impossible to cite translated work

hi,

will it ever be possible to cite something like this (just a made up example):

Freud, S., 1900, Traumdeutung, Leipzig: Deuticke (transl. The interpretation of dreams, 1999, New York: The Free Press).

I read somewhere that Zotero is not that "internationally friendly", but frankly speaking the fact that you cannot cite the original title, publisher and place of publication of a reference is a shame and an unforgivable limitation for any serious scholar.

lch
  • Your basic point is reasonable, but the rhetoric ("unforgivable"? did you really write that?) is a bit much. It's not the most straightforward thing in the world to do right, but I do believe they're working on it.
  • i did write that.
    it is unforgivable to me because:
    1-it prevents me from using zotero
    2-it shows an english-is-all-that-counts approach which is both so outdated and self-centered to be embarrassing
    3-i am not a native english speaker and i hardly know what 'unforgivable' means ;)
  • As another non-native speaker I think that especially point 3 makes your initial choice of words forgiveable.
  • it shows an english-is-all-that-counts approach
    No, it really doesn't.

    It shows the current lack of a complicated, planned feature with wide-ranging implications across many parts of Zotero and any software or system that gets data to/from it.

    But perhaps you'd like to chime in on the thread where it's being discussed? Incidentally, I asked several questions on that thread nine months ago and didn't get a response until a few weeks ago, from Bruce...
  • edited November 28, 2008
    3-i am not a native english speaker and i hardly know what 'unforgivable' means ;)
    OK, fair enough.

    Just to clarify what I meant (as a native speaker), to forgive someone or something suggests that they have violated some moral or ethical code. In this context, I take "unforgivable" to suggest that Zotero has deeply offended you such that you would never use the software again, even if they were to fix the problem you've identified.

    As Dan hints at, for those of you that care about this functionality, you need to go beyond just complaining and help to solve the (difficult) issues involved in actually implementing it. This is NOT just a question of adding some fields; it's much more than that that. Dan and I both proposed some ideas of how to do that, but nobody with real expertise in this area has commented on whether these are good ideas.
  • @Dan
    "It shows the current lack of a complicated, planned feature with wide-ranging implications across many parts of Zotero and any software or system that gets data to/from it."

    i have no competence in software, so i would not know what to say about this.
    i notice that all reference software i know of have this functionality -- and i take it
    that's the case exaclty because it is a VITAL functionality to many.

    @bdarcus
    "to forgive someone or something suggests that they have violated some moral or ethical code. In this context, I take "unforgivable" to suggest that Zotero has deeply offended you such that you would never use the software again, even if they were to fix the problem you've identified."

    oh no, i did not mean that. i will be so glad to use zotero when this issue is solved!
    as i said, it is only that i do not understand how some software can go so far and forget something so basic as this.

    to my understanding, the lack of this functionality prevents any non-english scholar writing to any non-english readership from using zotero. especially in the humanities, but not only, nobody would even consider not citing original title-place-publisher.

    and even within the anglo-saxon world, there are many fields where that would be not acceptable.

    @bdarcus
    "As Dan hints at, for those of you that care about this functionality, you need to go beyond just complaining and help to solve the (difficult) issues involved in actually implementing it."

    i am sorry i cannot be of any help in this regard. i know nothing about software.
  • i am sorry i cannot be of any help in this regard. i know nothing about software.
    You don't really need to; it's just basic logic that then gets translated into code. Dan knows how to do the second; the more basic issue is the first, which you can (and should) help with.

    Here's the way I understand the issue now:

    1. all resources have titles

    2. titles may be translated or transliterated

    3. resources may themselves be translated or transliterated

    Your case is 3.

    Remember, Zotero is designed not just for individual users, but with the goal of being a social networking tool for scholars. So imagine if you have multi-lingual communities sharing resources. The trick is how to keep all these distinct such that in any given language context, you know how to present the different titles.

    My answer is:

    1. all Zotero items have a title and language field, each of which always share the same language

    2. all Zotero items may have zero-or-more "alternative titles" fields of some sort to indicate translated or transliterated titles, each of which can have languages attached to them

    3. Zotero items can be linked, such that one can easily indicate that one item is a translation of another; in other cases you'd use the same mechanism to say item A is a review of item B, etc.

    So you're case is the last one. Am not sure exactly how one would do the linking, but from the perspective of the database, you have two separate entries: the original German version, and the English translation.

    Does that sound sensible?
  • edited November 28, 2008
    " 3. Zotero items can be linked, such that one can easily indicate that one item is a translation of another; in other cases you'd use the same mechanism to say item A is a review of item B, etc.
    So you're case is the last one. Am not sure exactly how one would do the linking, but from the perspective of the database, you have two separate entries: the original German version, and the English translation."

    yes, that makes perfect sense.
    this would be a very good way to go about it.

    another way -- perhaps simpler from a software point of view -- would be to allow the end user to manually insert the missing data about the original publication (and modify the output style accordingly).
  • @lorgg

    Just so you know, it's no problem already to edit the actual citation to include just the extra information you want. But this isn't automatic, and you'd have to do it manually for each citation, and no one thinks it's a full solution.

    And as to the larger issue:
    Bruce, the above solution sounds good as far as what data and connections get stored. I assume that the association mechanism would include the capacity to associate more than two editions with each other. (e.g. an original German edition, the first English edition [which introduced the concepts to the Anglophone world], and the more recent reprint I made use of). I can't think of anything more that I'd need for my own work, which does involve this sort of thing.

    It would be nice to add to that a way to give the user considerable choice in what actually gets cited in each individual case. Just because I have original edition publication information, doesn't mean I want it included in every context. Perhaps there is a 'per document setting' (use original [chinese] titles and include Pinyin transliterations where available) which can be overridden on a per-citation basis.

    The other (practical, simple) way I've seen two different editions referred to is by citing a double year only (Freud 1999[1900]). This gives the reader the year of the edition I cited (essential, to find it) and the year the original came out (important for other reasons). Not sure how common this is, but I'd be glad to be able to do it.
  • edited November 28, 2008
    @scot:
    the above solution sounds good as far as what data and connections get stored. I assume that the association mechanism would include the capacity to associate more than two editions with each other. (e.g. an original German edition, the first English edition [which introduced the concepts to the Anglophone world], and the more recent reprint I made use of).
    I can't see any way to do that without a more complicated FRBR-like model. In that case, you might have the work, and then different expressions of that work, linked together like a tree.

    But that sort of model is probably too complicated for Zotero. In what I'm proposing, you'd have two distinct relations, where each English version points to the same German original. So there would be a link between those latter two; it would just be inferred, rather than direct.
    It would be nice to add to that a way to give the user considerable choice in what actually gets cited in each individual case. Just because I have original edition publication information, doesn't mean I want it included in every context. Perhaps there is a 'per document setting' (use original [chinese] titles and include Pinyin transliterations where available) which can be overridden on a per-citation basis.
    This is an interesting and complex problem that I think we'll have to come back to ;-)
    The other (practical, simple) way I've seen two different editions referred to is by citing a double year only (Freud 1999[1900]). This gives the reader the year of the edition I cited (essential, to find it) and the year the original came out (important for other reasons). Not sure how common this is, but I'd be glad to be able to do it.
    This is indeed very common, and part of why I'm proposing this solution. CSL already has support for this notion of "original" title, etc.
  • Zotero items can be linked, such that one can easily indicate that one item is a translation of another; in other cases you'd use the same mechanism to say item A is a review of item B, etc.
    Right. This is, of course, the "semantic relations" support we've discussed previously, and it's also related to the fabled future "hierarchical item type" support, which we've talked about as a separate feature but which may just be a way of describing special UI and CSL handling for certain predefined semantic relations.

    In terms of the UI, it will probably be implemented in such a way that the referred-to item could be hidden—it would exist in the database but not necessary show up separately to the user by default (unless it was added first as an independent item or manually detached).

    We probably won't have time to work on this until after syncing support is stable, however, since there are some tricky issues to work out here. (For example, things as seemingly simple as item deletions become more complicated.)
    another way -- perhaps simpler from a software point of view -- would be to allow the end user to manually insert the missing data about the original publication (and modify the output style accordingly).
    This would indeed be simpler, and as far as I know it's how most other bibliographic software handles this. We haven't implemented this as a stopgap solution because 1) we haven't been able to make any changes to the Zotero data model in Zotero 1.0 for the last year due to database instability issues in Firefox 2 (fixed in Firefox 3), and 2) it's not really a good solution, and it'd be better to come up with a more elegant, flexible system first rather than having to deal with data migration issues down the line.
  • "In terms of the UI, it will probably be implemented in such a way that the referred-to item could be hidden—it would exist in the database but not necessary show up separately to the user by default (unless it was added first as an independent item or manually detached)."

    I think that makes a lot of sense. In most cases people don't really need a second entry for the original or translated version but really only need that data in one single entry. To put this in one entry, probably with the possibility to split it into two different entries, would certainly be a good idea.

    "We probably won't have time to work on this until after syncing support is stable, however, since there are some tricky issues to work out here. (For example, things as seemingly simple as item deletions become more complicated.)"

    So are we still talking about version 1.5 here, or will this hierarchical stuff come in a alter version?
  • @scot
    "Just so you know, it's no problem already to edit the actual citation to include just the extra information you want. But this isn't automatic, and you'd have to do it manually for each citation."

    how do you do that?
  • In Word or OpenOffice, click on the citation or bibliography and click Edit Citation or Edit Bibliography from the Zotero toolbar.
  • do you mean i will have to edit the citation and related bibliography WITHIN winword, without affecting the actual database?

    i am not sure i understand.

    as an example:

    if in my database i only have the english edition of freud, then it will show (freud 1999) in winword no matter what i do, will it not?

    instead, it should show (freud 1900).

    and how do i get

    Freud, S., 1900, Traumdeutung, Leipzig: Deuticke (transl. The interpretation of dreams, 1999, New York: The Free Press).

    in the inserted bibliography at the end of the document?
  • do you mean i will have to edit the citation and related bibliography WITHIN winword, without affecting the actual database?
    For now, yes.
  • @simifilm:
    Just to put a fine point on the discussion:
    I think that makes a lot of sense. In most cases people don't really need a second entry for the original or translated version but really only need that data in one single entry. To put this in one entry, probably with the possibility to split it into two different entries, would certainly be a good idea.
    How it appears to the user is just a UI question. At the data level, however, it's important that they be represented consistently.

    An aside: it's just occurred to me that there may also be room here for unanticipated advantages; like, say, automatic linking or association.
  • @bdarcus

    I'm not a programmer, so I know how is this best solved at the data level. What I know is that no one wants to create a complete new entry just for a few fields like "Original year of publication" or "Original title". So while this is "just a UI question", I'd say it's still important. For me, as a user, it's actually much more important how this is solved UI-wise than whether this is done with a flat or a hierarchical model. On the GUI level, this definitely should be part of the same entry as long as the user doesn't explicitely wish otherwise.

    I don't know, on the level of development, what the pros and cons of a hierarchical model are. But I know from biblatex, that you can do quite a lot with a flat model in a clean way, but there also are things you simply can't do.

    Would it make sense to mix a flat and a hierarchical model? For example: If you only have one field like "Original year of publication" have it flat, once more data is added, make it hierarchical. Is that doable? Would it actually have any advantag in terms of internal handling or performance-wise?
  • So while this is "just a UI question", I'd say it's still important.
    Sure; am not disputing that. Just wanted to make sure we understand that these are really two separate issues.

    [...]
    Would it make sense to mix a flat and a hierarchical model? For example: If you only have one field like "Original year of publication" have it flat, once more data is added, make it hierarchical. Is that doable?
    Sure, but I'm not sure why one would want to do that; encode data differently depending on how much of it there is.

    The advantage of the more relational/semantic/hierarchical approach is mainly that it's more flexible. For example, what if you're dealing with an original work that was published in Mandarin, and you need both the Mandarin title and its Latin transliteration? You could create fields like "original-title" and "original-title-transliterated," but that starts to get awkward. Even more awkward is if you need to deal with things that are published in other things: say a story originally published in one book, in one language, and then published in another book, in another language.

    Ultimately, as I say, how well this would work would depend a lot on how well the UI was designed, and I could imagine there's a lot of possible ways to implement this idea. Maybe, for example, relations of this sort work something like the current contributors widget: one clicks an "+" button, and can select the relation that gets added (In this case, it would be "original" or "original publication")? Maybe there's some checkbox that allows one to treat (essentially from the UI perspective) that original item as a standalone item? Maybe, too, Zotero becomes smart enough to know the original publication link, an dot provide you hints?

    But then again, I'm no UI designer ;-)
  • edited December 1, 2008
    @simifilm
    What I know is that no one wants to create a complete new entry just for a few fields like "Original year of publication" or "Original title"
    Not sure about that. The original poster, for example, wants basically a full citation of all relevant data for the translation (without repeating the author's name) within the citation of the original. My style requires the same. Besides that, depending on the focus of the project, I'm reasonably likely to consult and use the originals. In that case I'd want to cite them as "original [+translated as...]" like the OP, rather than as "translation [+originally published as...]".

    You're right that it should be made as easy as possible to enter the info, but if you have to store full publication data (and if you need the ability to cite both original+translation and translation+original) than it makes sense to think in terms of a separate but related item.

    In my case it really is no harder to import a whole record, since 90% of the time I'm just going to go to my favorite online library catalog and import all the publication data for the supplementary edition just like I did with the original edition.
  • It does, of course, depend on the citation style, but at least in my field, it's common to cite the original year of publication (and only this) for later editions. So, to take an first example of this thread, Freud would be printed something like this in the bibliography:

    Freud, S. (1999 [1990]): Traumdeutung, Frankfurt a. Main.

    In this case (which is not for translations), the only additional field I need is "Original year". I don't need to know where or by whom the first edition was published. I have dozens of cases like this in my PhD thesis, and having double entries for all theses cases, would be compeltely pointless and just clutter my database.

    But I guess we basically agree. Both ways should be possible. As I said earlier: I think the best would be if some fields like "Original year" would be displayed as part of one entry by default, but would already be part of a new entity in the database. And once you see that you want a seperate new entry, the original could be split, so that the entries which already exist in the database are now als displayed as discrete but linked entities.

    I hope I made myself clear.
  • Sure, I think we're both on the same page. (And I think that if all you need is a hack, you can do what you need now by just putting the whole double date in the date field, but we both consider that beside the point, I guess.)

    As to the larger problem: the solution does seem to need to work for a variety of cases. At least:

    1. the inclusion of double dates only
    2. the inclusion of embedded citations (in either direction: original-subsequent or subsequent-original)
    3. the inclusion of translated/transliterated titles (which are not titles of a printed translation).
  • International publishing is painful anyway. It has ever been a mess with Reference Manager or Endnote. For years I am struggling with this issue. So this is no fault of Zotero or the programming team - its just a challenge.

    To my experience it would be good if there was a field for the original title. This field should be friendly with all kinds of international characters. Even if someone saves the OTs from Medline or any other international database, s/he can take it for granted that there is a need for correction (e.g. German Umlaute äöü). It is essential to save the OT in the database in case you need it. If you publish " at home" in a non-angloamerican country you will want to use the OT, if "abroad" you will want to add the translation.
    e.g.:

    Freud, A (1936) Das Ich und die Abwehrmechanismen. Fischer, Frankfurt.
    or
    Freud, A (1936) Das Ich und die Abwehrmechanismen [The Ego and the Mechanisms of Defense]. Fischer, Frankfurt.

    Speaking technically, it could be an option of the output styles which version they produce. If the language was a "child" of the OT field, it could be determined if an OT had to be added. For example:

    Publication language English >> all OTs in [].
    Publication language German >> German titles without OTs, all other than English OTs in [] (if available for all Italian, French etc.). English OTs as "Modern Standard Latin" are untouched and left as they are.
  • I too would like to encourage the dev team to make translations of titles in the reference list possible. It is really needed for us who speak more than one language...

    Let me add that I from a non-native English-speaker point do understand that some people react quite emotionally - I find it odd that many journals require translations of titles. What's the point of reading the translation of a Flemish title to English if you're not able to read the Flemish book, anyway? If you don't understand Flemish (or German, or Danish, or Swedish, or Spanish...) then learn it! But that has nothing do to with Zotero - the people behind it are doing a great job.
  • are there any solution for this citation?

    --
    Contamine, Philippe., "La Guerre au Moyen", Paris:Presses Universitaires de France, 1980
    (Jones, Michael(trans.), "War in the Middle Ages", Oxford:Blackwell, 1984)
  • not at the moment, but there will be relatively soon.
    Look in this thread
    http://forums.zotero.org/discussion/1798/multiple-versions-of-a-field-in-a-record/
    for a discussion of what will be done in csl 1.0, which, as I understand, will be part of Zotero 2.1
  • @sonnet, @adamsmith: I'm a little late with a response, sorry.

    The mutlilingual support I've shoehorned into the new CSL processor is a simpler thing than the example above; it's just meant to provide a transliteration or translation for titles that would be totally incomprehensible to many readers if presented in the original language or script. Common use cases might be for Chinese, Japanese or Thai sources referenced in an English-language manuscript, say. To embed cites to works that are themselves translations, we'll need hierarchical or linked item support, which is a more interesting proposition, if still a glint in the eye.
Sign In or Register to comment.