Flashcard learning is NOT efficient

[@jeffmcquillaninla]

James - I misspoke at 33:10. The correct calculation for 2% unknown words read at 150 wpm is approximately **22** words per hour for incidental acquisition, not 27 words per hour. I had in mind the 15% and read out the wrong number. The 5% calculation was correct. The basic point remains the same, however: When we look at vocabulary acquisition rates, the supposed dramatic differences between rote memorization and incidental virtually disappear. Thanks for having me on! Jeff

[@futuremultilingual6134]

Thank you. It is so lovely to see that people are really reflecting on these ideas.

[@songandwind72]

Thank you for your various contributions. I'm a Japanese-to-English translator and have experimented with various "learning" methods over the years. Anki never really helped me. I think what happens is that people who use Anki do so much immersion that they acquire the words/sentences that they added to Anki and then think that Anki was the cause.

[@michaeldmytriw1047]

That math at the end comparing rote memorization and reading was eye opening. Context, context, context. “If it’s not hard, then your not learning!” Forget that! I want to acquire it anyway!

[@futuremultilingual6134]

We look at how you can and can't use knowledge you have memorised in the second part.

[@johnsch8634]

Excellent interview! Dr. McQuillan is my favorite. I've always wondered about how researchers approach the question of gradual/partial learning of words. It's so easy to give people a list of words and say "memorize these" and compare that with a group who read a 600 words text. What do you know?! The memorization group did better at 'learning' the words. But it is so much harder to measure the gradual progress in acquiring a word that people make over many hours of reading or listening.

[@futuremultilingual6134]

Great. We go deeper in the second part.

[@AConnorDN38416]

And do these studies actually study if people really understand the words on a conceptual level or do they just know a general meaning? When I spent a lot of time on flash cards I “learned” a lot of words, but there were so many in that list of learned words that I really had no clue how they were used. As a result, I would come up with some really nonsensical phrases based on these incomplete understandings of the words I had learned using flash cards.

[@futuremultilingual6134]

@@AConnorDN38416 what Dr McQuillan suggests in the second interview is that even when it looks like someone has learnt a word through flashcards, it doesnt help them comprehend texts. In terms of actually engaging in a language it seems to be useless knowledge

[@andrewrobinson2985]

For the most part, I agree. You will never learn to speak a language, or even a single word, well from anki alone. However, there is a blind spot for languages like Japanese, which has a non-phonetic and logographic script. I'll say up front that I am biased as a Japanese learner in my stance on flash cards, and if I were learning any other language, I don't think I'd use anki. Reading is an extremely helpful activity for second language acquisition, as I'm sure you'd agree. It's the perfect opportunity to see words in context, understand them at your own pace, and pick things up incidentally. In most languages, you can get started immediately. However, given the nature of written Japanese, there is no way to reliably guess the pronunciation of a new word, and learning by ear is also very difficult for English speakers because there are very few borrowed words from English. This poses a huge barrier to entry for reading, and those words that you cannot pronounce end up quickly forgotten, as if just noise. Enter the flash card. This is all anecdotal, of course. While having a word and its definition and pronunciation on a card and going over it isn't going to help you learn to use that word, it will help you to recognize it when you see/hear it in context. Over my first year of studying the language, I put just under 12,000 unique words on anki and kept up with my reviews each day. As I review those cards now, I fail to recall ~15% of the cards each day. So, in the past year, let's say I've learned to recognize and pronounce roughly 10,000 words. Through that process, my ability to read books went up drastically. Where it would take me north of an hour to read a couple pages of a Japanese childrens' novel before, by the time I quit making new flash cards, I had become able to cover the same ground in an adult-oriented novel in less than half the time, with greater comprehension, and without needing to look things up. My total time spent reviewing in anki was just under 500 hours, so not a small time-commitment, and about a third of my total time spent that year learning the language. The ability to read that doing flash cards gave me led to me reading more, and that reading more led to improvements in my listening, writing, and speaking, improvements that would have been much smaller without those ten-thousand trips to the dictionary. In this way, I am convinced that using flash cards was an extremely efficient way to prep myself to absorb the language. For phonetic languages, for languages with lexical similarity to one's native language, and for languages you're already relatively proficient in, I agree. Flash cards aren't important and 100% of your time as a learner is better spent consuming level appropriate content. But with languages for which that isn't true and for which there is a barrier to entry that could be broken through quickly by dedicating 20-50% of your study-time toward rote memorization, I don't think the same applies. I hope my perspective can provoke some thought about how flash cards may help language acquisition in a more indirect way than is addressed here. Great interview and great video as always. Lots of interesting data points to chew on. I look forward to hearing the next part!

[@jamesmccloud7535]

Honestly for languages like Japanese, what learners really need the most is more beginner content that can be understood from zero. Something like the channel "Dreaming Spanish" where you assume the listener knows nothing of the language and can incrementally build up comprehension over time without resorting to translation. Even in popular languages, there is still a lack of quality comprehensible input available. On the other hand, there is an oversaturation of books and videos teaching you grammar and vocabulary. Since Japanese uses a different script, reading from the start would not be reliable of course, so listening is likely the main activity in the beginning. Studying the new alphabet system is fine I think, but taking words out of context to memorize them is different. I mean, most people in history were illiterate as education was a privilege, yet people still become fluent. Japanese children didn't need to spend hundreds of hours painstakingly using flashcards to acquire their language, they just received input from their environment. Japanese is a language like all others so it can be acquired the same way as well by adults.

[@johnsch8634]

So you were able to read at a rate of a couple pages in a half an hour after dedicating 500 hours to using flashcards? And you figure you've learned 10,000 words in that time, so 10,000/500=20 words per hour... doesn't sound like "quickly dedicating your study-time" to me. And that's 1/3 of your time, so you've spent 1,500 hours total in a year? so about 4 hours a day? and 20/3=~7 so you've hit 7 words per hour like Dr. McQuillan talks about. I fail to see your claimed advantages to rote memorization...

[@andrewrobinson2985]

​@@johnsch8634 average reading speed is 8-9k characters/hour, to be honest I'm just unsure of the page count because I often read digitally. Consider the alternative, however, that is not being able to read at all because I can't pronounce any of the words. > I fail to see your claimed advantages to rote memorization... The advantage is that you literally cannot read words you don't know how to recognize or pronounce. You can sit and look at a sentence containing 「氷山の一角」 all day and maybe even figure it out by context, but if you don't ever memorize those word's readings, you won't be able to recognize or reproduce it. Because the script is so foreign and unpredictable, the alternative to rote memorization is just ignoring the words, and you don't exactly learn incidentally doing that. Rote memorization at some point in the process of learning this language isn't an "advantageous thing" it's a necessary thing.

[@andrewrobinson2985]

@@jamesmccloud7535 That's a very important point too. Because there is a lack of this content, the gap ends up being from textbooks to native level content, and the only way to bridge that gap is to memorize things. I will say though, in my experience, after switching from a primarily listening-based approach to one of rote memorization and lots of reading that my holistic understanding of the language improved at such a greater rate than it had been improving at before ("lots of reading" is key here as well. Memorizing is pointless if you aren't also then taking these words to be found in context). >but taking words out of context to memorize them is different once again, what i'm describing as important is not rote memorizing the words meanings, but memorizing the pronunciations. You cannot memorize the pronunciations of the individual kanji characters, there is no reliable way to predict a reading without simply recalling the specific word's pronunciation (and pronunciation is exceedingly rarely contextual). > Japanese children didn't need to spend hundreds of hours painstakingly using flashcards to acquire their language (actually, they do. rote memorization of the writing system, often involving handwriting sheets and flash cards makes up a large part of Japanese childrens' language education) we aren't Japanese children. Children get thousands upon thousands of hours of input and years of trial and error before they become capable in and commanding of a language. While it is an option to learn from audio only, when reading is such a great tool for language learning, you're shooting yourself in the foot by not breaking down the barrier it takes to get there.

[@jamesmccloud7535]

@@andrewrobinson2985 You are talking about the writing system and the script that they use, all of us as children also studied these things as languages have different writing systems. This is not unique to Japanese. What I am referring to is language acquisition, note the word I used. You can be fluent in a language without ever learning how to read and write, not saying that you shouldn't learn these things as well. It might be even faster if you can read because you get more exposed to the language.

[@Alec72HD]

After switching to MONOLINGUAL dictionaries it seems I never forget a word I look up. When a word appears in content that I LIKE, I look up it's definition and usually never forget it. My memory was terrible when I was using translation.

[@jasonkappes9013]

I like when you bring on researchers and discuss with them. Good video!

[@emmcdermott574]

This is exactly what I needed to be reminded of. I spend far too long on Anki each day at a cost to immersion. I will probably never give up on Anki completely (I *may* be a slave to the streak!) but it's a reminder of where I need to be focusing. I do find that finding words in my reading that I've been revising on Anki solidifies them in a way I haven't otherwise found. Off to watch the next one...

[@AConnorDN38416]

My roommate questioned if the reading speed accounted for the time it takes to look up unknown words. I think reading on a Kindle I would say that 150wpm is probably about my reading speed including lookup time when I’m reading on a kindle (though I typically won’t look up every unknown word and will just take a guess at what it means and move on so reading doesn’t get too laborious). The numbers really are pretty compelling for just replacing whatever time I’d spend on Anki with more time reading. These days I typically spend about 5 minutes a day on Anki and the rest is all reading and listening mostly just because I know I have learned a lot of words that I saved to Anki and it’s hard to want to take the leap and just leave it behind entirely. It probably has to do with that whole sunk cost fallacy since I’ve spent so many years building my Anki decks. Reading books is certainly more enjoyable than doing flash cards.

[@futuremultilingual6134]

So, the assumption is that we are reading texts with 2 or 3 percent new words. The idea is that the reader concerns himself with meaning of the text and doesnt lo9k up the words. You strengthen knowledge of words you have seen before and incrementally acquire the new words

[@AConnorDN38416]

@@futuremultilingual6134 that is a remarkably high retention rate then if we're talking learned words without even looking them up. I've wondered for a long time how often we should be looking up the meaning of new words considering the tradeoff of how much it slows down your reading. Now I feel encouraged to look up words even less, haha.

[@IulianYT]

5:20 - yes, it is very good and fun and useful to meet the word in a context and figure out what it could mean. But, only *after* you know some words. And to learn *some* words - spaced repetition is a thing. Of course you can't learn a language using flashcards, but they can improve your vocabulary.

[@futuremultilingual6134]

The process of figuring out is not at a conscious level. You have to remember that this is reading and not contextualized listening. Most people can deal with 5 percent unknown words when reading. They will incrementally and implicitly (not at a conscious level) acquire these words. Comprehensible input (videos, stories) doesn't need this. Also, cards give you a different type of knowledge that can't be used in the spontaneous processing of language. We need to be careful of confirmation bias.

[@IulianYT]

@@futuremultilingual6134 but still, what is the aproach when ratio of known words is near 0%?

[@arccosinusopinion2323]

Great interview. I've seen it probably a dozen of times but it is still as exciting as it was the first time

[@Kate-vd3hl]

I think SRS is like jamming your foot in the door to be able to acquire words while reading or listening faster. You memorize the meaning in your own language - understand it immediately when you see it in context - and then through continuous exposure it gets ingrained in you. I also think with things like lingq the ability to immediately click on a word instead of straining to try and remember actually does hurt you. For me at least there seems to be something vital to acquiring a word in the act of racking my brain for its meaning.

[@futuremultilingual6134]

There are two things going on here 1. Front loading words which if what you suggest is tight "understand it immediately when you see it in context " works. Dr McQuillan suggests this is not the case in the other video.He refers to research that says words memorized up front aren't available for spontaneous comprehension. The second is the LinQ point which means using translation to make words comprehensible. I agree with you here.

[@Paulus8765]

I'm convinced. So where do find understandable language materials when you're starting to learn a foreign language?

[@futuremultilingual6134]

The best way is socially. Find a crosstalk partner. You can see Pablo (dreaming in Spanish) for an explanation. Otherwise find videos on KZbin that you can understand through context. If you tell me which language I will tell you which I think are good

[@Paulus8765]

Thank you. Swahili, my son's first language, and Latin, the language at work.

[@futuresekkai]

This video is so interesting, earned a sub, currently doing my 20 words of japanese a day on a SRS whilst watching 🤣

[@futuremultilingual6134]

Thanks. I am putting out part two tomorrow

[@jonallen7619]

Clearly didn't learn anything from the video and will not be able to use those 20 words

[@James_zai_dongbei]

All I know is that I was getting nowhere with reading Chinese before using flashcards.

[@futuremultilingual6134]

This is a problem for education researchers. They can go through the process, read the ideas, reflect, discuss, try to eliminate bias and habitus. Then somebody comes along and says. I have reflected on my own process which I believe is objectively transparent to me and that is enough to refute you.

[@James_zai_dongbei]

@@futuremultilingual6134 I thought you were "keen to hear from those who wish to take the centre ground or prefer to say memorising is more efficient"? Your response doesn't seem to be fostering discussion to me... What strategy would you use to read Chinese? If perfect graded input materials existed, I would use them, but they don't.

[@futuremultilingual6134]

@@James_zai_dongbeiLet's discuss then. In terms of the Chinese reading, I will have to refer you to videos I made with Dr Diane Neubaur. Here's what I would like you to consider in terms of how you think flashcards have helped you. 1. It is impossible to reflect on how we did something and see clearly how it happened. This is an epistemological issue. Coming to knowledge of something is a long and difficult process. 2. If you are going to try and reflect on what you did you need to go through a critical process. Put down everything you think happened and then highlight every assumption contained within. 3. Some of those assumptions will be correct, others will be wrong 4. They are all socially constructed so it might be worth thinking about where both sets of assumptions come from. It is interesting to to hear from people who take a different position but I have now received quite a lot of messages saying 'I have reflected on my own process and you are wrong" I will probably come back to the channel in the new year and this will be an area I will come back to

[@stevencarr4002]

Professor McQuillan publishes videos where he explains things like 'the verb is to wake up' and 'A pillow - p-i-l-l-o-w - is what you put your head on when you sleep.' A typical lesson from his ESL podcasts runs something like :- 'Your “neighbor” is the person who lives either right next to you or very close to you. Nancy says, “You’re new to the neighborhood,” meaning you have not been in this area before. Eric says, “Yes, I just moved in last weekend.” “To move in” means to move to a new location.' It is this kind of superb explicit teaching of vocabulary that built up Professor McQuillan's reputation as one of the finest teachers of English in America. Jeff even gives you 'a glossary with definitions of the key vocabulary'!

[@futuremultilingual6134]

That does sound pretty bad. I am going to have a look. That said I still agree with what was said in this video

[@Alec72HD]

​@@futuremultilingual6134 We really need to define what is Comprehensible Input or CI. Here it is: INPUT is anything in the Target Language or L2. COMPREHENSIBLE means it is understood through L2 and WITHOUT the use of Native Language or L1. Visual cues and context are perfectly acceptable. In this regard any discussions regarding L2 done entirely in L2 are still Comprehensible Input, but simultaneously explicit institutions also. You get 2 for the price of 1.

[@futuremultilingual6134]

@@Alec72HD This is the assumption I think. However, research by Krashen and lots of evidence gathered together by Garcia show that in general multilingual classrooms are more successful than L2 monolingual classrooms.

[@Alec72HD]

@@futuremultilingual6134 Do you know what is a "multilingual classroom?" That's when students don't share the same Native Language. It's very common in ESL language schools in US, where Krashen worked. This way the only common language for all the students becomes L2 English. In comparison (in US) there are private English language schools just for one ethnicity. They teach through grammar-translation and the teachers are NOT Native speakers. Needless to say those students are total failures.

[@futuremultilingual6134]

@@Alec72HD What all this research found is that Spanish/ English translanguaging classrooms lead to better results for English acquisition than English only classrooms. If you don't like those conclusions you have to show some actual evidence and not just speak down to people.

[@Luna________]

Don't all those efficiency measurements fall a bit flat just by them not being able to assess how well you really know the vocabulary? Understanding a word in multiple different contexts and the nuances of is a different thing than being able to give a definition of it or remembering the translation. I once found anki deck for spanish slang and i realized even had i used it extensively it would have never tought me how or when it is usable or adequate. I think when you learn a language through real contact with it, there forms a kind of meta-knowledge that is crucial. In flashcards almost all of that gets lost

[@futuremultilingual6134]

This is essentially what we move onto in the second part of the interview. It gets worse for memorising. Not only os it inefficient but the knolwedge you are left with is not useable and not just speaking. People who had memorised lots of words couldn't even comprehend them in a text.

[@Recoveryanonymous]

Everyone is always recommended anki too

[@Recoveryanonymous]

What is your opinion on something like Pimsleur Japanese?

[@futuremultilingual6134]

@@Recoveryanonymous I don't know what that is I am afraid. There is a secind part to this interview and at the end of that I hope people will feel able to reflect completely on menorising. For me it is a total waste of time

[@Recoveryanonymous]

@@futuremultilingual6134 all good friend. Pimsleur is a language learning software focused on audio learning only

[@williambudd2850]

If this fool James Stubbs would sut up and let us listen to his guest, we might learn something!!!

[@futuremultilingual6134]

🤣🤣🤣😘

[@ericsmith5919]

About 10 years ago, I attended the Defense Language Institute and went through the Korean Basic course. In 15 months, I went from 0 knowledge to a 3/3 on the DLPT (Defense Language Proficiency Test.) This was accomplished using study, flashcards, and all of those other traditional methods that supposedly don't work. This video, and the genre of video to which it belongs, appear to me to be simply a way to excuse laziness. "Don't bother actually trying or putting in hard work. Not only is it unnecessary, it actually doesn't help at all!"

[@futuremultilingual6134]

This is bias. How much like hard work you judge something to be and how effective it is are not related. It's just the cognitive bias of someone who is perhaps too lazy to examine such things and so leaves them unchecked. So lazy. "Don't bother actually trying or putting in hard work. Not only is it unnecessary, it actually doesn't help at all!" You know what else is laziness. The idea that just doing something gives you sufficient knowledge of how it is done to be as confident as you are. Dr McQuillan and I have both spent many years studying language acquisition, researching and teaching. You took the lazy way out and demand that egocentric reasoning is enough. So lazy

[@IulianYT]

Well, I lost half an hour for nothing, if you came up here, my IMO tl;dr: using *only* flashcards to learn language is next to impossible learning words by just meeting them in context is more efficient for long term memory but, they spoke about scenarios when reader knows 95% of the words, 98% of the words, 99.5% of the words. And didn't say a single word about people who start to learn a language, so they know only 0(zero)% of the words in a text in new language. Zero. None. Nothing. If the human knows the alphabet, he/she will see letters arranged in some random order, not sure how even to read correctly. Also they recognize that flashcards give you at least some results. So, IMO, if you start learning a new language by yourself, you can absolutely start with flashcards with spaced repetition, so after a few weeks, in very very simple sentences you will start to know 30-40% of the words, you may find out the remaining by dictionary. And after a few months, you may get to 50% in sample texts. And maybe after few years you will get to 80%. At this point you may leave flashcards and return to this video, and improve your language when you already know many words.

[@futuremultilingual6134]

There's a massive but there. In the second video we talk about the type of knowledge flashcards give you. It is not the type of implicit knowledge we can use spontaneously but the type of knowledge you can recall when you are able to focus on the recall. So not in the processing of texts or audio or in the production of language. I think the other video is called "flashcards don't give you the type of knowledge you need". The reason he talks about 95 percent known words is because it is reading on a page. Listening in context would not require this amount of known words. It wouldn't really require any if the input was comprehensible

[@gidmanone]

Both of you couldn't have been more wrong. The best approach is to frontload with rot memorization (I'm not recommending a specific strategy here) and then follow it up with what the guys in the video are suggesting. It is the ultimate game changer. It's shame these language acquisition channels are like religion now.

[@futuremultilingual6134]

You need to offer more. Why? How does this relate to language acquisition, cognition, motivation etc. Does this produce implicit or explicit knowledge or doesn't it matter? What is the source of this revelation is it years of research, reading, study or a claim to be able to gain knowledge from self-observation. We go into a great amount of detail to explain our position. You need to do the same if you want to convince

[@thomasbayer2832]

Your video is NOT efficient!

[@futuremultilingual6134]

Lamentably, I am a rambler but that is more fun tha memorising words on cards