I think the title (and the article to a degree) is misleading. Something like "Japanese bureaucrats finally plan to update severely out of date official romanization scheme" would be far more accurate.
The examples in the top table make it clear, I've never once seen judo romanized as zyûdô, it's always jūdō. Out of curiosity I asked a native Japanese speaker and although they could figure out what word it was they had never seen the former romanization either.
The title also makes it sound like a new romanization system was just created, when in fact it's just modern Hepburn, which is how everyone other than the Japanese government has been romanizing Japanese words for ages. The article does mention Hepburn to be fair, but then it makes it sound like the proposed changes are something new and not the de facto standard.
> The article does mention Hepburn to be fair, but then it makes it sound like the proposed changes are something new and not the de facto standard.
Well, it's not the de facto standard, since there are differences, for instance:
> The revision plan allows for not using any diacritic for a long vowel, as in “judo,” or adding “h,” as in “Ohtani” for the Los Angeles Dodgers star. Both are widely used internationally.
When typing in Japanese I tend to type "tu" instead of "tsu" or "si" instead of "shi" because the end result is the same (つ/し) and I save a keystroke, I assume most people who regularly input Japanese with a QWERTY keyboard do the same. The revised romanizaton from the article prefers "tsu" and "shi" though.
That's true—the actual IME is much more flexible than this.
My point is more that this isn't really anything new. It's a fairly common way to Romanize Japanese in practice, even though it doesn't strictly follow Hepburn or Nihon-shiki. In fact, this style seems to be more widely used in the Western "fandom" world—on platforms like AniDB, Danbooru, and others—than the two "official" systems.
Tokyo Metro romanizes 日本橋 as Nihombashi, you can look up the station on Google Maps and check out the photos, the signage showing "Nihombashi" shows up pretty soon.
Google Maps itself also romanizes it as Nihombashi. Same applies to 神保町 (Jimbōchō), for instance.
You're perfectly right and it's what occurred to me when I lived in Japan, too: you do get to see 'abberant' spellings quite often and that's because train companies decided many, many decades ago on what system to employ for their English station names, and those stuck of course because of printed schedules and so on. Depending on how you count Nihombashi either gets lots of points b/c of the many people who get to see this particular name each day, or it just gets a single point because it's one spot on the map. Observe that when you click through the pictures to get to the actual map—you'll find "Nihonbashi Bridge" spelled in the more common way (including the adorable repetition). So arguably 'Nihonbashi' it is in general but the station name is spelled with an 'm'.
Yeah I think you're right, unless you often visit those places it's more common to just see their names only when you pass by their train stations or see them on the train maps, so the "mb" spellings stand out more.
Try not to get peeve-hooked by it but take it as an expression of the democratic, open societies that the folks in S Korea, Taiwan and Japan try to be and the jumble that is the many divergent transcription systems becomes a way to express their freedom...
Try Thailand. People will just make up whatever romanisation they want to. "Fookoosheema" would definitely happen there. They also sometimes randomly throw in an r, not exactly sure why (maybe hypercorrection, because casual speech tends to drop rs).
The 'r' is to the best of my understanding not thrown in randomly but a reflection of British spelling. British being a non-rhotic dialect of English doesn't sound post-vocalic, pre-consonantal 'r's. Hence in spellings like 'Vajiralongkorn' the 'r' just serves to indicate a long o [ɔː] which contrasts with the short o [o] of 'long'.
That's part of it but doesn't explain why Phat khaprao is sometimes spelled as phat krapow or similar. Although maybe that's just non-Thais mixing it up.
> "phat khaprao" or "phat krapow", which one is correct Thai?
> In Thai, the correct spelling is "phat kaphrao" (ผัดกะเพรา). This dish, known for its delicious combination of stir-fried meat (often chicken, pork, or seafood) with holy basil, garlic, and chili, is a staple in Thai cuisine.
Might of course be an often-heard variant name of a popular dish.
And for an appearance of comparability the "#8" programs are un-optimised single-thread transliterated lowest-common-denominator style into different programming languages from the same original.
What a great resource, thanks to whoever put this together.
To anyone who enjoys contorting the language to make it do weird things for fun, quines, and those sort of things, I really recommend Tomoya Ishida's talk at RubyKaigi 2024. It's in Japanese, but there are subtitles, and the slides speak for themselves. Some of the more whimsical uses of Ruby I've seen in a while. Keyword to peak your interest: animated quines.
I suspect that not only relieved some of their own stress, but also sent a very clear message to the community that someone needed to take up that slack if they wanted OSX builds.
For me it's gotten a lot better since one or two versions ago. There's still some lagginess, but it's nowhere near the horror show that it used to be, and they fixed some excruciating screen resolution issues as well. I used to avoid Inkscape on macOS like the plague, now I can use it fairly comfortably.
I haven't used Graphite, but from what I've seen it allows for non-destructive procedural editing using nodes (à la Blender), which Inkscape mostly doesn't do.
I knew that that the <html> tag can be fully omitted (and some others, like <body> and <title> IIRC), but if I read that it can also be partially omitted. So you're allowed to open it but not close it, or close it but don't open it. That's new to me.
I've tried omitting those tags, but I decided that in the end things are easier to read when you do include them, so nowadays I always include them when I write HTML.
> This language server is stricter than the HTML spec whenever it would prevent potential human errors from being reported.
If your goal is to output minimal HTML, let a tool automate that. Forgetting to open <html> is more likely a mistake than shorthand, and as shorthand it's not valuable.
The examples in the top table make it clear, I've never once seen judo romanized as zyûdô, it's always jūdō. Out of curiosity I asked a native Japanese speaker and although they could figure out what word it was they had never seen the former romanization either.
The title also makes it sound like a new romanization system was just created, when in fact it's just modern Hepburn, which is how everyone other than the Japanese government has been romanizing Japanese words for ages. The article does mention Hepburn to be fair, but then it makes it sound like the proposed changes are something new and not the de facto standard.
reply