Problems with title case

Hi there,

Here's my issue: I have a _lot_ of primary texts stored with short title abbreviations like "S" or "RN" or "ToB". I'm using MLA and there are a lot of in-text citations of primary texts, so I need them to be relatively unobtrusive. HOWEVER, Zotero reformats my short titles into title case. So, for example, if I have an abbreviated title: (RNNS 30) it turns into (Rnns 30) in the in-text citation. Since it's an abbreviation of the first letters of the title, this is now ungrammatical.

Can anyone offer me any help? Would be much appreciated :)

Best,
Richard
  • I have the same problem. My citations often have acronyms, and the all-caps get transformed into title case when I put them in a document. I use Chicago (full note) and have the most recent update. Is there a better fix than manual editing of the citation? I tried turning off "capitalizeTitles" in about:config, but it had no effect. -Eric
  • (capitalizeTitle affects import, not export/citations)
  • This sounds like something that shouldn't happen. Would any damage be done if the processor simply didn't touch letters that are already in uppercase?
  • It would fail to fix titles entered in all caps - but my (and I believe Zotero devs) position would be that they shouldn't be in all caps in the first place - so I'd say yes, go ahead and leave capital letters alone.
  • Sometimes when copying titles from docs (I use a lot of "gray literature"), the titles are in all-caps, but the transform feature makes changing those to title case in the collection very easy. The "autocorrect" for citations is more of a problem, since I have to check each citation to make sure there are no inappropriately converted acronyms. So I, for one, would support leaving caps alone if that's the best solution.
  • IMO the auto capitalization should only happen if all letters in the word are lower case. (i.e. word.toLowerCase() === word)

    Consider titles that contain words (from bio contexts) like "mRNA" (acronym) or "alamarBlue" (registered trademark). The initial letters of either of those should never be capitalized. I can't think of cases where words containing upper case letters or non-alphabet characters (i.e. numbers) should undergo automatic capitalization, but maybe someone can present certain scenarios.
  • the transform feature makes changing those to title case in the collection very easy
    I would argue that Zotero should not offer the Transform to title case option. This should only be done by the CSL processor. But I guess that's another discussion.
  • edited June 6, 2012
    IMO the auto capitalization should only happen if all letters in the word are lower case. (i.e. word.toLowerCase() === word)
    That's a good thought. I've taken care of the lowercasing issue already, I'll add this catch in the next release.

    (Edit: Checking the code and then the behavior, I see that the processor already leaves the initial character of mixed-case words untouched in "title" case tranforms, so there's no need for action there.)
  • I would argue that Zotero should not offer the Transform to title case option. This should only be done by the CSL processor. But I guess that's another discussion.
    Continued: http://forums.zotero.org/discussion/23601/remove-transform-text-title-case-option/
  • @aurimas:
    IMO the auto capitalization should only happen if all letters in the word are lower case. (i.e. word.toLowerCase() === word)
    It turns out that text-case="title" already has this restriction coded into it. Should the constraint be extended to "capitalize-first" and "capitalize-all"?
  • edited June 6, 2012
    I'm pretty sure we discussed this already a while back. I could try looking it up [edit: I found a private email exchange between Frank and me of 2011-05-12 where we agreed to the following text], but see e.g. the current description in the CSL specification trunk for the text-case options:

    - "capitalize-first": capitalizes the first character (other
    characters remain in their original case)
    - "capitalize-all": capitalizes the first character of every word (other
    characters remain in their original case)
    https://github.com/citation-style-language/documentation/blob/330939e8940861cb7e23c501c5f824c87ed4222e/specification.txt#L2244
  • The specification text at the link in Rintze's comment reads:
    "Title Case Conversion
    ^^^^^^^^^^^^^^^^^^^^^
    ...
    "Title case conversion (with ``text-case`` set to "title") for English-language
    items is performed by:
    1. For uppercase strings, the first letter of each word remains in uppercase.
    All other letters are lowercased.
    2. For lower or mixed case strings, the first letter of each word is uppercased.
    Words entirely in uppercase are left untouched."...

    So does this mean that the processor *should* be retaining all-caps for acronyms but is not for some reason?
  • So does this mean that the processor *should* be retaining all-caps for acronyms but is not for some reason?
    That seems to be what the specification is saying and I believe Frank fixed it. it just might take a while for the end-user to get the update.
    That's a good thought. I've taken care of the lowercasing issue already, I'll add this catch in the next release.
    It turns out that text-case="title" already has this restriction coded into it. Should the constraint be extended to "capitalize-first" and "capitalize-all"?
    I would say yes. while it is not explicitly covered in the specification that Rintze links to, I think this was the intention when writing it. Maybe Rintze has some reasons to disagree.
  • @aurimas, what exactly are we discussing here? Whether

    ---
    - "capitalize-first": capitalizes the first character (other
    characters remain in their original case)
    - "capitalize-all": capitalizes the first character of every word (other
    characters remain in their original case)
    ---

    should read

    ---
    - "capitalize-first": capitalizes the first character of the first word, if the word is lowercase
    - "capitalize-all": capitalizes the first character of every lowercase word
    ---

    instead?
  • I changed the language of the upcoming version of the CSL specification a bit. Any comments?

    https://github.com/citation-style-language/documentation/commit/c34700363a0a05507e191654253151605f31838b
  • When using Chicago (note), journal titles entered into the database in sentence case are automatically changed into Title Case. This is a serious error, which did not occur in previous versions of Zotero. There are many languages (notably French and Chinese pinyin) for which capitalization is not used in titles. When citing book titles, Zotero does not make this mistake and leaves the lower case if it was entered that way. Has this problem been noted by others and is it being addressed? At the moment, it basically means I have had to stop using Zotero for citing journal articles in my book manuscript.
  • Please ignore previous comment. Just found out the issue is addressed here:

    http://forums.zotero.org/discussion/21974/problem-with-foreign-language-citation/

    Adding the correct information in the "Language" field sorted the problem. Problem did not occur with books I was citing, because book data were imported from Google Books and had the language field completed.

    Apologies for wasting people's time!
  • edited June 25, 2012
    http://forums.zotero.org/discussion/18352/title-case-for-nonenglish-titles-call-for-info/

    If your CSL style has default-locale="en" or no default-locale, entries are presumed to be English. In the Language field of each non-English item, enter the two-character code of its language (i.e. "fr" for French, "zh" for Chinese).

    If your CSL style has default-locale="fr" (or any other non-"en" value), entries are presumed to be non-English. In that case, set the Language field to "en" for the English entries.

    You may wish to set the correct Language value for all of your entries, so that citations will come out correctly regardless of the style applied.

    (Edit: No problem; that's why we're a community.)
  • edited June 28, 2012
    Rintze said:
    I changed the language of the upcoming version of the CSL specification a bit. Any comments?

    https://github.com/citation-style-language/documentation/commit/c34700363a0a05507e191654253151605f31838b
    A little more insight on this. A lot of titles fetched from websites and otherwise are in title case, which is often their "original" case and I think should be preserved in Zotero. However, that kind of breaks sentence case conversion. I think the following should happen:

    sentence and capitalize-first: all words beginning with upper case letters (\b[A-Z][a-z]*\b) are lowercased. The first letter of the first word of the sentence is capitalized.

    This of course introduces the problem that certain words, like proper nouns, should always have first letter upper-cased. One solution to this would be introduction of <pre> tags, which would prevent any case conversion.

    Edit: maybe the condition above can be applied only if the title is entirely in title case. That is, we check that all the words (with the exception of a few skip words like a, the, with, etc.) have first letter in upper case. I understand that it becomes more complex with other languages.
  • A lot of titles fetched from websites and otherwise are in title case, which is often their "original" case and I think should be preserved in Zotero.
    Why? To be clear: Only English has title case, so this only applies to English titles in the first place.
    So far, the only citation style we're aware of requiring English titles "as is" is one French citation style.*

    On the other hand, problems with proper nouns in sentence case would be a common concern, with many major citation styles requiring sentence case titles.

    Yes, we could improve sentence casing by only forcing it on title cased titles, but I'm not sure that's going to create more good then harm.


    * And I'll stick to my claim that this is only the case because the editors of that and similar styles don't understand title casing in English. E.g. you'll find many books that have title case titles in the front, but are sentence cased in the LoC, the most authoritative source of bibliographic data. The idea of a "real" casing for titles simply doesn't exist in US publishing (can't say anything about the UK).
  • edited June 29, 2012
    I agree with adamsmith. Title case holds less information than sentence case, so it's just a bad choice for storing item metadata.
  • I hope there will soon be an option to specify the case for saving new items. I prefer the sentence case myself.

    Is there any means to (batch) convert the case for the whole library at once? I am not looking forward to going through several hundred entries and changing the case one by one :-(
  • Zotero saves in sentence case whenever the data is presented in sentence case.
    It just can't auto-convert to sentence case (also, again, auto-conversion is imperfect bc. of proper nouns etc. so you'd have to manually check anyway). There's currently no option to batch edit, no, but it's planned.
Sign In or Register to comment.