Taylor and Francis+NEJM translator: What must T&F do to allow Language (Also NEJM)
I seem to have the ear of someone with Taylor and Francis indexing department. (I correspond with T&F twice a year concerning my license to abstract and index the contents of their journals. T&F wants a summary of the number of articles from each journal that SafetyLit indexes. I am working with a very cooperative person who is new to me.)
I understand that the reason that Zotero does not capture a language tag is because the few non-English T&F journals have metadata that incorrectly labels the language as "en". (see:
https://forums.zotero.org/discussion/79441/taylor-and-francis-translator-and-language-field#latest )
Please @zuphilip or someone else, give me the specific and detailed instructions that I may relay to T&F so that they can fix their metadata so that it meets Zotero translator standards. It is tedious to need to hand-edit all T&F-metadata records to include the "en".
Also: I see that the translator for the New England Journal of Medicine is the same as that for T&F journals. Why reject the New England Journal of Medicine language tag in the page header: meta name="dc.Language" content="en" ? Please consider making the NEJM translator discrete from the T&F translator. NEJM does not publish non-English articles.
This is the text of my request to T&F Abstracting and Indexing Administrator sent last year:
Thanks
I understand that the reason that Zotero does not capture a language tag is because the few non-English T&F journals have metadata that incorrectly labels the language as "en". (see:
https://forums.zotero.org/discussion/79441/taylor-and-francis-translator-and-language-field#latest )
Please @zuphilip or someone else, give me the specific and detailed instructions that I may relay to T&F so that they can fix their metadata so that it meets Zotero translator standards. It is tedious to need to hand-edit all T&F-metadata records to include the "en".
Also: I see that the translator for the New England Journal of Medicine is the same as that for T&F journals. Why reject the New England Journal of Medicine language tag in the page header: meta name="dc.Language" content="en" ? Please consider making the NEJM translator discrete from the T&F translator. NEJM does not publish non-English articles.
This is the text of my request to T&F Abstracting and Indexing Administrator sent last year:
I hope that someone will be able to suggest different language for my request such that it is likely to be heeded.
Although some of the Dublin Core elements in the page headers for T&F non-English language journals indicate the true language of the article the most important DC element falsely labels the article language as English. [meta name="dc.Language" content="en"] While there are relatively few non-English journals, I'm sure that you want these to be indexed properly with their accurate language label.
For example see:
https://www.tandfonline.com/doi/full/10.1080/14682737.2017.1377467
https://www.tandfonline.com/doi/full/10.1080/00806765.2019.1586582
This not only affects what I bring into the SafetyLit database but the incorrect language labels are imported into each user's personal bibliography management software (EndNote, Mendeley, Zotero). These bibliography managers use the language label to assist with properly formatting the article title so that it meets the requirements of the various style manual requirements. Improperly styled references can cause serious problems with manuscripts submitted for publication or student reports where some of the grade will depend upon a properly formatted reference list.
For those of us at SafetyLit, we edit your metadata to provide the correct language. This is tedious and slows our indexing of your published articles.
Thank you for passing this along to the proper department so that SafetyLit can properly index your articles in a timely way.
Thanks
We can use any information on the page, but it'd be cleaner if we could use the formats we already use.
FWIW, two things in your quoted email aren't quite right:
a) There isn't really anything in the metatags that indicate the article/title is in Spanish, unless you count the fact that the abstract is available in both Spanish and English (which does not reliably indicate a Spanish title/article).
b) It's actually not a big issue for English language tags to be missing for Zotero and Mendeley purposes (I don't know about Endnote but would be surprised). Title casing happens when the language field is empty or starts with en, so those two are equivalent. The missing Spanish language tags are the problem.
Also, if you end up doing this a lot, check out Zutilo's ability to batch-write fields, which should make this pretty painless.
edit: I have numerous records from a Japanese journal each has near perfect metadata except that the language tag is missing. I hope to use Zutillo to paste "ja" into the language field of all the selected records so that I do not need to paste each record individually. Similarly, when I capture records from Taylor and Francis journals I would like to add the missing "en" to the language fields of each record. THanks
https://pressbooks.library.yorku.ca/masteringzotero/chapter/an-easier-way-to-manage-tags-and-related-items/