[Userscript] WaniKani Pitch Info

I am a noob in pitch, but… Does it help to specifically say (and make me remember) that the word is 尾高, that is, number of morae == pitch number?

Nailed it. :thumbsup:

1 Like

Yeah, sure that works for most, except 中高 of course

No. But I’d say those words are rare to begin with.

My copy of the 「NHK日本語発音アクセント辞典」 places them side by side:

Yeah, the audio on those is not correct. They are the other way around.

As for the “middle” tone, there is some academic discussion on that, but basically, no: currently people don’t normally analyse Japanese as a three-tone system.

I’d suggest displaying pitch accent similarly to how japanese.stackexchange.com does it! Example:

On a related note, I’ve heard what pitch a word starts on somehow depends on the word before it. How does that work out? Could / should that be indicated in some way, perhaps?


An unaccented word has no pitch fall. If the word that followed it started with a low mora, then it would mean that the unaccented word had a pitch fall at the end (because you went from high to low). Since this cannot be, then moras that are preceded by an unaccented word are high.

But a more general rule would be hard to come by. Compound words do not necessarily inherit the accent pattern of the words that compose them. So if you have eg. noun + noun, then the above might not apply (because the first word might not be unaccented when used in a compound).

What I mean is some info is not shown by listening to WaniKani audio… It is only shown if the word is attached to an article, that is, in a sentence.

Diagrams like jjatria/obskyr suggested would help.

@Invertex Looks like the pitch accent information doesn’t show up during the quiz part of the lessons.

To reiterate what carinama and Leebo were saying, the text you put for the pitch accent numbers doesn’t seem accurate. For example, with [2]:

Pitch starts low, stays high for 2 mora, ends low.

It doesn’t stay high for two mora. It stays high up to and including the second mora, which means it stays high for only one mora since the first is low. Also, if the entire word is only two mora, the word itself doesn’t really end low (even if the following particle is low).

This is why I personally prefer just the numbers (perhaps with a general explanation) or the images shown above. Text like this can be tricky to get right programmatically.

Eugh, damnit WaniKani, why isn’t your code same patterned across the site… The “Reading” section alone already uses 3 different kinds of html elements depending on the page type lol, it’s so stupid just for some text.

I’ll look into all this stuff soon, bit swamped with work work right now.

I’d like to bump this comment - this confused the heck out of me before I learned about 中高 and 尾高. The text actually might be doing more harm than good right now. Let’s say I was a beginner and looking at 足:

あし - Pitch starts low, stays high for 2 mora, ends low. [2]

Since the word itself is only 2 mora, あ is low then し is high, but then… it ends low? Do I change my pitch in the middle of pronouncing し? Okay I’ve listened to the sample audio WK gave me, but it seems to end high, what is this low business about? (It’s obvious to me now it means 足 is 尾高 and the particle attaches low but there is no way I could have figured that out from just the description).

Maybe ditch the text and link the number to a description of what the number means? A diagram would be ideal, but I know that’s a lot more work to implement.


Yeah, that text description is just wrong

1 Like

@Invertex Any plans to update the text?

Yes, when I have the time, I work a lot of crunch in game development and often don’t have much time for side projects. I’ll try to at least get the text adjusted for <3 mora words soon, or if someone wants to layout some more specific rules in one post for me.

@Invertex I know you’re busy at work, so I’d love to implement the visual lines myself, as in Dogen’s videos (like at this time position: Japanese Phonetics #3: Common Japanese Misconceptions - YouTube).

If you don’t mind me jumping in on your script, that is. Let me know!

1 Like

OOoooo. Where do you work? What games have you worked on?

Me again. So I couldn’t wait and wanted to see what I could do with @Invertex’s script. I implemented the chart you can find in the Apple dictionary, but I only sort of know how to use it. From the discussion above, it looks like 三つ is incorrect? My knowledge is extremely limited so I’d appreciate some feedback and guidance. Also, the white dot is the particle.


三つ looks correct. They show the extra particle dot to differentiate between (in this case) [0] and [3]. Since it’s [3], they have to indicate the the particle following the three mora in 三つ drops instead of staying high (as [0] would indicate).

Also, that looks nice! But what are you doing to handle when a word has more than one valid pitch accent?

I guess showing them all above the word in rows would be the correct thing to do.

1 Like