for the obvious reasons that there were no Zotero equivalent and that it was "controversial" according to the BibLaTeX manual.

    If you think it should be done it could be quite easily added the same way that mvbook is now (through a check for "volume").
  • pprw
    I would definitively recommend to allow the possibility to create @inbook entry. @collection entry type is also very important (perhaps more) for me (for the reasons mentioned above).

    @collection could be created by a check for "collection" inside a book entry in zotero. If this check is activate, then the reference will be translated into @collection, if not, into @book.

    @inbook could be created by a check for "not in collection". If this check is activate, than the reference will be converted into @inbook, if not, into @incollection.

    This solution is similar to my proposal to create an option which aims to indicate in which entry type you want to convert this reference (see above, pasted below).

    A nice and general solution to this problem could be to keep the automatic translator the way it is now *and* add a Bib(La)TeX field into each reference. This field would allow the user to indicate explicitly in which entry type he want his document to be converted. For our example, if we have a book and if we know it is a collection, we can override the automatic behavior by saying explicitly we want this reference to be converted in @collection.

    Of course, this field should be available only for users who check the corresponding option in settings (we can call it "Allow manual definition of Bib(La)TeX translator" and add it in the export tab). If this field is blank, then the automatic rules are apply.

    This could be a way to combine the possibilities of BibLaTeX in terms of entry types and the exigence of simplicity of Zotero.

    What do you think?
  • adamsmith
    edited January 20, 2014
    again, there won't be any checkmarks/boxes or the like, especially since they wouldn't have any functionality in Zotero proper. While it's great if people find Zotero useful to work with BibLaTeX, Zotero is not a BibLaTeX frontend.

    @anjo - I think given what nickbart says, adding @collection and @inbook makes sense.

    So far my understanding is that a Zotero "book" item should be @collection if and only if it has both
    1) no author and
    2) one or multiple editors

    and a Zotero "bookSection" should be @inbook if and only if it has
    1) one or multiple authors and
    2) no editors
    3) technically also: a title - though basically all Zotero items do, so I think you can skip testing for that.

    (not sure about 1) here) does that make sense?
  • pprw
    edited January 20, 2014
    @adamsmith I was only talking about "check" because anjo7539 mentions it and I didn't think about boxes (perhaps we misunderstood ;English is not my native language).

    For @collection : ok (no author and editor(s).
    For @inbook : we need a bookauthor (not just author(s)) like nickbart sayed.
  • adamsmith
    anjo uses "check" as in "test for the presence of a variable" when exporting.

    For @inbook - what would a book section without a book author be, then? Wouldn't that be better as @inbook than as @incollection?
    Especially since Zotero doesn't collapse author and book author, i.e. to generate a citation like:

    Phibbs, Brendan. “Herrlisheim: Diary of a Battle.” In The Other Side of Time: A Combat Surgeon in World War II, 117–63. Boston: Little, Brown, 1987.

    (CMS 14.111) you would want to add a book section without a book author.
  • pprw
    edited January 20, 2014
    You are right.

    So I guess @inbook is defined by one or several authors and no editors.
  • nickbart
    edited January 21, 2014
    For @inbook - what would a book section without a book author be, then? Wouldn't that be better as @inbook than as @incollection?
    From a systematic point of view, the presence or absence of a book author is the distinctive feature between inbook and incollection (the presence or absence of an editor does not matter).

    I do see the point that a book author must not be printed if s/he is identical with the author, so currently, if only to get the output format right, it should better be left out in Zotero.

    Since the main difference, in biblatex, is that @inbook can print a bookauthor, and @incollection cannot, I do not see a problem in following the systematic approach and translating *all* book sections without a book author to @incollection (if there is no book author in the first place, it doesn't matter that it cannot be printed). As a sort of rescue action, book sections with just an author, and no book author and/or editor could of course be translated to @inbook.

    However – and that's not just motivated by the translator issues we're discussing – I'd like to propose a more general solution, and that is adding a citeproc routine that collapses author and book author if they are identical, very similar to the existing one that collapses editor and translator if identical.

    Collapsing would have two main benefits: It would require no user interaction and would thus reduce errors, and users could even systematically include book authors to highlight, at a glance, the fact the that a given book section in the database belongs to an authored (as opposed to an edited) book. And, of course, all book sections from authored books, provided the book author is included, would be translated correctly to biblatex @inbook.
  • adamsmith
    From a systematic point of view, the presence or absence of a book author is the distinctive feature between inbook and incollection (the presence or absence of an editor does not matter).
    why? BibLaTex has
    "Required fields: author, title, booktitle, year/date"
    for inbook. bookauthor is listed as optional.
  • adamsmith
    but yes, bookauthor is sufficient for @inbook, so we can (and should) just test for that and automatically do
    if (bookauthor) --> @inbook
    the question is what we do with the rest.
  • andersljohansson
    I'm working on implementing these improvements (along with some other discussed elsewhere) now and just want to confirm what you think about what has been discussed here.
    Are these the two rules that you think should be implemented to get support for @collection and @inbook (or did I miss something)?
    A Zotero "book" item should be @collection if and only if it has both
    1) no author and
    2) one or multiple editors
    A Zotero "bookSection" item should be @inbook if it contains a "bookauthor" field.
  • adamsmith
    yes, I think that's right.
Sign In or Register to comment.