Can you confirm if multilingual journals are supposed to stick with a constant title across all languages? That is, the metadata for
<journal_metadata><full_title>is not supposed to be translated when<journal_article><title>is translated? For example, if a journal publishes some articles in French and others in English, then the article titles obviously will change but the journal title shouldn’t?
Yes, that’s more-or-less correct. For practical purposes, based on the way our title record system is setup (the purpose of which is mainly to prevent duplicate records by duplicate publishers for the same journal) you need to pick one journal title and stick with it. And, the language of the journal title does not need to match the language of the article title.
That said, the schema allows up to 10 <full_title> tags within <journal_metadata>, so if you’re creating your own XML, you can choose to add additional titles, in other languages or as other variations, there. But whichever <full_title> comes first in the XML has to remain consistent from one deposit to the next.
If the understanding above is correct, then the options for a constant journal title, regardless of the content language, would be:
- any of the content languages: in the example above, always the English version (e.g., Canadian Psychology) or always the French version (Psychologie canadienne)
- a compound bilingual version: always both English / French, as in Canadian Psychology/Psychologie canadienne, which seems to be what that particular journal has adopted.
Yes, that’s correct as well. And the choice is really an editorial decision. There’s not a clear best practice we can recommend.
I also noted the CrossRef Title List recognizes small variations such as Canadian Psychology-Psychologie Canadienne (with a hyphen instead of a slash). Can we inform such title variants when depositing a journal-level DOI or is it something better asked via email? And what about diacritics or accents (such as “é” or “ê”): should they be recorded in a simplified journal title variant (with just “e”) or is this character simplification something done internally?
The title list is basically a reflection of all the title records we have at a given time. The variants are the result of both the additional <full_title> elements that are submitted via xml, and any past titles that were submitted before a title was manually changed by us at the request of a member. They’re mainly used for the purposes of citation matching, along with the title abbreviations which are also submitted in the xml.
Diacritics should remain in the titles you submit. We don’t simplify them on our end. The only characters we normalize out in those titles (only in the title records, not in the actual metadata) are the following:
’ { } ; : @ . , as well as line breaks, tabs, and page breaks. So, those characters won’t appear in the title list versions either.
Thanks for letting us know about the csv title list. I’ll alert our tech team.
Best,
Shayn