Cross-Posting: Furigana in Japanese Mini-Stories

Cross-posting here from the general Open Forum in hopes that it gets more visibility:

Someone really needs to clean up the furigana in the Japanese mini-stories. Kanji in Japanese often times has multiple sounds/pronunciations. However, I’ve seen several times where the incorrect furigana is written on top of the kanji, but the audio is at least correct.

E.g., High School Life, a JLPT 5 story.

Off the bat, the furigana says 4 tsuki for April. It should read 4 gatsu. This is really problematic for people studying for the JLPTs since one of the sections is selecting the correct kanji reading.

At least the audio is correct.


I’ve been refreshing Japanese here after a many-year gap.

I’ve been surprised by the lack of quality of the Japanese furigana.


I guess there really isn’t a mechanism for us to edit these stories/suggest corrections, is there? It’s concerning that there are people who pay for the Japanese course but are being taught incorrect readings.


Someone in my early days of Lingq recommended just biting the bullet and turning off the furigana. It took me a few months to get confident enough to do that but I eventually did. It was painful at first without the crutch, but I got past it. Afterward, I realized that I’d been unconsciously focusing my eyes on the furigana and ignoring the kanji when I read, which was preventing me from ever learning the kanji themselves.

I’m no expert on language learning and I’ve still got a very long way to go before I consider myself literate, but maybe give that a try and see what happens. Maybe no furigana is better than bad furigana in the long run.


I agree that turning off the furigana is probably best both in terms of actually practicing the reading but also not to reinforce bad kanji reading…

1 Like

I second turning off the furigana. I rarely pay attention to the generated audio, and instead just listen to the original audio or look at the Lorenzi’s dictionary. It’s common for the generated audio to be wrong (especially with numbers + counters).

If I want to know if I understand the reading correctly, I verify with the original audio and/or Lorenzi. I also put the reading into my definition so that I can verify that I know the reading, but I haven’t been consistent with that.


Yeah, I was figuring that I’d have to use an external dictionary like popjisho to read and verify.

I will say, the support team seems to be taking this seriously and has been in contact with me about the issue. Hopefully, something good will come of it.


T&C, thanks for sharing the info. By “taking it seriously,” are you referring just their understanding of the significance of the problem? Or are you also referring to a willingness and ability to act?

From the outside in, the LingQ technical team seems strapped by a huge code base, architectural inconsistency, a significant amount of tech debt, a high number of bugs, and business need for advancing and releasing cool new features in the age of AI.

Since I’m refreshing Japanese with LingQ, including with the ministories, it’s fairly easy for me to observe most of the word-splitting and furigana errors. Yet, with the commonness of these errors, I don’t think there’s any way I could recommend LingQ for anyone actually learning Japanese. This one problem seems an absolute show-stopper with what’s arguably one of the more important languages in the world.


Nihongo bakari ja nai desu yo. Watashi ha nippon-jin de furansu-go wo manande imasu ga, eigo igai ha, warito ooi nodeha nai deshouka? Ima nanajuu nichi wo koete sutoresu no piku ni tasshita watashi wa, kouyatte minasan ga doko ni sutoresu wo kanjite iru no ka shiritakute haiken shite orimasu.

It’s not just Japanese. I’m Japanese and learning French, but apart from English, isn’t it quite common to learn other languages? Having surpassed 70 days now, reaching the peak of stress, I’m here to see where everyone else is feeling stressed.