Steve Kaufmann: Don't Try To Memorize Vocabulary

[Le Baron]

Steve Kaufmann always has interesting things to say, but like with Krashen you have to pay attention wearing your critic's hat. I remember some time ago reading a blog article where a fellow was stating as a 'fact' that Anki simply doesn't work. He laid out his argument, which seemed reasonable, yet the comments were full of people who said they'd used Anki very successfully. So are they liars? Deluded? Perhaps just different learners or ones who found a productive learning groove with Anki.

The only situation that should make you question a method is if you've been plugging away at it for ages and it has produced measly or no results. Or if it dries up as a method. If Anki or some other kind of recall is making you learn words, why would you stop using it just because someone came along and said 'it doesn't work'? We all know nothing is given as freely as advice and it's worth remembering that when Steve Kaufmann says 'what works' he's really saying what (he estimates to have) worked for Steve Kaufmann, because he's not running a teaching programme and we know nothing about how successful LingQ users really are.

As stated above Iversen uses direct wordlists and this appears to have worked, and since he isn't the only one it's hardly an 'exception that proves the rule'. The trick for words is to encounter them and then have a situation where you are forced to review them. Several words roll into one concept: retrieval, recall, review, repetition, context, understanding... The only dispute is: how is this done?

All that rests on the type of learner, time and lastly the question of whether the chosen methodology is productive. Reading works, because it places words into meaningful context, but you need to at least gather together some sort of useful corpus for yourself of whatever size, either before or during reading and do something with it; even if that is only skimming over them to discover their meanings so you can encounter them again in further reading/listening. Just reading and listening and not thinking about it is like a non-swimmer merely getting into the water without learning about how to swim and hoping it all works out. Luckily you only figuratively drown with TL learning!

Mass exposure, yes, but with a background of organisation including review and understanding what you're exposed to. Conscious work as well as unconscious exposure.

[einzelne]

As Confucius said, we won't go anywhere without first defining terminology.

He says language learning is not about memorization.

But At 2:18 Steve says, "I sometimes do review words in lists".
And at 2:42 he mentions he does flashcards, but with all the info on the front.

Both reviewing words in lists and doing flashcards are deliberate attempts to reinforce your memory.

Hot take: If you find a comfortable repetition routine you call it vocabulary review.
Otherwise you call it memorization.

[leosmith]

Do you know where I could read more about the roll of recalling?

Unfortunately I'm pretty bad at finding studies. But Kraut posted a useful link a few posts below yours.

Passive study is always less efficient than active recall

Yes, in a way, but we also need massive input to learn a language. I just want to be clear that I advocate a mixture. I don't think doing only anki to learn a language is a good idea, for example. For me, recall work might start out to be as much as 50% in the very beginning, and drop down to less than 5% over the course of couple years.

[Le Baron]

For me, recall work might start out to be as much as 50% in the very beginning, and drop down to less than 5% over the course of couple years.

For anyone I would hope. A few years down the line any learner ought to have learned enough vocabulary that they've slackened-off that sort of vocabulary input.

[bolaobo]

On topic: Kaufmann's approach of mass reading, reading like hours everyday, getting his vocabulary and grammar from mass exposure in the context of his texts. I tried it. I think I rather need to hear or speak, maybe if I would read the texts out loud, but they are not that interesting to me - especially at the beginner level and I die of boredom from reading.
It probably works well for certain types of learners, maybe a curious type who quickly skips over interesting texts and is content with not getting everything, and also has good pattern recognition.

.

Out of curiosity, I tried his LingQ approach to learn Arabic. It didn't seem feasible to me for someone starting from zero. "Who is She" is clearly adapted for Western languages and introduces the dual in lesson 1 with no explanation! It also doesn't mark vowels so you have to be a slave to the recorded voice to actually know what you are reading. There were also orthographic errors that even someone as beginner as me could spot.

[lingohot]

I think you have misunderstood what is meant by "recall" here. It means actually putting an effort into recalling something. Passive reading, listening and Anki do not require enough effort to optimize your learning.

Is cloze deletion active recall according to you?

He build his system into the language platform LingQ. I used it a bit for translating Japanese songs 2-3 years ago, also I added my university course texts. It was helpful for that. You can import and annotate texts, the problem is that the platform will delete all of that data after you cancel your subscription, and I feel that is like blackmail to me and I'm still angry for losing all my data from that time. That is just a shameful concept.

Use Learning with Texts. It's free and all the files are stored on your device.

[leosmith]

Is cloze deletion active recall according to you?

Sure, any flashcard is as long as, if you don't know it immediately, you make an effort to recall it.
That being said, I much prefer single word or entire sentence retrieval because it works my brain harder. I think Duolingo is worthless because it is actually too easy to be effective.

[lingohot]

Is cloze deletion active recall according to you?

Sure, any flashcard is as long as, if you don't know it immediately, you make an effort to recall it.
That being said, I much prefer single word or entire sentence retrieval because it works my brain harder.

Then I've been doing active recall on a daily basis for over 10 years now. I totally agree that it's much more effective than just pure recognition cards. That would also be my criticism when it comes to professor Argüelles' approach which is very "passive" (not sure if scriptorium counts as active recall, I guess it doesn't).

[leosmith]

That would also be my criticism when it comes to professor Argüelles' approach which is very "passive" (not sure if scriptorium counts as active recall, I guess it doesn't).

He uses different kinds of "active recall" activities. From what I remember, he works through grammars and text books like Teach Yourself actively. He makes notes in textbooks and other books. Scriptorium sort of counts, but the way he does it, it's more like he is recalling how to spell. Also, keep in mind that he has natural talent - he can do stuff most people can't. From his recent AMA he stated that he never explicitly memorizes vocabulary, his languages never atrophy (maybe I misread that, because it seems to conflict with other posts he's written), he advocates learning certain languages, like the major romance languages, in parallel from scratch, etc. While I find it very interesting to read about his accomplishments and his point of view, I think it's unrealistic for most people to try to model their studies after his.

Kaufmann's method is even more passive than Arguelles'. But keep in mind that he didn't use it to learn the languages that made him famous. He used "normal" techniques to learn Mandarin and Japanese when he was younger. He used a lot of active recall back then. The language learning that he did after he started his website is not as remarkable as his earlier learning. And even with those, he eventually did a lot of production, which requires active recall, so it's hard to credit his way of starting a language passively for everything.

[PeterMollenburg]

For me, recall work might start out to be as much as 50% in the very beginning, and drop down to less than 5% over the course of couple years.

For anyone I would hope. A few years down the line any learner ought to have learned enough vocabulary that they've slackened-off that sort of vocabulary input.

Sounds a bit pompous... Fyi, that's your wife speaking, not me.

There are of course learners who progress at much slower rates than those you likely have in mind above, whether through choice (eg 1 hour/week of Spanish or 10 min/day of German is enough for me), or due to legitimate time constraints, lacking consistent effort, naïveté regarding how much time and effort is needed to learn a language to a particular level, learning casually, learning difficulties, access to teachers/content/learning materials etc etc

I'd imagine organised learners that reach lower percentages of required 'recall work' are the exception.

I'll have you know I've called the police, you'll be going to prison for several years for your above comments. Don't worry about your wife, she's fine here.