Any advice on distinguishing some Japanese consonants?

In particular, I don’t feel I can differentiate し, じ and ち at all.

To a lesser extent, perhaps す, ず and つ; but usually I can tell z.


Probably something similar in english with c and s. It’s just something that will come with time I’d imagine.

Hm, I can’t say I’ve had this problem, what are you doing for listening practice? Potentially speaking using them will help too, but not sure.

Anki with audio on the front.

I can’t remember exactly, but I think sh ch is also different from English.

Can you share an example where you can’t tell the difference?

Yeah, they’re a bit different than the english equivalents. Are you able to say them or is that also a bit off? When I say them し seems to be a bit more clear and ち sounds like there is a bit of white noise. Not sure if that’s a good explanation though.

Long before I started Japanese, I had been watching anime for a long time so I guess that counted as listening practice for me. While I don’t use Anki, I generally don’t think the robot voices or even if there is a voice actor is going be sufficient for listening. I’d recommend something else like a podcast or a JP youtuber.

1 Like

For saying them, shi and tchi are fairly useful, imo.


The main problem is that if people don’t speak the same native language, it’s going to be difficult to give advice on hearing differences, because it’s your native language that sets the baseline ability to hear various sounds.


Yeah I think that’s fairly accurate for ち. Unfortunately though for し it’s not so obvious.

I’d say for し maybe start off with the english sh, then spread your mouth open a bit more kind of like a smile, it changes the sound. Once you know that you could probably replicate it without the english part.

I share a native language with everyone, the language of love. :smiling_face_with_three_hearts:


Or do shhhh :shushing_face: if you mean that it’s sometimes more like ひ, then yeah, but that usually depends on the person, I think.


Perhaps the problem isn’t as severe as I thought, but some imagery may help to imagine better. Otherwise, IPA or vocal organ chart.

I might reply to the thread later if I notice the same problem again.

A little imagining helps a lot.

I think that for English too. But it probably only works a little more than 50% at a time, both for pronouncing and recognition.

It’s because I imagine the IPA, or some variants; where I originally studied English, outside school, from BBC’s course series.

I used a premade deck, because collecting sentence audio (from ImmersionKit or perhaps, Youglish / Youtube) is some work to do. Vocabulary audio attached, to tell which vocabulary to focus.

I unsuspend by each Wanikani level, at least one Kanji with that level, now at level 7 (skipped level 1, and added no-Kanji vocabularies). For Kanji as well, to do handwriting.

I haven’t looked into sub2srs yet, but I already have too many items from Yomichan.

My review is, shorter audio helps to focus on pronunciation subtleties.


I find it’s easier to tell them apart in context than when hearing vocabulary words in isolation.

Of course, but that is true for seeing vocabularies in isolation as well, like Wanikani-styled SRS.

I purposedly add sentence audio on the front of this purpose. (So, 2 audio files.) Although changing sentence audio is an option, I have no need to do that yet.

If I find something amiss, I might also suspend those cards.

Are there any good prebuilt decks that use minimal pairs to help train your ears to distinguish similar sounds, perhaps? I bought a Fluent Forever anki deck for French a few years back that I figured was worth the $12 I paid for it; they seem to do a Japanese one though I have no experience with it. The French one would present you with audio and the two words, and ask which of the two possibilities it was. On the answer side of the card you could play the audio of both words to compare them.


In really fast spoken Japanese a lot of syllables are kind of smooshed together, and your brain isn’t understanding it by breaking it down at the syllable level, you eventually learn to use context to figure out which words it thinks are being used. It does take time to get there, though. It might help to listen to Japanese speaking with Japanese subs.


I think examples where you get confused would definitely help, even if it’s just for everyone else to try to suggest words you can practise with. It’s certainly true that it depends on your native language though: I know I have a really hard time with the three H-like sounds in Arabic, for example, even if I kinda know the difference now that a friend has explained them to me.

This is true though. Sometimes syllables really aren’t pronounced clearly, especially when people aren’t being recorded for the sake of broadcasting. There are moments when you hear the right sound because you know the word, as opposed knowing what word it is because you can identify the sound.


This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.