For those who aren’t familiar, Open Library (https://openlibrary.org/) is the official book-lending arm of archive.org. It is a digital library with thousands of titles in dozens of languages available freely to the public, but as it is a non-profit organization, book listings are of varying quality and are improved mostly by volunteer editors.
Open Library can potentially be a fantastic option for accessing books you find on Natively for free, however, discoverability for books written in languages with a non-Latin script is extremely poor. Most book pages are automatically seeded with English library catalogue entries, meaning that the title, author, and any other pertinent information is only presented using inconsistent romanization and is completely unsearchable using the book’s native language.
I’ve confirmed with site staff that this is undesirable and that they encourage re-writing book information in its native script if users can accurately do so. And because Natively is a community of people who like books and might be familiar with the material, I thought I’d put the call out (particularly to Japanese and Korean learners who are most affected by this limitation) to look through the catalogue and improve the entries of any titles you’re familiar with.
Every entry that has its information corrected will make Open Library a more useful resource to other readers!
As in replace it completely? Or will both be still available?
(It doesn’t really matter in the end, but if someone got used to search for a given book/series in romaji, they would be confused if it suddenly switched to an other script)
Edit: I see a lot of examples of [Japanese title]: [English title], but there’s no standard I guess.
Replace titles completely with the correct ones. In general, the romaji information isn’t useful to searchers, as native readers are the primary demographic who would be accessing the titles; the number of users looking for the romaji title would be very small in comparison.
There’s a little FAQ on the main site also clarifying that author names should be written in their native language if possible, with the romanization given as an alternate name. “If the native spelling of an author’s name uses non-latin characters, please use that as the primary spelling. Transliterations and alternate spellings may be added to the field labeled “Does this author go by any other names?”. Anything entered into this field will be searchable.” [x]
I was afraid I would break something, but looking at the latest addition, I can see someone added a Chinese book titled 大周皇族 2 as content for a completely unrelated Japanese book Why would you do that.
I was pretty focussed on the CJK romanization issue when I wrote the OP, but it’s worth mentioning that there are many other ways book entries can be improved that apply to ones in Natively’s other languages, too. Every little bit helps!
The most useful one to learners browsing would be fixing missing/incorrect language tags, so books actually appear in the catalogue for the target language. And if a book is a translation of one in another language, filling in that information (including the name of the translator in the Contributors section) and linking the “edition” to the correct “work” in the original language particularly will help surface books that people browsing might not otherwise find.
What I’ve mostly been doing is searching for books I own by their ISBN, rather than searching the title/author or browsing the language catalogue, as it’s more reliable in bringing up the entry I want to fix if it exists.
I figured that it was added by a bot, from the internet archive (where it was also recorded that way). I flagged it on the internet archive side. We’ll see what happens.
I also edited one book ( 騎士団長殺し 1), but it feels nerve wrecking as everything I saw was in romaji In particular, someone took the time to transcript the whole description in romaji, while I just copy/pasted the one directly available from Amazon (after checking it was the exact same content). Since copy/paste is a lot less efforts, I can’t help but think that the other person did that on purpose
Dang. That’s so much work for so little reward. You’d have to be able to read Japanese to read it anyway, and those who know Japanese wouldn’t want to read it in romaji… I have to wonder what the point was.
From what I understand, the absolutely-everything-in-romaji is a limitation of the library catalogues that Open Library imports their data from. I know when I look up Japanese books at my local library (in Canada), the entries look similar, and I think it’s just that whenever the systems were designed, they didn’t have non-Latin script books in mind. I don’t think it’s someone taking the time to intentionally transcribe everything into romaji, it’s probably an automatic process of converting whatever Japanese data they (whatever library system Open Library was importing from) started with into something Western library catalogues could use.
For reassurance, this is what was told to me when I contacted them for clarification on whether the romanized titles should be changed or not:
That is the way the records come to us. Do feel free to convert such titles to their proper titles. In doing so, please be sure to only change the edition’s information and not the work’s information, unless that is the work’s original language.
Little quirk I’ve discovered while finding book entries to fix up: a lot of non-English books get imported with the language written in the title, instead of having the proper language tag to appear in the catalogue. But a lot of the entries that are like this are actually pretty famous works that probably a lot of learners would be interested in borrowing, if they could find them using the language filter:
(Shout-out to the collections of old Clamp and Hiromu Arakawa works I found that were literally just labelled “(in Japanese)” with no titles or other information given.)
Even without correcting anything else in the entry, just searching for something like “in (language)” or “(language) edition” and adding the right language tag to any results that look interesting or popular will do a lot to surface books for learners browsing their target language’s catalogue.
I’ve been going a bit overboard on OL’s Japanese books since it was added as a provider for Natively Finding books to fix and cross-list here is kind of fun, though, like a scavenger hunt.
Yeah TBH a lot of those 600 aren’t super useful to have linked, like only the 下 volume of a 2-book set or volume 19 of a 30-volume manga, but I get satisfaction from adding them and seeing the numbers go up anyway. If I’m just doing it as a time-killer it’s not too onerous, just a lot of copy-pasting.
This is really good advice, thanks! My MO for a while was just searching random romanized words and seeing what came up
Something else I’ve noticed for Japanese books is that sometimes the scans are uploaded upside-down or back-to-front (i.e. with the binding on the left instead of the right). I’ve been reporting them to site staff when I find them and they’ve said they’d be fixed “shortly”, but everything else I’ve reported that needed elevated permission to fix (e.g. merging duplicate authors) has just gone into a backlog, so I’m not sure how quickly those kinds of problems can be addressed.
Do they have the notion of a superuser on archive? I know back when I used to do stuff like this for other sites occasionally you could get elevated permissions if you applied and showed them that you could properly use the site. I figure at this point you’ve done enough to demonstrate your knowledge/good intentions.
I think those are called “Librarians” on Open Library, but honestly I don’t have a very good idea of what they do, what responsibilities they have, or how to apply to be one, so I’m more comfortable just being a standard user and fixing up titles in my downtime.
So far they’re great at communication, though! Any time I contact with a question about improving book entries or to report a problem, I get a human answer within a few days max. Gives me confidence that issues are being seen by someone and will be addressed eventually, even if they have a backlog to work through first.
2 Likes
As an Amazon Associate, Natively earns from qualifying purchases through any Amazon links on the site.
Why would you do that.
In particular, someone took the time to transcript the whole description in romaji, while I just copy/pasted the one directly available from Amazon (after checking it was the exact same content). Since copy/paste is a lot less efforts, I can’t help but think that the other person did that on purpose 
