267 karma · joined November 22, 2022
I tried the best I can to test specially Japanese and Chinese without speaking those languages myself.
the app uses real human content (audio and text) for the most part, so hopefully it comes natural enough. there is still some AI though
But what you ask is not feasible, neither I nor OP have 700 millions of reviews to stack up against FSRS.
I have my own experience and a bunch of simulations, and if FSRS were better I would have switched to it.
The reason in my case btw is that FSRS is generic, mine is tailor made for languages - big difference.
It is a lot effort, I find Spaced Repetition actually very demanding – but effective, that's the point.
If you take Paul Nation's math (the famous https://files.eric.ed.gov/fulltext/EJ1044345.pdf) and instead of doing that you mine the hardest sentences you find to an SRS (basically, separating grain from chaff) you save yourself a lot of time, because you will never find a piece of text long enough where you know exactly 98% of the words.
In my case, I use flashcards primarily for reading and listening comprehension, but lately I am using it a bit more for pronunciation, and it can also work well for spelling.
From 0: phrase books + spaced repetition + easy recordings for beginners (many webs provide them for lots of languages). Don't forget to work on listening and pronunciation
A1 - A2: add active reading with graded readers + some writing
B1 – B2: keep on graded readers and SRS, add Italki tutor, podcasts and new learning materials. Mix passive and active reading / listening. Start looking at grammar
+B2: lots of comprehensive input + whatever you fancy (eg language exchanges, creative writing, etc)
If you want to know more I wrote a book here: https://thehardway.guide/ (with lots of links to other books / sources)
German -> Germany, French -> France, Swedish -> Sweden.
It's not perfect, but 99% of times it communicates what is meant to communicate, which is more than what you can say of many icons.
There are 2 big problems:
1. If you are studying seriously, you will encounter many words outside your SRS practice, so the FSRS prediction fails flat.
2. More importantly, language learning is not about recalling words.
The algorithm I use for my app has a different set of objectives, it's <30 lines of straightforward code, and it works better (I tested it for months to be sure about it)
I'll open source it soon
I did, but very differently from how most people do it.
Premade decks with single words sucks, yes. But creating your own flashcards with lots of audio and mined sentences from your learning materials accelerates learning better than anything I tried before.
And yes, it requires a lot of immersion / exposure and other things. But when done properly it works really really well.
For language, 10-20 mins / day is great, and if you are putting enough effort, 6-8 reps is enough to learn a card with 2 new concepts at ~15secs / rep, which really makes a difference long term.
I made a tool[1] with that approach in mind and tailor-made for language acquisition patterns. The key part is that creating cards (audio included) must be super-easy and fast, but they need to come from you nevertheless.
I read it because of your comment and I think it is actually quite reasonable and well laid out.
Everyone quotes the FSI estimate, but I find this one much more generalizable.
Soon you will start to find "Duolingo wrapped" posts on Twitter, and you will see that top 1% of active users spend ~5 mins / day avg. and learn a few hundred words.
Considering that to reach fluency you need around 8k words and 700-1000 hours of study + practice (depending on language, and assuming good mindset, methods, teachers and resources), I'd say that Duolingo does not work at all.
Duolingo is not for people that wants to learn a language, it is for people that wants to want to learn a language.
Yes.
Comfort: Duolingo and similar. UX designed to make you feel proficient when in fact your are not, gamification, drag and drop words, point and click on things, streaks, leaderboards, etc.
Discomfort: write your own content, talk to people a bit before you are ready, seek content a bit too difficult for you, listen to a podcast many times to see how your understanding improves with each listening, deep dive onto a subject, shadowing exercises, etc.
In short: prefer harder and more efficient things even when easier and less efficient things are available.
What, really? What am I missing? If you tell me video, yes, but a screenshot...
1x to actually write it + 1x to call it done.
If you need to pass an exam, obtain a certificate, etc. you will need a different approach than if you are just curious about a subject and explore what it is about.
There are commonalities, however. Much of the advice on deliberate practice (From the book Peak Performance) is valid even if you don't try to be a top expert.
With graded readers, thou, you can have good reading experiences with around 3k words (ofc depending on language, book, etc).
Non fiction: Maus
ie. the best and worst that humans can be.
I disagree. Duolingo makes you feel that you are being productive and you are learning (when you actually don't).
That is worse than nothing, because when you do nothing you know you are being indulgent, will feel bad about it, and react.
Duolingo is like a candy wrapped in package that says "this is a healthy vegetable".
That research you link has been conducted by Xiangying Jiang and Bozena Pajak (first and last authors). If you want to contact them about how totally objective and unbiased they are, here are their emails:
xiangying@duolingo.com bozena@duolingo.com
For me the worst is that when you indulge, you feel bad and will do something; but with Duolingo you are still indulging the same, however it makes you feel accomplished and that you made your work.
No, you didn't — You just watched the ads they wanted you to watch. Doing nothing is literally better than Duolingo.