How the masters learn

Open Forum in English
1

I guess I’m not alone in paying keen attention to the learning strategies employed by some of the most successful master-polyglots in the online community. It isn’t always entirely easy to figure out exactly what these guys do, but it seems to be something roughly like this:

1.) KAUFMANNISM
This one actually is explained in some detail by Steve in many Youtube videos. It is (rather obviously) the method behind LingQ. In a nutshell: lots of comprehensible input using LingQ’s tools; not speaking before building up a good passive understanding; not going to the country for immersion before being ready to gain maximum benefit.

2.) SIMCOTTERY
Richard is something of an enigma in that he seems not to have any one single method, rather he has stated that he has used different approaches for different languages. It would seem that he often starts by taking classroom lessons of one kind or another. I believe he has stated that he sometimes uses Assimil? It isn’t entirely clear what other approaches he has, however.

3.) LAMPARIELLISMO
Luca is, in my opinion, one of the most impressive guys in the online community, because in many cases (such as Russian, Mandarin and Swedish) he seems to have built up an impressive functional command of a target language without ever even visiting a country where the language is spoken or having had any full immersion. I believe he is currently writing a book about his method, but it seems to be based (initially at least) on translations to and from the target language, based on a course such as Assimil. Later on he uses other approaches (which may include LingQ?)

4.) BIEGLERISM
Robert is an elite language professional who graduated from interpreters school. These days I believe he self-studies new languages by building a kind of “core” using frequency lists, and a course such as “Instant Immersion”, before going on to do an immersion stay in the country where the target language is spoken? He has also stressed the importance of reading in the target language as soon as possible, I believe?

5.) ROHRISM
Friedemann has recently stated that he starts out by using a simple introductory course, before moving on to massive input based study using dictionary software and a reference grammar. He also keeps wordlists - but I’m not entirely clear how these work? (I suspect that it is these lists which are the magic ingredient, so to speak.)

So this is, very roughly, what some seasoned language learners with a track record of success do. I wish I knew more of the exact details. If anyone (including the individuals in question) have any additional information about this, I would be extremely interested to hear about it. :slight_smile:

2

I have to say I am not really comfortable being mentioned in the list above. I have always been interested in language learning but I certainly do not have the passion, stamina, dedication nor the achievements to show for that Luca and Richard have. I have many varied interests such as reading about politics, science, I love to play my guitars, to exercise and to travel. Language learning is only a small part of that. Furthermore, I do not regard myself as particularly talented when it comes to language learning.

I also have to say that I find it hard to invest the time to learn a language without a real application in my job or private life. I think I would never learn a language just out of curiosity like Moses or Benny seem to be doing. Another thing is that I have come to realize that languages are so deep and complex and the more you learn, the more you realize how much you don’t know. This is something I experience every day in Chinese of course, but lately even in my “easier” languages as well. Sometimes I wish we could just upload a new languages into our brains like we install a software onto a harddisk.

This realization of being a dilettant also goes for all my other interests as well, if you want to be good at anything, you almost cannot do anything else. Mastery of anything takes 100% dedication, be it music, science, sport, or language learning for that matter.

Regarding my wordlists it is very simple really. In the old days with Norwegian, French etc, I wrote down most new words and phrases I came across by hand, left column in the target language and right column in German. After a while the list grew longer and then I would review the list regularly, mostly hiding the right column and trying to remember the correct translation, that’s all really. These days I use MS word, which has many advantages of course: you can search for words, it is portable on mobile devices, you don’t have to carry a physical binder with you etc.

2 Likes
3

@Friedemann: “…I am not really comfortable being mentioned in the list…”

All modesty aside, there are not that many people in Europe or North America who have a superhigh (near-native) level in 3 foreign languages and a very high functional level in two more besides…

@Friedemann: “…In the old days with Norwegian, French etc, I wrote down most new words and phrases I came across by hand, left column in the target language and right column in German. After a while the list grew longer and then I would review the list regularly, mostly hiding the right column and trying to remember the correct translation…”

Interesting. So in a way it was a bit like the “Goldlist method” used by David James?

4

BTW
The common theme running through all of these methods seems to be the importance of input - so it would certainly tend to confirm Krashen.

What I would quite like to try out is a combination of “Lampariellismo” and “Rohrism” :smiley:

5

@ Friedemann - Just out of interest, how big do your lists tend to get after say a year? Do you have lists with 10,000 words on them?

6

Colin,

the handwritten ones I don’t have anymore but they were nowhere near 10,000. I think that is because in Norwegian, French etc. you have so many words you already know by default like government, solidarity and I never bothered to write those down. For Chinese this is very different and there I am now well above 10,000. But that took me years. I always found Steve’s self-reported word counts hard to believe, like some 20,000 words in Czech in less than a year. Now he would say that those are the total number of his lingqs and not necessarily all part of his active vocabulary and so forth but still I find these numbers very high.

7

I don’t find 20,000 particularly high for a year at LingQ. Last month I did three weeks of very intensive work at LingQ by importing and reading a huge number of news articles and Wikipedia pages and got to fourth place on the activity scoreboard. In these three weeks, I produced over 7,000 LingQs. If I had kept this up for an entire year, I would have made over 120,000 LingQs.

At the end of these three weeks, I completely burned out and couldn’t bring myself to look at the LingQ reading interface for a week or two. On the other hand, my German comprehension skyrocketed, both for reading and listening.

8

Maybe I havn’t fully grasped the concept of a lingq then.

9

If you are not too worried about the translation of the LingQ being particularly accurate, you will use the previously saved user hints for most words, and when they are not available, the Google Translate translations. With these, you can make the LingQ with a single click, or you can even do it with the enter/return bottom on the keyboard. It all happens very fast. You would make several hundred in an hour of reading.

10

@Friedemann: “…I always found Steve’s self-reported word counts hard to believe, like some 20,000 words in Czech in less than a year…”

In fairness to Steve, it should be remembered that all of the case-inflected forms of Czech nouns and adjectives would get lingQed as an individual word - as would all of the different verb forms. Therefore 20,000 LingQs might actually imply no more than (let’s say) 3500 real items of vocabulary?

11

Of course if you do want to look up every word in a dictionary, it will take a lot longer (maybe ten times longer). I rarely do this. Normally when I can’t find the translation of a word straight away, I just forget about it, and make the LingQ without a translation.

12

J_4_J has a good point. If I learned the word ‘haben’ in the system, the verb forms haben, habe, hat, hast, habt, gehabt, hatte, hatten, hätte, and hätten would be ten LingQs.

13

I can agree with both sides. If you want a high word count, it’s relatively easy to save anything you come across, including inflected forms, conjugated verbs, ordinals, names (even English has several possible forms: Johnson, Johnson’s, Johnsons, Johnsons’).

We’re free to save whichever word we like, but nobody is forcing us to save slightly different forms (e.g. habe, hast, as mentioned by ColinPhilipJohnstone) nor does it allow us to save “identical” forms (for instance, consider the English verb form “played” it could be the past tense (I played football yesterday) OR the supine form (have played/had played) - different usage, really).

Somebody might learn all the case endings of Russian, Czech etc. by just looking at a grammar table, and then, it’s not necessary to save all of them.

14

@Jeff: “…Somebody might learn all the case endings of Russian, Czech etc. by just looking at a grammar table, and then, it’s not necessary to save all of them…”

In my humble opinion, that is indeed the way a profi-learner would go about attacking a heavily case-inflected language! :smiley:

However, as regards LingQ’s methodology, surely the whole point is that you DON’T learn tables of endings - rather you get reminded of things which you have encountered before? But how would this work if you were only saving words in one case?

Also, inasmuch as one is lingQing words which have not yet been learned, it would be kind of strange to mark them as “known” in some (or even most?) of the forms in which they are found, wouldn’t it? :-0

15

I think there is a missing step, common to the masters mentioned at the top of the thread, that really makes a great difference:

1: Become a very experienced language learner

Having a decade or two of extensive immersion in language learning is, as far as I can tell, far more important than the specific method you follow.

I know that many of us - including me - would like method to be more important than experience, but it seems to not be the case.

This is why the best way to become good at language learning is most probably to spend a few decades learning a few languages.

16

That’s a fair point, Anthony. But there are some specific methods which one can use, right?

For example, I was watching an excellent video on your Youtube channel the other day about the L-R method. (I had heard this mentioned before, but I never really knew exactly what it was before seeing your video. Now I am thinking of trying this out myself sometime.)

BTW
The above list of masters is not meant to be exhaustive. For example: I would happily add “Lauderism” and “Donerism” to the roll of honour! :wink:

17

I certainly do not belong on the list. I describe myself as a PolyNot, since there are many things I am very bad at, including language learning.

18

"I always found Steve’s self-reported word counts hard to believe, like some 20,000 words in Czech in less than a year. Now he would say that those are the total number of his lingqs and not necessarily all part of his active vocabulary and so forth but still I find these numbers very high. "

Which word self-reported word count? The word count at LingQ is generated by the system and is a combination of the words that I choose not to save and the ones that I have moved to known, mostly the former. The known words total is closer to 53,000 words for the 15 months that I worked on Czech at LingQ. These are the white or unhighlighted words.

The total of my LingQs is 43,500 words and phrases. This is a different list, the yellow words. Most of these are, in fact, known, I just haven’t moved them to known on the system. Yes Czech is very inflected. What does this mean in reality. I don’t know. However, when I import articles from the Internet, the new words % is only 5% or so including names and forms of words that I already know. So, in one year I have gone from not understanding any Czech to being able to read the newspaper and understand interviews that I listen to. The known words count is just an indicator of progress in a given language, a target, a goal, something to keep you going. It cannot be compared across different languages. What is there that you don’t believe?

“all of the case-inflected forms of Czech nouns and adjectives would get lingQed as an individual word - as would all of the different verb forms. Therefore 20,000 LingQs might actually imply no more than (let’s say) 3500 real items of vocabulary?”

Which language? What do you call a “real item of vocabulary”? I think the conversion for English may be more like 2.5 to 1 based on studies that we did between different languages.

“I can agree with both sides. If you want a high word count, it’s relatively easy to save anything you come across, including inflected forms, conjugated verbs, ordinals, names (even English has several possible forms: Johnson, Johnson’s, Johnsons, Johnsons’).”

In fact most people only save what they need, words that they are not sure of, and that includes different forms of words. I doubt that anyone saves many proper names or names like Johnson, Johnson’s, Johnsons, Johnsons’. However, not everyone deletes names, numbers and other non-words from their known word count although the system enables you to do so. However the number of these is not great, perhaps 5%.

“Somebody might learn all the case endings of Russian, Czech etc. by just looking at a grammar table, and then, it’s not necessary to save all of them.”

I would find that impossible to do. I have tried. It is only by seeing these different forms in action, and getting a lot of exposure that they start stick, even with regularly looking at the tables.

19

@Steve: “…Which language?..”

Well, I understood Friedemann to be referring to Czech.

@Steve: “…What do you call a “real item of vocabulary”?..”

A good general example would be the Italian verb “parlare”: I don’t think of this as “parlo”, “parliamo”, “parlerebbe”, “parlato”, etc, etc, etc… I think of it as “parl_(+verb-pattern)”

20

J, aha, so you mean that 20,000 Czech words on the LingQ count is 3,500 word families?

I doubt that the number is that low. Not all words inflect. There must be an authoritative number somewhere.

The concept of word family is nebulous. In English are act, acting, actor, activate, react, active, activity all par of one word family?