Style based on Monumenta Nipponica style sheet?

nathan.hopson

How difficult would it be for me (or some charitable soul) to make a citation style based on the Monumenta Nipponica style sheet?

http://monumenta.cc.sophia.ac.jp/MN_Style.pdf

It's based on author-date, but adapted to the specific needs of citing Japanese (well, CJK) sources.

adamsmith

it's be a fair amount of work and getting the combination of transliterations and original Japanese titles right would require the MLZ version of Zotero. Remind me if you're using that already?

nathan.hopson

I was, but reinstalling caused MLZ to fail. Currently using the standalone.
If I can get MLZ working again, I'll follow up on this.
Thanks.

testest45679

I think this citation style would be a bit awkward to implement -- even with MLZ -- because it requires different name formats depending on the nationality of the author: Firstname Lastname for Western authors and Lastname Firstname for CJK authors.

For titles, MLZ is able to produce any sort of combination of transliteration, translation, or original titles, but I couldn't find a way to get it to handle publications with mixed Western and CJK authors.

The problem is that the language can only be set on a per-item basis, but would have to be set for each author for this to work.

fbennett

MLZ can handle this. Language variants can be set on individual fields and names. You would have to implement the style against a set of test references to see if there are wrinkles in the processor that need to be ironed out, but this is the type of referencing system that MLZ is designed to support.

fbennett

I've taken a look at the examples, and I see the issue: English works by an author who publishes predominantly in Japanese need to be set in romanised form but Japanese order, without the original kanji form of the name; but the kanji form does need to be provided on Japanese works by the same author.

It might work out of the box to tag the primary name (for English items by such authors) as being romanised CJK (i.e. "ja-alalc97"). If that doesn't produce the correct name order currently, it can be made to happen. To cope with English records that carry the kanji name form as supplementary data (for research convenience and accuracy), a small conditional can be added to CSL-m.

If anyone wants to tackle this style, I'll be happy to make the necessary adjustments in the processor, and document them in the CSL-m specification supplement.

As it happens, I will be attending an event at Sophia the weekend after next. If someone can send me an email for the editorial offices of Monumenta Nipponica in a direct message, I could try to get in touch with them directly about possible collaboration on style development.

fbennett

I've added a couple of options to CSL-m that would be relevant to this, and I've written to the Monumenta Nipponica editors to sound out their interest in MLZ support. We'll see how it goes.

fbennett

Have heard back word from the MN editors. They seem not to be aggressively interested, and express a moderate concern about the possibility that style development might require a commitment of time. I won't be picking this one up myself, but as I say above, I'll be happy to make any extensions to the processor that might be necessary to caste this style in CSL-m.

testest45679

Thanks for your great support. Could you pinpoint the modifications to CSL-m? I skimmed over the specification, but couldn't find the new additions.

I'd be willing to implement this style, but I still see one general problem: The alternate language version fields seem to be second-class citizens in CLS-m. It doesn't seem possible to check for their existence in a conditional or to use their content in a text variable. E.g., for the MN style, I'd like to do something like this:

<names variable="author">
<if variable="author#ja-alalc97">
<name variable="author#ja-alalc97" ...>
<name variable="author#ja" ...>
</if>
<else>
<text variable="author">
</else>
</names>

This assumes that Japanese authors are entered into MLZ with their kanji name tagged as "ja", the transliteration as "ja-alalc97", and Western authors would be tagged "en" or something else.

Tagging the primary name as "ja-alalc97" produces the correct Lastname Firstname order, but it would require users who have the kanji version as primary name to edit all their items.

Similarly, for titles I'd like to do something like:

<if language="ja">
<text variable="title#ja-alalc97">
<text variable="title#ja">
</if>
<else>
<text variable="title">
</else>

Conditionals for language tags would also make styles possible that work out-of-the-box. Currently, the format of multilingual titles etc. can be configured in the language settings tab, which is nice for retrofitting existing styles, but with the current way, users have to edit those settings even if they installed the Zotero style for the intended publication. E.g. if they install the MN style, the title won't be formatted properly if they have set it to, say, "<i>Original title</i> [Translation]". They'd have to manually change those settings for the style to work correctly.

fbennett

I skimmed over the specification, but couldn't find the new additions.They are these three -- the global name-order and short-form settings.

Tagging the primary name as "ja-alalc97" produces the correct Lastname Firstname order, but it would require users who have the kanji version as primary name to edit all their items.

The kanji name shouldn't be needed. A language tag can be set on the headline field itself (via left-click and the submenu). A tagged headline field should be picked up either via its proper tag name, or as "orig".

Conditionals for language tags would also make styles possible that work out-of-the-box.

Some means of expressing preferred language settings in the style itself would be good, but with conditionals written directly into the body of the style, it would be awkward to code fallback behaviour. It would be simpler to cast the settings themselves in a style header, similar to what is done for cs:locale. If that's eventually done, the settings should still permit override through the UI (to handle wrinkles in the input where say a user has entries set with "ja-Latn-hepburn" instead of "ja-alalc97", or with transliterated names set [incorrectly] as "en").

The behaviour of the language settings is not completely settled, though, and things will likely change [for the better] in light of bug reports and suggestions as the system comes under greater pressure in the field.

nathan.hopson

Thank you all for continuing to keep the ball rolling.

For my current article, I'm going to just manually edit the source list.

In the short term, I wonder if we could address the question of footnotes in the Monumenta Nipponica style.

Here's an example:
Aston 1907, pp. 164–65.
Florenz 1906, p. 343.
Unno 1994b, pp. 75–81.

Looks to me like "International Organization" (https://www.zotero.org/styles/international-organization) + "p/pp" for page # is the way to go here. Is that an easy edit?

nathan.hopson

10 years later, and I'm still wondering the same thing...

How hard would this be with the current CSL editor, etc.?

Reading the style guide again, it looks like this is the main thing that needs attention:

The reference list, located at the end of the article, includes all works—and only those works—cited in the footnotes. It contains the full relevant bibliographic data, including characters for the titles and authors of works in Japanese.

It would be important to set this only for sources coded as Japanese, I assume.

I've looked at the Chicago note style at citationstyles.org, but I'm not sure how to go about the editing. Looks like you'd need to add fields, which I don't understand how to do.

Any advice, or anyone who wants to take this up?