.

^The example line of deduction, at least under (an assumption based on personal anecdotal only) my understanding, seems to be parsed without having to really be processed on the fly. The more rules/general patterns known, over time and usage, seem, to make these processes well, a lot simpler overall as they act like mental shortcuts.

If you don't mind a potential stupid question, are you describing the above process as something that someone would consciously process; outlining the processes involved -or both?

Arupan Wrote:It would be a lot more convincing if she actually spoke on the spot. I do admire her pronunciation of course but can you speak that fast in your own language without writing your speech beforehand as well?

1. Me? Are we getting personal, or something?? [1][2]Some people speak fast quite naturally.
(For the record, if I push myself, yes, I can talk that fast or even faster in my language. Not for long though - you see, this would not be my natural pace of speaking.)

2. I thought your point was "purpose of learning language", not how fast you can talk - the girl from the video I linked doesn't talk about rocket science but usual things she is comfortable with. And I would love to be able to express myself in Japanese, the way she does in Russian (which to my ears, is her weakest language), for example.

[1] http://en.wikipedia.org/wiki/John_Moschitta,_Jr.
[2] http://en.wikipedia.org/wiki/Words_per_m..._listening

.

Arupan Wrote:I don't understand why people get offended this much by someone who doesn't share your opinion...

That's nice to know how I feel...

What about "purpose of learning language" question?

.

Edit: Nevermind.

Inny Jan Wrote:

Arupan Wrote:What's the purpose in language learning then? Just so you can get by if you go on a trip or to pick up women...?

Essentially, yes.

That may be your purpose. Others may have different goals in mind.

alright talk, but i think he overstates the percentage of known words you need to study effectively, leading to his recommendation for graded readers. I'm constantly trying to figure out whatever I can when encountering foreign languages and couldn't care less what percentage of words I know, or whether my guesses are accurate or not. They get better over time, and there are these handy things called dictionaries. If I encounter something in Chinese or Korean, I'll attempt to read it skipping over bits that are too difficult. Totally agree with a previous post who said that enjoyment is a far more important metric than known words.

I also think the idea of intensively studying the first 1000-2000 words thing may work well for some language with a lot of cognates with your native language. But for other languages you really need to learn thousands more, and they're not necessarily less important than the 'high frequency' words. Also high frequency words are by definition easier to learn so don't really require focused study.

Arupan Wrote:

uisukii Wrote:...

I just described the process. Everyone should (probably) do it unconsciously. It really is parsed to some extent I guess (0 is most likely a lot shorter), but the information isn't always the same, so you have to make (a lot of) adjustments on the fly. You can make the conversation even more complex if you add a bunch of other (unknown) factors to the whole situation: e.g. your companion's age / sex / status / ethnic group / your relationship / place / new people joining and quitting the conversation / virtually everything. The brain processes all this information in real time.

Okay, cool; thanks for that.

*returns to the pavilion*

Arupan Wrote:-> As far as I remember, you needed around 6000 words in order to pass the old JLPT Lvl2 and the new one should be approximately the same (you can pass it if you know less of course considering the fact that you don't need a 100% score and that there are only a few words appearing on the test itself). So in order to be a polyglot you should just take N2 and then go on to the next language (maybe take N1 5 years later or just don't...? who knows) cause the more the better, right?

Did you just stop watching the video at 24:15? He very clearly recommended that once you reach 6000 words you move onto extensive reading and listening instead of accumulating further knowledge through intensive study. He didn't say anything about moving on to the next language. Incidentally, I switched from intensive study to extensive reading 6 months after taking 2級, and I'm doing just fine.

Quote:-> I just deduced it from the general definition of "polyglot." He didn't really say anything about -continuing learning- either, so I guess we're even on that.

Yes he did. Just watch the portion of the video at 24:15. And what he talks about from that point on is not what is typically taught in schools.

Any idea about how can we build the "islands" he talks about in the video?
Can we make a study group for that?

Arupan Wrote:-> Yes, I did mean deductive reasoning. In my opinion it's even easier to think if it's not one-on-one, cause you'll have a lot more time to listen -> so you'll have more time to process information and figure out your own standing on the matter.

Let's say that you speak in Japanese (cause you said thinking in a second language is slower). You want to ask "what's this/that?" You process all kinds of information like:
0) Think what the rules for これ、それ、あれ are (which means everything after 0)
1) The distance between you and the "thing"
2) The distance between your companion and the "thing".
3) Whether you or your companion is closer to it. How close.
4) Whether it's on the TV.
5) Whether you've already talked about it or not.
6) etc...
7) Processing the whole information
As you can see, the whole process is quite long (the more rules you know, the longer)... and come up with これ、それ、あれ in... guess what? Realtime.

Yes, it would be quite long, and we'd get slower and slower at speaking the more we learned. This isn't how it works though, we develop a set of language skills. If you had to know the rules and process them one by one, then first language learners would have to be taught the rules of grammar one by one, and everyone would know very well all the rules of grammar.

What we do instead is develop a set of embedded understanding, coded very deeply in the brain, that intuitively produces correct speech. I never think about the rules for using 'this' vs. using 'that', and when I still thought about the rules for using これ、それ、and あれ I couldn't hold a natural conversation. (Well, I still can't unless the topic is quite simple, as I have to think quite a lot still about how to structure hypothetical or conditional sentences.)

The ability to develop natural language reflexes in a second language varies quite a bit from person to person, but it doesn't really connect to intelligence (of the ordinary IQ-test style problem-solving type). There's quite a lot of science being done on where and how in the brain language is coded, but not much (that I can find) that is very practical. One of the nice things about the video was that he cited so much from the science that -has- been done that is practical. Well, 'referred to' more than cited, he didn't do a lot of quoting. Still, he did a heck of a lot more groundwork than the average fellow theorizing about language learning.

Damn it feels good to be a polynot.

I'm not about to change my ways, but I at least agree with him on two points: that reading material where you understand less than 95% of the words is not very fun, and that you should study based on frequency in the beginning. What I think is dubious is this whole idea that you can guess your way through, and also the swapping of one arbitrary goal (20,000 words in anki) for another-- the [Dr. Evil face] ONE MILLION WORDS [/Dr. Evil face] in a year thing.

I still think its a reasonable method to look up and then add new words to anki as well as add from frequency lists for basically as long as you think is useful. You'll never be able to tell when you've reached his "95%", so you might as well keep adding until you see less and less new stuff. Of course all anki'ing requires a sanity check if the word is actually going to be useful to you in some way. Still, we should be aiming to understand and remember the words we come across, at least until we've reached our threshold for comfortable reading. And for me, 95% sounds like a gross underestimate. That means you get 1 new word every 2 sentences or so. I don't want to see more than 1 a page. I think maybe after I get there, I'll consider doing his guessing thing.

I tend to think of reading as the easiest skill in language learning, and also the only area where guessing would really work. If you miss a word while listening, you'll either have to nod politely or ask for clarification. If you're missing the word for 'shoelaces' while writing, you'll have to pull out the dictionary. The guessing thing sounds to me like he's basically encouraging people to aim for a flawed understanding of what they read because its an easy way to gloss over the 5% of words they don't know. I guess maybe it sounds positive to some, but it just sort of stinks to me as a sort of feel-good way of contenting yourself with mediocrity, as well as a missed learning opportunity.

And anyways, 20,000 cards in anki really isn't that big of a deal. Because of the exponentially increasing time frames, the amount of work you need to put in to sustain that knowledge goes to zero over time. I know because I already have that many mature cards from various areas (not all Japanese). Just thinking about it, 3 of my decks containing about 10,000 cards only take me about 5 minutes daily. No one is saying you should race to 20,000 cards because that number will earn you the golden merit badge of fluency, but if you're just casually accumulating them on your way to native-level fluency, its very doable over a few year time frame. I'm not sure if its necessary to have that many. Maybe I'll be able to give a better estimate later. But still, I'm not going to just quit after 6k. I'm going to keep at it for as long as unknown words continue to hinder my reading.

.

@Haych

Don’t knock it until you’ve tried it. Or perhaps I should say, you have already tried it to great success, and just fail to recognize it. I’m willing to bet that you learned more than 90% of your native language through “guessing” and continue to accumulate new vocabulary through “guessing”. It’s only when dealing with a second language that people tend to lose their nerve and think they can’t learn anything without being spoonfed from a dictionary or word list.

The speaker in that video is not advocating for a flawed understanding of vocabulary. Nothing could be further from the truth. The process he is describing is one of constant refinement – seeing words in multitudes of situations and contexts and gradually elaborating your understanding of them. This leads to a degree of understanding and nuance that dictionaries do not – and cannot – provide. It is in fact over-reliance on dictionaries and hard-and-fast definitions that leave people steeped in mediocrity, not the other way around.

I hardly think you would say that you have a flawed and mediocre understanding of your native language, and yet, as I said above, I doubt you’ve learned much of it through word lists and dictionaries. So give guessing a try. If you think it’s painful to read through text where every other sentence has an unknown word, I can guarantee that looking up every one of those words and putting them into Anki is several times more burdensome.

Arupan Wrote:I don't understand why people get offended this much by someone who doesn't share your opinion...

When individual opinion is threatened one (or both) sides try to justify themselves.
(for example: I might ask someone to add 17+34 together to prove my point)

Err.. back on topic. The four stages seem accurate.

1.Intensive Study (foundation)
2.Extensive Reading (amass vocab in context)
3.Practice writing and speaking
4.Fluency (develop automaticity)

From what I can gather the speaker suggests that people stop learning from arbitrary word lists past the 4000 to maybe 6000 vocabulary mark.

It is interesting how the speaker chooses the 6k mark as the point at which to stop anki-ing vocab and start reading intensively.

(also interesting that this is a method one of our more prominent members has been advocating for years. nadiatims I'm sure feels quite vindicated after years of tussling with anki-addicts)

so why 6k?...
Moreover, this number will probably vary depending on the language
This number will probably vary depending on the person

I think 6k is the point at which the speaker felt that the cost-benefit equation of anki-ing became relatively ineffective vs. high-volume reading, due to a number of factors:

anki con 1) the time required to create anki cards (I assume he is manually creating cards)
anki con 2) time spent outside native media, the 'classroom bubble' problem
anki pro 1) highly efficient method of memorizing words

JUST reading pro 1) exposure to high volume, natural rhythm, natural frequency, native prose, native context
JUST reading pro 2) depth and breadth of vocabulary
JUST reading con 1) no way to memorize unknown words efficiently
(remember that the speaker's recommendation is to stop ank-ing at 6k and JUST read)

I would say for a couple reasons that more vocab study is better for ME learning JAPANESE..

1)Japanese has a relatively high amount of vocab required to read normal texts.
2)with the core 10k deck we have a very efficient way of putting those next 4k words in our heads with minimal fuss... totally worth it in my opinion...I can read easily now.

After that I agree the advantages of anki-ing are outweighed by high-volume reading...
HOWEVER, others may feel reading while srs'ing unknown words is a good transition OR better period.

I do think it bears noting however that all the furious anki-ing needs to slow down at some point... like anything else in the world it has a decreasing marginal utility.

anki-addiction is a disease, folks...

Not really able to watch the video on this Internet connection, so please bear with me: in respect to people discussing the recommendation to stop adding vocabulary at around the 6 thousand mark in a SRS program... would I be accurate in assuming that this is being discussed as to no longer add more cards yet continue to review the 6 thousand cards into their maturity, while focusing most of the time using mass native input for further vocabulary?

A little thing I'd like some clarity, in order to continue to follow the discussion.

Quote:anki-addiction is a disease, folks...

*laughs* I don't think anyone is going to disagree there.

yes by 'stopping anki' I and I presume the speaker were talking about not adding any more new cards... still continuing reviews, but those will become nominal after a few months.

dtcamero Wrote:[...]
so why 6k?...
Moreover, this number will probably vary depending on the language
This number will probably vary depending on the person

I think 6k is the point at which the speaker felt that the cost-benefit equation of anki-ing became relatively ineffective vs. high-volume reading, due to a number of factors...

Agreed.

More background: one slide mentioned that word frequency past 6000 words is very low. The argument is that past 6k, we're able to learn words more effectively through reading. He cites it as "contextual learning". Through extensive reading, you can then eliminate the "deleterious superfluity of low frequency vocabulary".

To quote alenaroo's summary

Quote:[...]Repeated encounters in different context deepens and enriches the meaning. Polyglots use known congnates from other languages and contextual clues.

From what I can tell he doesn't really mention quitting or ending anki use. I assume 4-6k words [from frequency] is an average for many people using anki

@dtcamaro -- your pros and cons seem accurate, and the combination of "anki con 1" and "reading pro 1" was the very reason I decided to quit Anki after 6250 self-input cards. (I did, however, enjoy the process, but I became addicted, and added pretty much any new word I came over...)

Also, the 6000 mark seems reasonable. I remember reading somewhere that there are about 4000 words that are very frequently used in everyday conversation, and also occur in movies, TV shows, books etc. Also, learning some words related to your interests may be helpful, because they are useful to you, personally, and describing your interests is probably something that you would like to be able to do.

Question: since Core is based on newspaper frequency, wouldn't it be stuffed with stiff, formal kango words that aren't used as much in everyday conversation? I find newspaper to contain way more kango than informal texts, movies and any other form of (less formal) media.

Deleting Anki felt like getting the disease cured. :p And it freed some time (and energy) to focus on an even weaker skill; listening comprehension

Also, the sentence method didn't work that well with German, but my German comprehension is usually closer to 99% in most cases.

However, once I start learning another language, I'll re-install Anki at once, because it is a bloody fantastic piece of software.

JimmySeal Wrote:I’m willing to bet that you learned more than 90% of your native language through “guessing” and continue to accumulate new vocabulary through “guessing”

...

It is in fact over-reliance on dictionaries and hard-and-fast definitions that leave people steeped in mediocrity, not the other way around.

I suppose I'm an exception, since I've used dictionaries extensively since elementary school. Of course, I usually try to guess the meaning of an unknown word on the basis of its context, but I also almost always double check my guess in the dictionary. Although my guesses are usually correct, I'm sometimes wrong too.

Although I realize this may sound boastful, this practice has given me a much more nuanced and refined understanding of English than most other native speakers. Native speakers constantly misuse words, phrases, and idioms. They often do so because they've heard such terms incorrectly used elsewhere, and thus incorrectly guessed the meaning of those terms.

In addition, modern electronic dictionaries seem to offer just what you claim is necessary for successful language acquisition. They can be searched instantly and are packed with natural example sentences so you can see the various uses of any word. Doesn't this allow learners to "see words in multitudes of situations and contexts"?

Anyway, I've worked as a professional writer and proofreader, and I have many friends in the publishing industry. All of them use dictionaries and thesauruses daily, and all of them are much better at using English than any hapless fool who relies on guesses.

vileru Wrote:[...]

vileru Wrote:They can be searched instantly and are packed with natural example sentences so you can see the various uses of any word. Doesn't this allow learners to "see words in multitudes of situations and contexts"?

Do they provide 1000 example sentences for each word, or somewhere around 3-5? By "multitudes", I was thinking somewhere on the order of several hundred.