It seems like the core things that Anki needs are new user experience improvements, and algorithm updates. SM2 really shows its age as compared to other algorithms.
That said, even SM-2 is probably vastly superior to just not doing SRS at all.
They state in contrast:
We do not dismiss the work behind FSRS. It is a commendable open-source effort and a marked improvement over ancient algorithms like SM-2.
For context, Anki uses SM-2's algorithm (albeit apparently heavily modified for various special cases) if FSRS is not enabled.
For me really learning a word means:
(1) Knowing how to say it.
(2) Knowing how to write it, meaning the Chinese characters, of course.
(3) Still remembering (1) and (2) after at least a month.
(4) Being able to actually use the word correctly.
If changing their SPS or the promises about an SPS motivate the learner, then great, they're putting in the work and time to learn, but I doubt that the effect of changing the SPS is as large as some people claim.
For example I used a tool that supposedly uses FSPS, but it did have a low maximum for the duration you don't have to practice a word, and no way for me to "ban" a word, so that it asks me in 6 months or something, and simple words kept coming back, especially after not learning for a few days. I didn't make much progress using the tool, even though it had FSPS.
You learn the words for a day (you're able to match the sounds and meaning to it). You will forget a lot of them tomorrow, so now you have to re-learn them. This is just how Anki works. You keep learning and re-learning until they stick for a prolonged period of time. It's common for Japanese learners to add between 20 or 30 words to their learning queue.
If you understand how Anki works, you will also understand how the word learning is used in relation to its flashcard mechanism.
And with all due respect, someone claiming they learn 20 words per day, in Mandarin, is an almost outrageous claim. If you think that "learning" is commonly agreed upon to mean "memorize for a couple of hours", then please show me the research into the meaning of the word, that proves your claim. While I have explicitly stated, what _my personal_ definition is, you are claiming to be knowledgeable about a "commonly agreed upon" definition. That is an impressive claim in itself. Let us all hear that definition, that is so commonly agreed upon, so that we can gain from that.
What you call learning, I call "training" or "practicing" or "revising". Now the onus is on you to prove to me, that indeed as you claim there is some commonly agreed upon definition, specifically in the area of learning Mandarin, that proves, that my definition is off.
And I will have you know, that I am learning Mandarin for some 10+ years, and have a lot of experience in that area. I know what counts and what is important.
I don't bother with the Hanzi past being able to recognize them. I want to be able to talk to people and, if I have to, use a pinyin keyboard to write basic sentences.
So only 1 & 4 are really relevant, 2 is what Anki is designed to do.
When people in the language learning community say they "learn 20 words a day", they are referring to New Cards Added. It is a metric of input and initial encoding, not guaranteed permanent storage.
In Anki, "Learning" is literally a specific phase (the red cards). You introduce the card, you pass the initial threshold, and then the algorithm handles the retention over the subsequent weeks. You are conflating the process of learning (adding new information) with the result of mastery (long-term active recall).
Would you say a native English speaker doesn't know the word "they're" if they keep spelling it "their" even if they use it correctly 100% of the time?
How does this opinion hold up if you consider that spelling wasn't standardized three centuries ago. Did no one know any English words until spelling got standardized in the 1800s? Do illiterate native speakers not truly know any words? Do children not know words?
How would your opinion change if you knew that plenty of native Japanese and Chinese speakers can't write the characters they can read anymore? If you don't have to physically write anymore, you lose the ability to write the characters. This is true of even educated adults in Japan and China. When I was a university student (I'm not Japanese), I could write kanji that my 30+yo Japanese friends had forgotten, but no one would say I knew how to use the words better than they could.
EDIT: And in any case, 10 new vocabulary words per day is extremely easy. In my experience having studied two foreign languages at the university level, that's pretty much the bare minimum expected to get an B in class.
Why do you keep harping on Mandarin in particular? Do you think it's harder than other languages to learn new words? It's not like you have to learn new hanzi for every word. Most are compounds. It's like being surprised someone easily learned how to spell "lighthouse" because it's got a silent "gh" and a silent "e" and the "ou" is not pronounced the way you'd expect, and the "th" in the middle isn't pronounced like "th" should be.
The learner already knew how to spell "light" and "house" so it was effortless to learn "lighthouse."
My experience with Japanese is that you hit around 800 or so kanji and new vocab comes very easily. Even new kanji come extremely easily because they're all made up of the same parts ("radicals").
EdIT: One hour a day devoted to language study will yield 20 new vocab words a day that, over time, you'll have around 85% recall, which translates to over 6,000 new words per year (over 7,000 but then you adjust downward because of the 85% factor).
The issue is that people want to learn a language in five minutes a day, but they don't bat an eye at playing a video game an hour a day to be able to beat some level. I remember playing for hours to be able to get good at 1942 on the NES back in the early 90s.
Writing a language makes you more skilled at living in the modern world. It's not a threshold past which you must travel to count as a speaker of that language.
By cutting out the memorization of Hanzi, I am able to accelerate my actual goal of having conversations with people.
In Silicon Valley speak, I think the term would be "ruthless prioritization" .
Unless you're applying for something in China, you don't need to know how to write hanzi ever, except for very one-off instances like "I can write happy new year in Chinese"
You know how many times I've written "real" Japanese by hand since 2005? Zero. I've written my name and stuff, sometimes I'll write 愛 to show my daughters. Nothing else. Because it's a worthless skill unless you live in Japan. Not even visiting. You live there.
Of course I type all the time. But typing is speaking + reading. It's not writing. You type phonetically (i.e., you know how to say the word), and then you hit spacebar until the correct kanji comes up (i.e., you can read kanji).
The original claim was about 20 Mandarin words.
> The learner already knew how to spell "light" and "house" so it was effortless to learn "lighthouse."
This kind of comparison doesn't work properly for learning Chinese characters. Simply combining characters like that only works ~half of the time or less.
> EdIT: One hour a day devoted to language study will yield 20 new vocab words a day that, over time, you'll have around 85% recall, which translates to over 6,000 new words per year (over 7,000 but then you adjust downward because of the 85% factor).
Delusional for Mandarin, unless you have some kind of special brain putting you in some 0.001% of the population. Not even natives learn that many words in a year. That many characters they might know when reaching university, and then later forget many again. Most native adults don't know that many.
> The issue is that people want to learn a language in five minutes a day, but they don't bat an eye at playing a video game an hour a day to be able to beat some level. I remember playing for hours to be able to get good at 1942 on the NES back in the early 90s.
Well, at least on that we agree. If one doesn't put in the time and effort, then the results will reflect that.
Bad example, but to roll with it: In that case I would say they don't know it properly, since it is apparent, that in their mind there is no difference between "their" and "they're" or even "they are".
> How does this opinion hold up if you consider that spelling wasn't standardized three centuries ago. Did no one know any English words until spelling got standardized in the 1800s? Do illiterate native speakers not truly know any words? Do children not know words?
I am basing my personal definition of when I consider a word "learned" on reality, not on some "what if". If I had to map that idea of no spelling standardization to Chinese characters, then it would mean, that characters don't have standardized lines/components/parts. If there was no standard, then I guess I would consider this kind of making up how to write it on the fly sufficient for having learned a word. Thankfully there is standardization, so that is not a reality we live in.
Since I strive to not be an illiterate, I do not count being illiterate as having learned a word.
> EDIT: And in any case, 10 new vocabulary words per day is extremely easy. In my experience having studied two foreign languages at the university level, that's pretty much the bare minimum expected to get an B in class.
That depends very much on the language and course, but if it is your major, then sure, such a time commitment seems reasonable, since it is the thing you are doing. If your major is anything else and you just take an additional language course, where I come from you have once or twice a week that course. Then maybe 1 or 2 weeks you finish one chapter of a course book, which might have 20 new words, so that makes 20 words in 1-2 weeks, not in 2 days.
Typically for Mandarin the speed will also be slower than other easier to learn languages. For example at school I almost never had to learn vocabulary in English or Spanish. I just saw the words and memorized them somehow. Usage in class and often their sound and structure was sufficient , and always had good grades, often very good grades in those languages and always had good grades, often very good grades in those languages.
It doesn't work like that with Chinese characters. You are not gonna learn them (including writing them) by just looking at them a few times, unless you got an extraordinary visual, almost photographic memory. I consider myself already to have a pretty good visual memory, but still I need to put in the time and effort, and 20 words a day is way out of my league. But then again it was already cleared up in another comment, that the OP cuts out writing entirely. That's definitely a choice one can make and explains how 20 newly added words (I would still debate that that's "learned") make any sense.