11 ms·

(Everything that follows is my personal opinion.) I'm technically working on a competitor to Duolingo (for Japanese), so I'm definitely biased, but personally

(Everything that follows is my personal opinion.) I'm technically working on a competitor to Duolingo (for Japanese), so I'm definitely biased, but personally I'm a little sad with how much success they're having considering how bad their system is (for Japanese, can't comment on other languages, but I can't imagine it's a lot better?). There are a lot of considerably better options for learning Japanese (and I'm not even talking about my own app as it's still a huge work-in-progress), and yet for some reason I see posts from people using Duolingo on forums everyday. It's really popular. I make a point of trying out most of the apps/websites in this space (since it's always a good idea to see what others are doing better than you), and (to me) it's apparent that Duolingo is definitely on the bottom of the barrel when it comes to actually being an effective learning tool. It might be okay-ish for a total, 100% beginner, but it breaks down very very fast. I guess you don't need to have a good product as long as you have solid marketing, flashy UI and heavily gamify the process? That was somewhat eye opening for me.

ummonk 5y ago
Gamifying is an important feature though, as it makes it easier for students to maintain motivation to keep learning.
- kouteiheika 5y ago
  Yeah, I agree. Even if a tool is effective if you can't motivate the user it's all for naught. I just wish that concentrating on this aspect wasn't so disproportionately rewarded by the market; after all an educational tool should primarily be judged by how effective it is in actually educating, and the "fun" factor should only be a bonus, where with the current state of affairs it kinda seems like it's the other way around.
  - Vinnl 5y ago
    But the point there kinda is that gamification is not a "fun" factor, it's what makes them more effective in actually educating. If you have the best learning strategy, but you don't actually stick to it, it won't help.
    - kouteiheika 5y ago
      Indeed, it makes them more effective in actually educating, but the effect is multiplicative, and not additive. The tool that is gamified needs to be effective in the first place, so ideally you'd need both.
- JCharante 5y ago
  Except I play some video games that really motivate me to keep playing. Both video games and Duolingo accomplished this by eliminating the learning portion from their software to boost retention.
stu2b50 5y ago
What are some recommendations for learning apps/websites?
- davidzweig 5y ago
  Which language? :) Edit: I enjoy Audiolingual drills: https://fsi-languages.yojik.eu/ https://fsi-languages.yojik.eu/ These courses vary, some very good. Chinese and French in particular. For Russian, 'Modern Russian' (Clayton). 'French in Action' video series is excellent. Assimil books are helpful. Software.. eh. [I make language software :D]
- emsy 5y ago
  I’ve tried Duolingo, Babbel, Rosetta Stone, Lingq and some others I can’t even remember the name of. First of all, you won’t learn a language using only an app. And you will need a lot of intrinsic motivation and grit which an app can’t help you with. That being said, the best app I ever used and still use is Lingq, because it guides you towards using the language in real life situations with a lot of context, and you can use their audio clips to learn in the car or during Sport etc.. The other apps only ever have you translate words and context free sentences, which is a futile approach.
- wraptile 5y ago
  Generally app/website courses suck because of conflicting business models. You need to seriously read and understand grammar to learn language and employ spaced repetition memorization. The apps usually do the later not the first one because holding attention span is hard, writing good courses is harder and it's easier to monetize dictionary building than reading material. Common plan is to pick up a well written book for your language and do flash cards + media/live tutor suppliment. Currently I'm learning Thai: I have a modern book course (Becker's series) + voice Anki flash cards + some media on youtube or netflix movies (I've tried pimsleurs but it made me fall asleep)
- yepguy 5y ago
  Check out Refold[1]. IMO it is the best place to learn how to acquire a second language. It's not an app or a course or anything. Its approach is almost all self-directed, but it lays out everything you need to do and provides a Discord community to help you. [1]: https://refold.la/ https://refold.la/
emsy 5y ago
As someone who has learned and is learning multiple languages this rings true to me. Duolingo is very good for getting started and acquiring the basic vocabulary. but it plateaus rapidly. After that you don’t get much from it and you should use something that uses mini stories with a larger context and preferably audio read by natives (Small aside: Duolingo has awesome mini story podcasts for Spanish and French which they should incorporate into their app). The best tool I found for the mini story approach is Lingq, which is not so good at getting you started though. You’re also right about the gamification. This is something Duolingo absolutely nails. A friend’s mother is addicted to doing her English lessons and has a 2 year streak. But she also barely speaks English yet, which is unsurprising to me given the fact she only ever uses Duolingo.
- mike_h 5y ago
  Hadn’t heard of LingQ til now, this seems really good.
  - emsy 5y ago
    The founder (Steve Kauffmann) is a polyglot with a youtube channel where he talks about effective strategies for learning languages, sometimes even in other languages. I can’t recommend it enough.
  - personlurking 5y ago
    It does seem good, til you compare price ($8-40/mo) with its limited use cases.
    - emsy 5y ago
      What do you mean by limited use case?
      - personlurking 5y ago
        When I used the free trial, the main activity was centered around pasting text from other sources/sites, then marking the words you recognize. To me that's not worth $8/mo, much less $40 at the Premium Plus level. I realize there are a few other features to LingQ but the cost-benefit of the lowest tier didn't make sense to me. I have free dictionary apps that remember every word that is looked up. I'd rather input those lists into a spaced repetition app and receive essentially the same benefit without paying.
bawolff 5y ago
I mean, at some point, success is the definition of success. Perhaps you're misinterpreting what the average consumer wants in a language learning app.
- kouteiheika 5y ago
  Initially that might have been true, but now that I thought about it for a little while I don't think I am. I just personally find it sad that for your average consumer the educational aspects in an educational app are significantly less important than non-educational ones, and that the perceived value of a product vastly overshadows its actual value, that's all.
AussieWog93 5y ago
What app, if I may ask?
- kouteiheika 5y ago
  This one: https://jpdb.io https://jpdb.io But be warned - it is still very rough around the edges, and still very much a work-in-progress. (I'm only working on it in my spare time as a side-project.) It's definitely not ready for prime time yet, and won't be for quite some time still. That said, I do already have users. One of my users is using it to help him consume native Japanese content really early. He's learning Japanese since September of last year, and he started using my app a few months ago (probably a month or two ago; not sure, I'd have to check) and he's already reading some of the easier Japanese visual novels and web novels, and watching anime with only Japanese subtitles. This - I personally believe - is the best way to learn a language: by immersion. So I always wanted to make an app that will short-circuit the process and will allow you to start consuming content a lot earlier that you'd ever be able to do using only traditional methods.
  - devilsbabe 5y ago
    Very cool work! Paradoxically, I've been slacking off on practicing Japanese diligently ever since I moved to Tokyo. I've been meaning to get back into a good routine so I'll give jpdb a shot. I'm curious how you did this: "We have 16785 prebuilt decks with vocabulary from 1124 different anime waiting for you." Did you write a script that calls subs2srs?
    - kouteiheika 5y ago
      Well, essentially yes, it is based on text analysis, but it's a lot more complex than "a script that calls subs2srs". (: (My whole codebase is over 100k lines of code, all written by me.) I have a unified morphological analysis engine that I use for every type of media that I have in my database, and I use it to generate stats and vocabulary lists.
      - 5y ago
ehnto 5y ago
I am currently learning Japanese on Duolingo, I am wondering how long it has been since you tried their Japanese course? I noticed it changed significantly between now and when it first released. It certainly doesn't feel like a course like traditional education provides, more like a flashcard vocabulary and grammar builder. I have had to Google plenty of things while using it because nothing is ever actually explained and Japanese has plenty of arbitrary rules, but I don't feel that's all that bad. I am also curious if you need beta testers for your app, I would be open to giving it a go.
- kouteiheika 5y ago
  Yeah, I did look at it recently too, and I still think it's not very good. I mean, for a total beginner it is okay-ish for some basic vocabulary, but at that stage of learning a textbook approach is a lot more effective in my opinion. You're better off just grabbing a copy of Genki, going through that with a teacher, and once you have the fundamentals down and can efficiently study on your own start doing vocabulary flashcards in a better app and consume easy native content. Sure, I'm always looking for beta testers. (: Although my app is very much public and open to everyone. Just don't expect miracles since it's currently only a work-in-progress side project that I'm working in my spare time and slowly improving: https://jpdb.io https://jpdb.io
nosefrog 5y ago
Hey! I went through your comment history, and you seem pretty similar to one of my friends haha. If you're down for a fun experiment, email me at my username @gmail.com.
- kouteiheika 5y ago
  That depends on what exactly the experiment entails. But since you're the one suggesting feel free to email me with the details. (: My address is @exia.io; you can use any username since I have a catch-all for the whole domain.
shusaku 5y ago
Don’t you think the whole market of language learning apps should be looked at with suspicion? I talked to my brother-in-law the other day, and he said he hired a tutor in the Philippines to teach his kids English. Unsurprisingly, it was super cheap. Low paid labor might be hard to get for Japanese, but it seems like for most languages people want to learn lessons over Skype will be far greater value than even a well designed app.
- kouteiheika 5y ago
  I think it depends on what exact level you're currently at in your language learning. If you're a total beginner then I personally think that having a teacher who can actually explain things is invaluable. If you don't have the basic rules of the language internalized and you don't have the fundamentals down you're going to have a harder time doing it on your own than someone who can just ask a teacher and immediately get an answer. I think the sweet-spot for language learning apps (at least for Japanese) is learning vocabulary and kanji. Those are two things where having a teacher doesn't really help all that much, and where an app can really help. But, again, a total beginner (which is the market that Duolingo targets since that's where the money is) doesn't need to grind vocabulary nor kanji; they need the fundamentals, which they won't really get from it.
markussss 5y ago
I've been practicing Russian on Duolingo almost every day since February 2020, and before that I had a few 30 to 100 days streaks over a few years (can't remember 100% for sure). Making you force yourself into learning a new language is what makes Duolingo stand out among all the other tools I've tried. I think that the tool that is the most efficient in teaching you a new language is the tool that makes you force yourself into learning, at least a little, every day. I'm quite sure that consciously practicing 10 minutes every Monday-Saturday for a year [1] is more efficient compared to practicing two hours every day for one week every three months [2], even though it adds up to approximately the same amount of time spent yearly on learning the target language. [1]: (10 minutes * (365 days - 52 Sundays)) / 60 minutes per hour ≈ 52 hours and 10 minutes [2]: 2 hours * 7 days * 4 quarters = 56 hours per year
globular-toast 5y ago
I see strong similarities between language learning and gyms. The vast majority of people who sign up to gyms do not use them to become athletes. Many of them don't even go. Some of the largest gym chains are objectively rubbish gyms. They don't have the best equipment and often forbid effective exercises because they are perceived as intimidating for casual members. Their business model is in attracting new paying members. They have no interest at all in making their users fit. Ironically the presence of fit, strong and healthy people at the gym discourages unfit people from joining. These language learning cons are all the same. They reel you in with the promise of language learning being easy. This is essentially a lie. Language learning isn't exactly "hard", but it takes dedication and persistence over a long period of time. They don't tell you that. The few who "make it" from Duolingo do so via other means in the end.
engineeringwoke 5y ago
I agree that the support for east Asian languages is horrendous, but Duolingo is optimised for romance languages, as it should be. The average person just wants to learn a little French as a hobby, and that's okay. I personally love the app and have used it as a stepping stone twice to get to content on YouTube or Twitch where you need an A2-ish level to start participating. Too many programmer types lose the forest for the trees. They think purely in terms of their definition of "better", not their customer. Flashy UI, marketing, and gamification are what most people want.

ummonk 5y ago

Gamifying is an important feature though, as it makes it easier for students to maintain motivation to keep learning.

kouteiheika 5y ago
Yeah, I agree. Even if a tool is effective if you can't motivate the user it's all for naught. I just wish that concentrating on this aspect wasn't so disproportionately rewarded by the market; after all an educational tool should primarily be judged by how effective it is in actually educating, and the "fun" factor should only be a bonus, where with the current state of affairs it kinda seems like it's the other way around.
- Vinnl 5y ago
  But the point there kinda is that gamification is not a "fun" factor, it's what makes them more effective in actually educating. If you have the best learning strategy, but you don't actually stick to it, it won't help.
  - kouteiheika 5y ago
    Indeed, it makes them more effective in actually educating, but the effect is multiplicative, and not additive. The tool that is gamified needs to be effective in the first place, so ideally you'd need both.
JCharante 5y ago
Except I play some video games that really motivate me to keep playing. Both video games and Duolingo accomplished this by eliminating the learning portion from their software to boost retention.

stu2b50 5y ago

What are some recommendations for learning apps/websites?

davidzweig 5y ago
Which language? :) Edit: I enjoy Audiolingual drills: https://fsi-languages.yojik.eu/ https://fsi-languages.yojik.eu/ These courses vary, some very good. Chinese and French in particular. For Russian, 'Modern Russian' (Clayton). 'French in Action' video series is excellent. Assimil books are helpful. Software.. eh. [I make language software :D]
emsy 5y ago
I’ve tried Duolingo, Babbel, Rosetta Stone, Lingq and some others I can’t even remember the name of. First of all, you won’t learn a language using only an app. And you will need a lot of intrinsic motivation and grit which an app can’t help you with. That being said, the best app I ever used and still use is Lingq, because it guides you towards using the language in real life situations with a lot of context, and you can use their audio clips to learn in the car or during Sport etc.. The other apps only ever have you translate words and context free sentences, which is a futile approach.
wraptile 5y ago
Generally app/website courses suck because of conflicting business models. You need to seriously read and understand grammar to learn language and employ spaced repetition memorization. The apps usually do the later not the first one because holding attention span is hard, writing good courses is harder and it's easier to monetize dictionary building than reading material. Common plan is to pick up a well written book for your language and do flash cards + media/live tutor suppliment. Currently I'm learning Thai: I have a modern book course (Becker's series) + voice Anki flash cards + some media on youtube or netflix movies (I've tried pimsleurs but it made me fall asleep)
yepguy 5y ago
Check out Refold[1]. IMO it is the best place to learn how to acquire a second language. It's not an app or a course or anything. Its approach is almost all self-directed, but it lays out everything you need to do and provides a Discord community to help you. [1]: https://refold.la/ https://refold.la/

emsy 5y ago

As someone who has learned and is learning multiple languages this rings true to me. Duolingo is very good for getting started and acquiring the basic vocabulary. but it plateaus rapidly. After that you don’t get much from it and you should use something that uses mini stories with a larger context and preferably audio read by natives (Small aside: Duolingo has awesome mini story podcasts for Spanish and French which they should incorporate into their app). The best tool I found for the mini story approach is Lingq, which is not so good at getting you started though. You’re also right about the gamification. This is something Duolingo absolutely nails. A friend’s mother is addicted to doing her English lessons and has a 2 year streak. But she also barely speaks English yet, which is unsurprising to me given the fact she only ever uses Duolingo.

mike_h 5y ago
Hadn’t heard of LingQ til now, this seems really good.
- emsy 5y ago
  The founder (Steve Kauffmann) is a polyglot with a youtube channel where he talks about effective strategies for learning languages, sometimes even in other languages. I can’t recommend it enough.
- personlurking 5y ago
  It does seem good, til you compare price ($8-40/mo) with its limited use cases.
  - emsy 5y ago
    What do you mean by limited use case?
    - personlurking 5y ago
      When I used the free trial, the main activity was centered around pasting text from other sources/sites, then marking the words you recognize. To me that's not worth $8/mo, much less $40 at the Premium Plus level. I realize there are a few other features to LingQ but the cost-benefit of the lowest tier didn't make sense to me. I have free dictionary apps that remember every word that is looked up. I'd rather input those lists into a spaced repetition app and receive essentially the same benefit without paying.

bawolff 5y ago

I mean, at some point, success is the definition of success. Perhaps you're misinterpreting what the average consumer wants in a language learning app.

kouteiheika 5y ago
Initially that might have been true, but now that I thought about it for a little while I don't think I am. I just personally find it sad that for your average consumer the educational aspects in an educational app are significantly less important than non-educational ones, and that the perceived value of a product vastly overshadows its actual value, that's all.

AussieWog93 5y ago

What app, if I may ask?

kouteiheika 5y ago
This one: https://jpdb.io https://jpdb.io But be warned - it is still very rough around the edges, and still very much a work-in-progress. (I'm only working on it in my spare time as a side-project.) It's definitely not ready for prime time yet, and won't be for quite some time still. That said, I do already have users. One of my users is using it to help him consume native Japanese content really early. He's learning Japanese since September of last year, and he started using my app a few months ago (probably a month or two ago; not sure, I'd have to check) and he's already reading some of the easier Japanese visual novels and web novels, and watching anime with only Japanese subtitles. This - I personally believe - is the best way to learn a language: by immersion. So I always wanted to make an app that will short-circuit the process and will allow you to start consuming content a lot earlier that you'd ever be able to do using only traditional methods.
- devilsbabe 5y ago
  Very cool work! Paradoxically, I've been slacking off on practicing Japanese diligently ever since I moved to Tokyo. I've been meaning to get back into a good routine so I'll give jpdb a shot. I'm curious how you did this: "We have 16785 prebuilt decks with vocabulary from 1124 different anime waiting for you." Did you write a script that calls subs2srs?
  - kouteiheika 5y ago
    Well, essentially yes, it is based on text analysis, but it's a lot more complex than "a script that calls subs2srs". (: (My whole codebase is over 100k lines of code, all written by me.) I have a unified morphological analysis engine that I use for every type of media that I have in my database, and I use it to generate stats and vocabulary lists.
    - 5y ago

ehnto 5y ago

I am currently learning Japanese on Duolingo, I am wondering how long it has been since you tried their Japanese course? I noticed it changed significantly between now and when it first released. It certainly doesn't feel like a course like traditional education provides, more like a flashcard vocabulary and grammar builder. I have had to Google plenty of things while using it because nothing is ever actually explained and Japanese has plenty of arbitrary rules, but I don't feel that's all that bad. I am also curious if you need beta testers for your app, I would be open to giving it a go.

kouteiheika 5y ago
Yeah, I did look at it recently too, and I still think it's not very good. I mean, for a total beginner it is okay-ish for some basic vocabulary, but at that stage of learning a textbook approach is a lot more effective in my opinion. You're better off just grabbing a copy of Genki, going through that with a teacher, and once you have the fundamentals down and can efficiently study on your own start doing vocabulary flashcards in a better app and consume easy native content. Sure, I'm always looking for beta testers. (: Although my app is very much public and open to everyone. Just don't expect miracles since it's currently only a work-in-progress side project that I'm working in my spare time and slowly improving: https://jpdb.io https://jpdb.io

nosefrog 5y ago

Hey! I went through your comment history, and you seem pretty similar to one of my friends haha. If you're down for a fun experiment, email me at my username @gmail.com.

kouteiheika 5y ago
That depends on what exactly the experiment entails. But since you're the one suggesting feel free to email me with the details. (: My address is @exia.io; you can use any username since I have a catch-all for the whole domain.

shusaku 5y ago

Don’t you think the whole market of language learning apps should be looked at with suspicion? I talked to my brother-in-law the other day, and he said he hired a tutor in the Philippines to teach his kids English. Unsurprisingly, it was super cheap. Low paid labor might be hard to get for Japanese, but it seems like for most languages people want to learn lessons over Skype will be far greater value than even a well designed app.

kouteiheika 5y ago
I think it depends on what exact level you're currently at in your language learning. If you're a total beginner then I personally think that having a teacher who can actually explain things is invaluable. If you don't have the basic rules of the language internalized and you don't have the fundamentals down you're going to have a harder time doing it on your own than someone who can just ask a teacher and immediately get an answer. I think the sweet-spot for language learning apps (at least for Japanese) is learning vocabulary and kanji. Those are two things where having a teacher doesn't really help all that much, and where an app can really help. But, again, a total beginner (which is the market that Duolingo targets since that's where the money is) doesn't need to grind vocabulary nor kanji; they need the fundamentals, which they won't really get from it.

markussss 5y ago

I've been practicing Russian on Duolingo almost every day since February 2020, and before that I had a few 30 to 100 days streaks over a few years (can't remember 100% for sure). Making you force yourself into learning a new language is what makes Duolingo stand out among all the other tools I've tried. I think that the tool that is the most efficient in teaching you a new language is the tool that makes you force yourself into learning, at least a little, every day. I'm quite sure that consciously practicing 10 minutes every Monday-Saturday for a year [1] is more efficient compared to practicing two hours every day for one week every three months [2], even though it adds up to approximately the same amount of time spent yearly on learning the target language. [1]: (10 minutes * (365 days - 52 Sundays)) / 60 minutes per hour ≈ 52 hours and 10 minutes [2]: 2 hours * 7 days * 4 quarters = 56 hours per year

globular-toast 5y ago

I see strong similarities between language learning and gyms. The vast majority of people who sign up to gyms do not use them to become athletes. Many of them don't even go. Some of the largest gym chains are objectively rubbish gyms. They don't have the best equipment and often forbid effective exercises because they are perceived as intimidating for casual members. Their business model is in attracting new paying members. They have no interest at all in making their users fit. Ironically the presence of fit, strong and healthy people at the gym discourages unfit people from joining. These language learning cons are all the same. They reel you in with the promise of language learning being easy. This is essentially a lie. Language learning isn't exactly "hard", but it takes dedication and persistence over a long period of time. They don't tell you that. The few who "make it" from Duolingo do so via other means in the end.

engineeringwoke 5y ago

I agree that the support for east Asian languages is horrendous, but Duolingo is optimised for romance languages, as it should be. The average person just wants to learn a little French as a hobby, and that's okay. I personally love the app and have used it as a stepping stone twice to get to content on YouTube or Twitch where you need an A2-ish level to start participating. Too many programmer types lose the forest for the trees. They think purely in terms of their definition of "better", not their customer. Flashy UI, marketing, and gamification are what most people want.