Polyglot Conference Video on PolyNots
Yes, I go through cycles. Exposure, grammar, exposure, grammar, make fool of self, exposure, grammar, etc.
Steve as a Jedi knight!?
Does that mean that Friedemann is Count Dooku? Or Darth Sidious perhaps?
(I can think of a pretty obvious candidate for Darth Maul too, but he's going through a rough patch right now so I'll say no more...)
Steve & others....I couldn't resist...check out Steve's "new" profile pic! Haha!
You guys must be talking about Ben Kanobe.
I had to google his name to find the correct spelling!
You may judge how out of touch I am with modern culture by the fact that I had to look up Obi Wan Kanobe on google to find out who he was.
I look forward to our conversations in Japanese and Chinese. Meanwhile I enjoy reading about the different paths that people take towards their language learning goals. I see LingQ as a community of people with common interests, and a constantly improving source of resources for language learning.
@Steve -Oh, you are so right. I joked with my husband the other day that I wanted to call you Obi Wan Kanobe (except you’re *way* better looking). Haha! I’m using Assimil for both Japanese and Chinese at the moment, and it’s incredible what I’m noticing. Plus I can read the Japanese script fluently anyhow. I can do the hell I like on LingQ as you said. I don’t need anyone to come along and start scrutinising stats to fit CEFR levels et cetera. I plan on sitting JLPT & HSK exams, but won’t get hung up on LingQ stat correlations either. Your ‘Linguist on Language’ posts are particularly interesting of late.
In my case, I have accumulated years of passive experience in Japanese, commencing in 1975 (yeh, during the Vietnam War, wow). Then I studied it on and off over the years. I was reading & writing Japanese before all my native Japanese Uni lecturers were even born, which feels strange. I've stopped feeling regret for time wasted, and now focus on all the wonderful language study I'll be doing in the next 51 years! That will take me to age 102...
By the way, I am *so* looking forward to speaking to you on Skype in both Japanese and Chinese some day! Noooo, not yet! :)
I think the reason for this is that when we are first confronted with these beginner programs, however much they try to gently introduce us to the language, the experience is overwhelming. Everything is new and strange. We don't know what to focus on or what to grab onto. We think we will never learn this language.
When we go back to these beginner materials at a later stage, after we have had experience with the language, we now know what to look for. It's as if we were able to take our life experience as a 50 or 60-year-old and go back and become a 20-year-old again. We would do a better job.
And so it is with language learning. As experienced learners, as people who have already been confronted with so many different aspects of the language, we now know what we want to focus on. We don't have to worry about acquiring unknown words or wondering what things mean. We just zoom in on those aspects of structure or grammar which we know cause us trouble.
It is interesting that often those things that are directed at beginners such as easy graded readers, or grammar or Michel Thomas or Pimsleur, are actually most effectively studied after we have had enough exposure with the language
I found Anthony’s presentation highly entertaining, and so very inspiring. I didn't need to have to agree with everything said in order to benefit from it. I didn’t go along with the ‘short term memory’ thing for example, but who cares? I would personally love to see more presentations done by Anthony.
Interestingly, the passive/active vocabulary discussions have made me realise that I’ve been unconsciously recording words as known for LingQ purposes only when in fact they were really my active vocabulary. [not up to date yet either]. I suspect several members are doing this ie. not marking words known until they think they have learned them – eg. in my case, being able to comprehend and use them correctly in tests and written work, and also orally. Maybe even handwrite them. However, I do agree with Steve that it’s not necessary to know words actively for LingQ vocabulary tally purposes.
I took a break from my University language studies to turn my passive vocabulary into active vocabulary before I advanced any further. Except that I originally didn’t couch it in those terms in my mind. All I know is that I was frustrated from acing exams, but couldn’t say much, even though I’m really B2 in Japanese say.
Now I find going back over the nuts & bolts, and using such courses as Michel Thomas and Pimsleur very good for acquiring active vocabulary. They won’t necessarily give me a large active vocabulary, but they certainly are giving me more confidence to speak the language and use what I already know. We have to find what works for ourselves.
I loved the presentation and was lucky to be present in Budapest. Lots of things to think about and many useful ideas. Anthony is a hilarious presenter!
The presentation is being transcribed and the subtitles will be translated to other languages. Hopefully it'll appear here on lingq as a lesson if the license permits...
Not so simple. I have never seen estimated levels of active vocabulary required for the CEFR levels. Not simple to estimate active vocabulary. You would have to either record people speaking or takewhat people write, and analyze it. The sample would have to be quite large. It could be done but I don't see us doing it here at LIngQ.
The passive vocabulary levels cannot predict the active vocabulary levels, in my view. Passive vocabulary will become activated through lots of speaking and writing. There is no obvious correlation.
Paul Nation estimated that the ratio between word families and words in English was 1 to 1.6, I believe. One would have to look at his work to see how he defines word families.
So I prefer to stay with the passive vocabulary, or our known words count, because it is easy to measure. It is a valid indicator of one's language potential. The rest is up to each learner.
I agree regarding CEFR. The best way to get a reasonable correlation is testing, imo. But there is an easier/less accurate way. Use "active" vocabulary levels established for CEFR (assuming these exist). Compare them to active vocabulary levels predicted by the LingQ word number. The hard part is estimating the word families type of vocabulary levels given the LingQ word number. This can probably be estimated by a simple ratio. And you could probably assume active vocabulary is 50% of passive vocabulary.
Here is Krashen's view on the relative unimportance of grammar instruction instruction with some references to research.
Success in language learning depends on the attitude of the learner, the time spent with the language and the attentiveness the learner. Therefore, it is difficult to compare the effectiveness of different techniques, since we usually don't know enough about the different learners' attitudes, time spent, and degree of effectiveness. So I think it is best to focus on doing those tasks which we most enjoy doing. This keeps us motivated. This ensures that we spend enough time. Probably this also heightens our attentiveness to the language.
I am closer to Krashen in terms of what I like to do. That is how we have designed LingQ. I do review grammar, but I don't do exercises. If I enjoyed doing exercises, I would do them. I don't think my language learning suffers because I don't grammar exercises.
The important thing about input-based learning, is that it creates the potential for us to become good speakers. It builds up our familiarity with the language, or ability to understand, and our vocabulary. Ultimately, however, in order to speak well we have to speak a lot.
The known word count at LingQ is relevant to the CEFR levels. For each level there is certainly a minimum vocabulary level required. I refer to passive vocabulary. These passive vocabularies are not enough to attain the CEFR levels, they just make it possible. As to how many known words would be needed for each level in the CEFR, we would have to do a study it to determine that.
@creimann - agreed. I have seen videos and listened to people who have focused on listening and reading, and claim to do very little grammar study. Compared to a saturation (balanced) learner, their vocabulary is a little more extensive, but their grammar decidedly worse. They also tend to struggle a little more, and appear less smooth, which I attribute to less time spent on conversation...probably nothing to do with grammar. Basically, if you want to have good grammar, study and practice grammar at some point. If you don't care, then ignore this advice.
Once again about the lingQ word count. Yes, you can use it to tell that you have made progress. For example, if it says 10,000, you probably know twice as many words as when it said 5,000. And you can use it to help you choose appropriate articles to read and listen too. But you can't use it to tell you how close to your goal you are, because you can't tie it to the real world, CEFR level, etc. This is why I suggest testing users.
Studying grammar, which I do, is not the same as practicing grammar or doing exercises, which I don't do. But in the end it is up to each person to decide what they enjoy doing and what works for them.
I think there's a trap either way. One can pin one's hopes on immersion, or on grammar books. But some combination is needed. The genius of LingQ is not that it obviates the need for studying grammar, but that it brings some order to the chaos of immersion.
creimann,in my view, language is not like science. It is not something you learn academically, with structure and experiments. It is not a matter of understanding concepts. It is a matter of acquiring habits, something you just get used to through exposure and eventually lots of usage. Most learners of English who get the third person singular present tense wrong when they speak, understand the concept.
I also read grammar books, often, but not in the hope of retaining anything, but in the hope that it will help me notice the patterns of the language what I'm reading and listening.
Your approach is different, understood. However, I don't think you can see that others "have to" do what you like to do.
I think Colins German is amazing in this short time!
I have always had the same experience. People who learn intensively with LingQ speak after a short time in a very natural way. But those who study hard, write for hours or cram grammar exercises, do not speak after years in a natural way.
"2. Who ever said anything about discarding grammar entirely?"
I certainly didn't. When I started learning German, I did four months at a language school intensively learning nothing but grammar. At the end of that time, I could have debated Noam Chomsky on technical linguistics issues and won. I still couldn't speak much and I couldn't understand a thing anybody said to me, but I found this knowledge of grammar very useful anyway. I don't know if it was the most efficient way to learn the language though.
Oh okay, not Colin. Yes I think a conscious effort must be made to activate. Out-go.
Your point 2 reminds me of Steve's video, about knowing something before you learn it. Which he got from some Eastern sages. So your massive exposure to French gives you ample experience upon which to build the grammar. It's a great combination.
1. I remember the post you are referring to. It was not Colin, but another user I will not name who somehow has around 100,000 known words, yet an inability to speak. It turned out she had never really tried to activate her vocabulary.
2. Who ever said anything about discarding structure/grammar entirely? I learned using lingq, along with the occasional glance at a grammar book. I just took an advanced French grammar class for foreigners (my first ever class) here in France, and somehow outgrammered most of the other students there, including several French majors from good British Unis who were in the midst of doing their year abroad.
Anyway, if you ask me, which nobody did (lol), LingQ is a great system for encouraging and managing "massive input" (as Steve puts it). But I see no reason to abandon structure. Take the comparison: When people seek to become doctors, they aren't enjoined to 'just absorb' the structure of anatomy and physiology. No. They must sit their tuckus in the chair until they know the seven layers of this, and the arrangement of that, and this is inside of something else, and the hormones cause this, and the electrolytes are needed for that, etc etc. If they can't hack it, they must go and settle for being psychiatrists, where they can just make it up as they go along! lol Anyway structure is beneficial and real, so why not make the attempt?
"I don't want to make Colin feel bad or anything, so I hope I'm not speaking out of turn, but I recall a thread in which he mentioned having tremendous trouble expressing anything even after some years of LingQ work. I'm paraphrasing of course, but my gut reaction is, he's not doing enough structure and output work. I would look to that."
You must have me mixed up with somebody else. I only started using LingQ back in December. I only started learning German a year ago and I have very little trouble expressing myself on most topics that I have interest in.
I am told that I make few grammatical mistakes when I speak German. When I speak, I often have trouble expressing myself if the topic is unfamiliar to me, but this is usually due to a lack of experience talking about this subject and a lack of vocabulary and has nothing to do with a lack of grammar.
Another way to be assured that the word count on LingQ is a meaningful metric, is to observe how it correlates with languages you know. My Spanish is far superior to my French, and it was no trouble at all to race upwards in the Spanish word count.
@Steve, I've watched a hundred or more of your videos, and I know that you don't believe in practicing grammar. In some of your videos you suggest using grammar books as ways to expose yourself to patterns: compare the exercise with the answer key, etc. I believe that actual practice is essential, but, I don't know of any rhetorical way to resolve the matter. I just don't think reading and talking are enough. Without some structural work, your talking will just be rubbish, IMO. I know mine is, in German!!! lol Anyway, people are free to emphasize different aspects. That's why this is better than school. In school, it's One Size Fits All! But yes, I think you have to get in there and work the patterns, and the exceptions. And there's no reason why it shouldn't be enjoyable. You get the pleasure of putting new patterns into your brain. No need to fear work! :)
I don't want to make Colin feel bad or anything, so I hope I'm not speaking out of turn, but I recall a thread in which he mentioned having tremendous trouble expressing anything even after some years of LingQ work. I'm paraphrasing of course, but my gut reaction is, he's not doing enough structure and output work. I would look to that.
First of all, I don't believe that you have to practice grammar or practice pronunciation, unless you enjoy doing so. You will get enough practice naturally while you speak. Speaking with our tutors here, where you get a report containing the words and phrases that gave you difficulty, is an excellent way to start speaking. To speak well you need to speak a lot. The known words count is an excellent measure of your potential in the language.
Assuming that you use LingQ properly, your known words count should reflect the degree to which you have listened to and read in your target language, and therefore your vocabulary and familiarity with the language. These are the factors that will determine your ability to have meaningful conversations in the language, in other words to practice the structure and the pronunciation of the language.
For me, it is also a meaningful measure of my activity and progress in a language. I just don't think it is a meaningful measure of my known words.
I think a lot of the disagreement stems from not knowing what the interlocutor really knows of a language, and from the fact that just reading a lot isn't enough. You have to practice grammar, practice pronunciation, and put yourself into situations where talking occurs. So for example, if someone who has merely read and listened a lot is commenting on whether or not the word count is meaningful, that isn't the same as hearing the opinion of someone who has read and listened a lot, but who has also practiced gramar and pronunciation, and tried to talk a lot.
For managing a massive input environment, word count as used at LingQ is a perfect innovation, that serves as a meaningful measure of progress.
I consider the "known words" count to be an important statistic. I explained how it is arrived at. If you delete non-words, the count is accurate, based on our definition of what a word is. Beyond that, it is up to each learner to decide how to use this statistic. For me, it is a meaningful measure of my activity and progress in a language.
I don't generally think the known words count means much. I think of it more as a motivational tool. For that reason, I suggest making it more motivational by having it give crazy astronomically high numbers. Each new word I set as known should count as 50 known words on the stats. Each day that I do some work on LingQ, I should get as a present 5000 new known words. This way, after starting a new language, I will be on 50.000 known words after a few days. I can't think of anything more motivational.
LingQ vocabulary count is very different from what I believe most linguists use. But I don't really care. I've gotten used to the tools, and know that I prefer articles that have 10 or less new lingQ words per minute. I don't go around saying I know 10,000 russian words, because that's about as useful as saying I'm fluent. It could mean anything.
That being said, it would be interesting to test people, whose word levels are "up to date" in lingQ (per Julz), to determine their CEFR in reading and listening. I don't believe the 6 levels of the avatar stats correspond to the CEFR levels. It would be nice to test more than that, but I don't think any meaningful conclusions could be drawn from graphing lingQ words vs speaking/writing CEFR. There is too much going on outside of lingQ to make that meaningful imo.
Without testing, I don't see the point of doing statistics. But I'm interested in hearing what other people think they're going to learn from them.
One of the most fascinating things about this topic is, we don't know what a word is, yet we use them all the time! :)
With regard to schoolgirl dictionaries, I can only say touché!!
With regard to your passive word count, let me know when you get to 50,000.
I will start a separate thread on this subject.
On the subject of vocabulary, I checked out the concise Oxford dictionary of the English language. It apparently has 240,000 words. The full 20 volume Oxford dictionary probably has many more words.
If I open my concise Oxford dictionary, vintage 1952, to any page, I know most of the words. Some of these words are foreign words, references to Greek mythology etc.. If I only count one third of these words as words that I know, then my vocabulary is over 80,000 words. I suggest others try the same thing with other dictionaries.
I have the Lexin Swedish English dictionary. Right on the cover it says that it contains 28,500 words. If I leaf through it, I know a lot of the words, I don't know what percentage. I have a Czech-English dictionary published by SPN for schools. It doesn't say how many words there are, but a rough estimate would place it at about 15- 20,000 words. This does not include the inflected forms of words. My rough guess is that I know about half of these words, but I am not sure.
I think the average schoolboy dictionary contains 15 to 20,000 words. I think an educated person knows 50,000 words, passively, in their own language. But this all depends on how we count words.
If anyone is interested, I just went through LingQing this article. It took about seven minutes. I added 35 new saved LingQs. My known words total increased by 143. It turned out that a lot of the blue words were in fact variations of words involving the component "velo", "velociped", "velosharing" , "velomarcheroute" etc. and corresponding inflections. This will no doubt unfairly inflate my "known words" count.
On the other hand I deleted any city names, proper names, or words for which I did not get a user hint or Google translate definition. (In Russian, unless a word is of particular interest to me, if there is no user Hint or Google translate definition, I just pass. I rarely look up the dictionary. I am more likely to use a dictionary in Czech, because there are fewer reliable user hints in that language.)
I think it would be great - and I assume for everybody here at LingQ - if you could give the people here some more insight about your daily language learning routine: how much time do you spend with desktop LingQ, or mobile LingQ, how much listening while doing other things, how many new texts with what ratio unknown/known words, how many new LingQs per day, how much time reviewing old lessons, reviewing vocab, etc, etc.
Maybe we can learn something new....
So to summarize, the accuracy of the total "known word" count number is not important. The fact that the "known word" count in some languages will appear inflated compared to other languages means that we cannot compare from one language to another. Furthermore, our ability to use words actively is not necessarily an indication of the size of our passive vocabulary.
The gist of this argument or disagreement should be on how important a large vocabulary is. Anthony said in the beginning of his presentation that vocabulary was the key to understanding any text, more than grammar or familiarity with the subject. Yet later on he seemed to imply that the goal of language learning was not just to acquire words, that one could be fluent with 400 or even 10 words.
So, I simply want to make the point that, to me, acquiring vocabulary is the most important activity in language learning. It is also, to some extent, measurable, even though the measurement is not absolute nor unambiguous. If we acquire this vocabulary through massive reading and listening, we will also acquire familiarity with the language and the ability to understand well.
As to the accuracy of our "known word" count at LingQ, to some extent it doesn't matter. It's enough to know that the more words we know, the further we have progressed in the language, the closer we are getting to our goal of fluency.
Fluency also implies the ability to use the language. Therefore as our word count grows, we should be engaging more and more with native speakers and tutors, and using the language. The languages that I speak the best, are the ones that I have spoken the most. If I want to improve my level in a language, I sign up for lots of discussions in that language with our tutors.
With regard to our personal statistics here at LingQ, it is not possible to know with any accuracy what has been included in the "known words" count, in other words, how many numbers, names, non target language words etc.. It is also not possible to know exactly to what extent different forms of the same word have inflated this number.
However it is possible to look at our LingQs. We can see them in the vocabulary page. We can choose to view the vocabulary page 200 terms at a time. We can choose to look at "phrases only", and count the number of pages of phrases we have saved. We therefore can deduct this from our total LingQs to arrive at the number of words we have saved. We can then look at our LingQs and see how many of them we now know. This is also not a bad way of reviewing our saved LingQs.
I did this for Russian just now. I did a quick sampling. I know perhaps half, or slightly more, of my saved words, it seems. On the other hand, I seem to know the vast majority of the yellow LingQs that show up in my lesson texts. This suggests to me that the more common LingQs are the ones that show up regularly in my texts, and the words that I don't know in my vocabulary list may not be that important. I have relatively few learned words (less than 10% of my LingQs) since I am not an active user of flashcards.
Just out of interest, I imported an interview from Echo Moskvi into LingQ about Moscow as a bike friendly city. I just shared this lesson. The URL is http://www.lingq.com/learn/ru/workdesk/item/5901600/reader/.
Before I start studying this lesson, I can see the following statistics. Total running words 3524, new words 143, yellow saved links 121 (11.6%), known words 984. This share of new words is higher than normal for me at this stage in my Russian. I suppose this is because the text contains a higher number than usual of names, or words like "bikesharing" and "carsharing" in both Russian and English.
I will now quickly go through the blue new words. Any words that are not genuine Russian words I will delete with the X key. Words that are genuinely new to me, or words, that I have any doubt about amongst the blue words, I will save by hitting enter on my keyboard. This takes about 10 minutes. Then I will try to find time to read the text either on the computer or on my iPad. The sound file has been downloaded to iTunes and from there I will download it to my MP3 player for listening in my car, while doing the dishes, or while exercising. I will do a few of these interviews a day. I will read one or two of them, and typically listen to more.
Many of the blue new words will in fact be known to me, as forms of words I already know. Some will be non-words. I don't know how many genuine new words will be added. If I look at my profile, I see that I have added roughly 1100 new words to my known words count in the last two weeks. In the same period I have created roughly 1100 LingQs. The statistics say that I have read 65,000 words in this period. However, this is not true, as many of these interviews are on my iPad and I have not yet read them. Very often I save the new words (create LingQs) and then listen to the interview without reading the lesson. It depends how much time I have.
Here is what I wrote on a recent blog post.
I know that you feel that vocabulary is the key to language learning. But how many words do I need to know?
It is difficult to answer this. It is a bit like saying how long is a length of string. It depends on your goals. If you just want to be able to say a few things when visiting a country, you probably don't need many words. If you want to understand what people are saying to you, what's on the radio, and what's in the newspaper, you need a lot of words.
Yes, but could you give me a number?
The problem is that it is difficult to define what we mean by knowing a word. If I can recognize the meaning of a word in even one context, I considered that I know that word. By that I mean that I have started on the path of getting to know the full scope of the meaning of that word, and eventually being able to use it. I know the word, but I'm really only at the beginning stage of mastering the word. Most of the words that I know in a foreign language, and many even in English, I know only partially. I do not know all the different ways in which these words can be used.
Another difficulty lies in how to count words. Do we count the words "run", "running", "runs" and "ran" as different words? What about "outrun", or "also-ran", or "runner"? It is not clear whether we should count only word families, or each form of a word, as a different word. It is also not clear just what to include in a word family. So I prefer not to put too much emphasis on a specific number. Instead I just say we need to learn a lot of words. The more words we know, the better our potential ability to understand, and even to speak.
But you measure known words at LingQ. What does that number mean?
Our "known word" count at LingQ is an accurate measure of your level of activity and your progress in the language. The more words you know, the further along you are in the language. You will find that as your "known word" count increases, your ability to read and understand improves. In my experience, lots of reading, lots of listening and creating lots of LingQs, is an effective way of learning a language.
The best part about learning foreign languages is debating how many words you know.
For some reason, the debate is always carried on in English. ;)
Just teasing a little. Pretend I'm not here. :)
@Friedemann: "...I don't think one would refer to shooting someone as a collective act of a crowd. The sentence is grammatically correct though..."
You're absolutely right - shooting would not normally be considered a collective action. (However there may be contexts where it is possible.)
Yeah they refused it. :(
I guess those who made the test are anti-gun activists! It's been a long time since the last pro vs anti-gun discussion by the way. :P
EDIT: yes I knew "shot" wouldn't work but still, I tried.
"The crowd shot the prisoner..."
In fact the test is rather unambiguous. I don't think one would refer to shooting someone as a collective act of a crowd. The sentence is grammatically correct though.
To be honest, Friedemann, I'm not convinced that these tests mean very much. It says that I only know 88% of words known by university graduates. This is, frankly, bulls***! I am a graduate with first class hons, and there is certainly no university level text in English that I would have the slightest difficulty with.
Yes, "shot" is possible (but pretty unlikely!) :-D
The crowd shot the prisoner...
This is an interesting case: "yarn" was a new word for me, "to shove" and "to bellow" were kind of passive passive vocabulary. If these are in the 5000 vocabulary size range I can only imagine what a person has to know with a vocab north of 20,000 words. This again brings me full circle back to the vocabulary size as calculated by Lingq and Steve's (sorry for bringing it up again) vocab count for his recent languages.
ball of yarn
shoved the prisoner
Yeah, I also misspelled words but still I was far from a full score. Maybe you can help me out since you are a native speaker: Which word are missing here (from the test I took):
- The kitten is playing with a ball of ya...
- The angry crowd sho... the prisoner as he was leaving the court.
- We could hear the sergeant bel... commands to the troops.
From this post, I can tell that I know at least fourteen words in English.
Hmm...this is supposed to be my native language, but I only scored 88% at the 10000/university test! :-0
The score would have been higher if I hadn't screwed up the spelling of one word, but still...
English Test A / 3000-5000 level => SCORE IS 77%, need to work at this level. OK for me.
Just performed the English test. The result was that I am somewhere in the 3000 - 5000 region.
Looking forward to it!
@Friedemann The conference organisers have assured people that the remaining presentations will be uploaded in the coming weeks. Remember, there were many hours of conference material, and the videos are being edited by one kind volunteer during his limited free time.
Here are some samples of vocabulary size tests for English, based on work by ISP Nation (the author of the big fat boring book). The same webpage has frequency lists for vocabulary in French and English. http://www.er.uqam.ca/nobel/r21270/levels/
I was hoping that they would also upload the other presentations, no idea why it is taking such a long time.
@Friedemann - the Chinese LingQ stats are just that - stats for work done on LingQ texts - so if someone hasn't read much Chinese in the last 10 years, let alone on this site, obviously the stats are not going to show anything like the amount learned from prior years.
If I spend more time reading LingQ Chinese or whatever, then those stats will always be higher than my Japanese say, even though my Japanese is my strongest L2. If your Chinese LingQ stats reflect everything you know, good for you. Mine certainly don't. I'm not in any hurry either to get my LingQ stats to match my current ability.
What am I missing here?
[didn't update quick enough to see Steve's post just then]
With regard to the word count in Chinese, the issue is not inflections. But rather how the characters combine to form words in the sense that we understand words. I don't know how this is dealt when it comes to word counts in Chinese, I just don't know.
However that is not the important point. The important point is that my active vocabulary in Chinese is not relevant to how many passive words I have acquired using LingQ in the languages I have studied most recently.
you're right, I do not know the number of Chinese words you know, I simply made the assumption that it should be "qualitatively" at least equivalent to your Russian and Czech since you often refer to Chinese, Japanese and French as your first tier languages. I think words are really easy to count in Chinese since there are no flections.
I believe, based on my wordlists that my passive vocabulary in Chinese is well below 20,000 words and that is after 5 years studying and living the language.
Again, I hope that Anthony will weigh in on this here. In the presentation he said that more recent studies suggest that previous estimates of word counts were an overestimation. I think he said 20,000 words (in English?) by age 20 and then further accumulation will be much slower so I don't think we continue at a rate of 1000/year until we are into our 70ies.
Ah, I forgot to say it is a very interesting and informational presentation regarding both the content and the Takahashi method. By the way, I also ordered the book because of your presentation ;-)
Thanks for posting the video, Anthony - I'll watch it later tonight!
As for word count, size of vocabulary, passive vs. active vocabulary etc., HTLAL member Iversen found out that he didn't use that large a portion of his passive vocabulary. (~10%, maybe 2500-3000 words)
No, I wouldn't judge Steve's Chinese based on the fact that he used a certain expression 10 times in a video.
@AnthonyLauder - We are working on updating the API to include more functions. Send one of us an email with some of the things you had in mind and we'll see what we can do for you here.
No, it wasn't really a hatchet job, but you did say to Steve: "...you do not strike me as someone with a very rich vocabulary. For instance you must have used 经验 some ten times in your video to mean "having experienced something". First of all 经验 is not often used as a verb but rather as a noun, secondly someone with a rich active vocabulary would use words like 经历，体会，体验，to provide some variation and to employ the subtle differences in meaning of these verbs..."
(I guess I'm just saying this seems a little über-critical, considering that Steve hasn't actually lived in China...)
Anthony, I am intrigued. What are some of the features and functions that you would like to create for LingQ to make it a more sophisticated tool. We have lots of ideas and limited resources. From our experience, everything takes much longer than we expect.
If you can help us improve LingQ, we are very interested in listening.
Here we go again,replying to your specific points.
1. I don't remember the numbers that Anthony gave in his presentation. However, searching on the Internet shows levels of between 10 and 60,000 words (Pinker) as the average English person's passive vocabulary. Anthony did say that we increase our vocabulary by 1000 words a year. If this were true, I would have 67,000 words in English. Of course this assumes we are talking about passive vocabulary. It also sidesteps the issue of whether we are referring to word families or each individual form of words.
In any case, the important point is that learning words in a second or foreign language is not the same as adding words in your native language, for the reasons I already provided. If you disagree with this concept please say so. Don't just ignore it.
2. I don't really care about the accuracy of my word count in LingQ. It is a measure of the increasing size of my vocabulary, which enables me to read more and more difficult texts. If you believe that I am not telling the truth about the range of my increasing reading ability, please say so.
3. I don't know why you would make assumptions about my Chinese word count. I don't know how you count words in Chinese and how that relates to English or Russian word count. Do you?
4. Why would you assume that my active vocabulary in Chinese, a language I have used little over the last 40 years, should be particularly rich? Why is my active vocabulary in Chinese relevant to the acquisition of passive vocabulary in other languages as reported on LingQ?
5. What do you mean by true word count?
would you mind to elaborate a bit more on the methods to measure vocabulary size you were referring to in your presentation? Why is it such a complicated quantity to measure?
With regard to doing research on LingQ, I will be attending the multilingualism conference in Montréal in October. See you the URL below. I will hope to interest some academics and doing research on LingQ.
hatched job? Here is what I said: "...you certainly have very good flow, above average pronounciation but you do not strike me as someone with a very rich vocabulary."
I am not sure what exactly it is you don't understand but I'll try again:
1. In his presentation Anthony gave numbers for the amount of vocabulary typically used by speakers of a given language.
2. These numbers were way below what you usually report for yourself.
3. I assumed your Chinese word count to be in the same range as for your other languages.
4. Assuming that amount of known words I would expect richer vocabulary in your spoken Chinese.
5. This suggests to me you or Lingq might overestimate your true word count.
This idea of doing some kind of statistical analysis of the LingQ database is one that has been mentioned a few times. At one point, I even became excited about doing it myself. I could build a much more sophisticated tool than LingQ currently is. The problem is that the API that is made available is very limited, and does not allow access to very much at all - in fact, it is does not enable much more than the limited functionality currently in the iPad app for LingQ. This is a great shame, since full access to the LingQ database would enable some really great features to be developed.
Have you contacted any genuine academics who study language learning to see if they are interested in doing such research? I think there is a huge amount of data here that could be used for genuine research.
I for one would be interested in knowing how much you can learn about somebodies level in a language from whether or not they know the word 'shoelace', which the LingQ data could probably answer. I guess that's not really a serious piece of research.
I've gotta be honest, Friedemann, you are giving some slightly mixed messages here: first you do a hatchet job on Steve's Chinese, but then you say he is "really good"...??
Personally, I think Steve deserves extra credit for becoming fluent in this fantastically difficult language without ever having been immersed in a Chinese speaking environment. (I have real doubts whether I would be able to do what he does now - even if I had lived for a decade in China!)
Colin, we would love to have someone do research based on our learners here at LingQ.
Friedemann, the point is I do not understand your point. I have made it clear that my word count is a count of words that are passively known. You cannot say that I have not acquired that many words passively, in the different languages that I have studied at LingQ. You cannot say that you cannot do it, because you have not tried to do it. Different people will acquire words at LingQ at different rates. The rate will depend on many factors, including how much time they put in, how close languages are to ones they already know, and other personal considerations.
Combined reading and listening is a powerful way to acquire words, at first passive words, and eventually with enough speaking opportunity, active words. There is research to support this. LingQ makes this process particularly effective, in my view. There is no research to support this.
as I said before your Chinese is really good! My point is always the same: I have a hard time believing that one can build vocabulary as fast as you say you can. Furthermore the word count you typically report here for many of your other languages (which I assume to be comparable in your Chinese since it is one of your stronger languages) would produce different spoken output, I think.
Have you ever thought about using the huge amount of data available at LingQ for doing genuine research into language learning?
Well, sorry to intervene the discussion between Steve and Friedemann, especially about the word count, the use of LingQ, etc.
I would just like to focus on the usage of "experience"-group you may have doubts on. To be short, both Steve and Friedemann are right. As @u50623 described elaborately in his post, 经验 can be used as a verb, but in everyday conversation, this usage is relatively less applied. If you check the source which Steve quoted, the sentence comes from 艾未未(Ai Weiwei). If you know 艾未未, you know he is a writer, an artist, a critic, etc. in China. The usage of 经验 as a verb is, well, a relative written form.
Anthony, I was not speaking with reference to research studies. I was only speaking on the basis of my own experience. I am quite skeptical of much language learning research. A lot of it is contradictory. The context of most of this research is usually the classroom. I don't think there's any research out there which studies LingQ learners, who have access to our vocabulary acquisition functionality as well as the audio for every lesson, and who have the ability to choose to learn from whatever interests them. In other words I believe that most of this research is based on typical passive classroom learners.
Personally, I do not guess at the meaning of words when I am on LingQ. Any word or phrase that I am unsure about, I save and create a LingQ for. I do agree however that experienced polyglots, and I would include myself in that category, are quite willing to accept uncertainty and a lack of clarity. That is what helps them persevere in texts that they don't fully understand. But personally, I don't guess at the meaning, I just move on and leave the meaning unclear.
On the other hand, there were number of statements that you made which were more subjective. You suggested, for example, that we could be fluent with only 400 words and that this was somehow a useful target. This may be a valid strategy for some. I wanted to point out that it is not a strategy that I pursue.
I am not sure what points you're trying to make.
You suggest that 1000 words a year is a likely rate of vocabulary increase. This means that someone learning a foreign language would need 10 years to reach the vocabulary level of a 10 year old. Do you really believe that?
You challenge the credibility of my vocabulary acquisition rate. Yet you won't use LingQ the way I do. Try it. Try reading over 1 million words and creating over 30,000 LingQs. Try reading texts full of yellow highlighted words and phrases. Then try listening to these texts. And make sure the subject matter is of interest to you. You refuse to do this, and stick by your excel spreadsheet system of keeping track of new vocabulary, while at the same time challenging what I claim to achieve. You can't have it both ways.
Your criticisms of my Chinese may well be valid. When I type jingyanguo, however, these three characters come up on my Mac Chinese word processor, 经验过, When I Google 经验过 I find examples like the following of it used as a verb.
And of course, as Hape points out, the dictionary says that it can be used as a verb.
However, perhaps it is rarely used this way as you suggest. Perhaps this is an Anglicism on my part. That is probably not the only fault you can find with my Chinese. I will agree that my active vocabulary in Chinese today is limited. But I don't understand what point you're trying to make.
I have made it clear many times that if I can recognize the word in any context, I consider that word known. The more words I know in this way, the more I can understand of what I read and listen to. But it certainly doesn't mean that I can use these words. My focus in language learning is comprehension and massive input. The more words I know, the broader range of material I can read and listen to, and the more words I can learn. When I have an opportunity to speak a lot, my output skills quickly catch up.
I have never lived in a Mandarin speaking environment, where I would have the opportunity to speak Mandarin regularly. I had the luxury of studying full-time for 10 months, 45 years ago. I spent most of my time reading books with glossaries on literature, history, politics and economics. The first book that I read without a glossary was 骆驼祥子 by 老舍 after seven or eight months or so. To do so requires a fairly large vocabulary. I can't tell you how many words since LingQ was not available at that time.
Since that time, I have visited China a dozen times. I also have many Chinese books and CDs here that I use from time to time, to refresh or improve my Chinese. Over the last 10 years I have focused on learning other languages and have rarely spoken or read or listened to Chinese. I would love to get back to Chinese and if I were to do so I would use LingQ. This would enable me to increase both my passive and active vocabulary. But I fail to see the relevance of all of this to my rate of vocabulary acquisition at LingQ.
No matter what you find in dictionaries, I have never heard Chinese use 经验 as a verb. It really sounds 别扭.
"The speech may not sound "beautiful", but understandable."
There is absolutely nothing wrong with his Chinese, no issues with understanding it. Very good flow, natural melody and phrasing. It's just that his vocabulary does not strike me as particularly rich. I don't know what his word count is in Chinese but given that he knows 4000+ characters and based on his word counts in other languages I would think he should be in the 10.000 - 20.000 range for sure.
"In my opinion it is VERY difficult to "provide some variation and to employ the subtle differences in meaning of these verbs"..."
Well, knowing how to wield different but similar words may not be easy but it is part of becoming good in a language. I don't think that mastering synonyms is any more difficult in Chinese compared to other languages.
You are so right about 经验, 经历, 体会, 体验.
经验 n. experience ◆ v. go through; experience
经历 v. go through; undergo; experience ◆ n. experience
体会 v. know/learn from experience; realize ◆ n. knowledge; understanding
体验 v. learn through practice or experience
It's difficult, even for advanced learners!
What's more, there are regional differences in usage.
Most learners only learn one expression, and use it.
The other forms are only passive knowledge.
The speech may not sound "beautiful", but understandable.
In my opinion it is VERY difficult to "provide some variation and to employ the subtle differences in meaning of these verbs"...
Thanks you the entertaining presentation. I enjoyed the video very much.
Best wishes with your new post. I think you'll like Bayern - it's a great place. (And it's probably the most rightwing place in the whole of Germany too, so you may be able to escape Socialism! :-p)
Yes, I am back in Germany, will move to Munich in a couple of weeks for a new job. Language wise, still busy improving in the languages I already speak, especially Chinese. I have not started any new language project since I took on Chinese five years ago.
I agree with all you said in your response to Anthony, you made very good and valid points there.
Regarding the vocabulary issue, we have been over this a couple of times and the only thing I can say is that I feel that my own word acquisition rate is much slower than what your statistics suggest. I cannot judge your Russian, Czech nor your Romanian but when I listen to your latest Chinese video for instance you certainly have very good flow, above average pronounciation but you do not strike me as someone with a very rich vocabulary. For instance you must have used 经验 some ten times in your video to mean "having experienced something". First of all 经验 is not often used as a verb but rather as a noun, secondly someone with a rich active vocabulary would use words like 经历，体会，体验，to provide some variation and to employ the subtle differences in meaning of these verbs. Maybe it is all in your passive vocabulary, I honestly don't know, but from my own learning experience and your active usage of words the Lingq statistics are hard to believe for me (or maybe it is just my inferiority complex...).
The thing about 1000 words in a year definitely seems wrong to me. I can see it might apply to people learning in a traditional classroom context. But if you are exposed 24/7 to a language for 12 months, you will surely soak up several thousand words?
I'm absolutely certain that I learned way more than a thousand words during even my first 6 months in Germany.
(Of course, that's just a personal anecdote which therefore can't compete with fancy Clugstonian things like "research"... But still, it was my personal experience, and I doubt whether I'm so very unusual as a language learner?)
Remember, very little of what I presented was my own opinion. I was quoting research findings from several hundred studies from applied linguistics. As noted in the video, I preface the whole presentation with "allegedly" since research findings can be wrong. Having said that. I found some of the research findings pretty convincing - the importance of guessing, for example. At the same time, I recognise that, as a language dunce, I may be too easily persuaded.
One more thought on speaking. If we are lucky enough to live in an area where we have lots of opportunity to speak the language, such as in the country where the languages spoken, of course it is a good idea to speak early and often. However most language learners are not in that situation. That is why listening and reading (and LingQing) are such convenient and mobile methods of learning. They are very effective and build up our language potential for when we have the opportunity to speak more. Eventually, to speak well, we have to speak a lot.
In fact rather than organize my thoughts I will just present here those issues where I disagree or agree with you, more or less in the order that they were covered in your presentation.
I very much disagree with the polyglot and polynot dichotomy. I do not agree that short-term memory leads to better guessing by polyglots. I think that guessing and inferring and short-term memory are irrelevant. I think polyglots are simply more experienced language learners, people who have found their own ways of learning languages and who are more experienced at doing so. But more than anything else, polyglots are people who spend their time and a lot of time on useful learning activities.
I agree that vocabulary is the key to understanding a text in another language, more than grammar or familiarity with the context.
The fact that a native speaker accumulates new vocabulary at the rate of 1000 words a year doesn't mean that someone learning a foreign language can only learn 1000 words a year. In fact the foreign language learner is going to learn many more words per year. The adult native speaker is acquiring low-frequency words every year in his own language. The adult foreign-language speaker is acquiring high-frequency words in a foreign language, often cognates, and eventually medium frequency words in his or her area interest. Very often these are words that mean the same as words that the adult already has in his or her own language.
I agree that cramming is not a good idea, being hung up on one meaning or an authoritative dictionary meaning is not a good idea, worrying about how well you know or don't know what word is unproductive, learning vocabulary thematic groups is not useful, and being over concerned about learning slang is also not useful.
I think learning vocabulary is the key and fundamental activity in language learning. This is for the simple reason that in order to engage in meaningful output, in other words conversation with natives, we need to understand what they are saying. We cannot control which words they are going to use. The native speaker has a much larger active vocabulary than we do. Therefore we need a passive vocabulary that approximates their active vocabulary. I see no purpose in being fluent in 400 words. Whom are you going to have a conversation with if you only understand 400 words?
I agree that excessive flashcarding and Anki addiction is not an effective use of time.
I learn words in context. I learn words from reading things of interest. I don't worry about frequency or collocations. I simply want to read and listen to things of interest. I do very little intensive vocabulary study. Through my listening and reading, I will naturally get sufficiently intensive study of the most important words, and these words will be relevant to my areas of interest. This happens naturally without me having to make any special effort or needing to focus on lists of frequently used words.
I do not like graded readers other than in languages like Chinese where you simply don't have enough characters to read anything at the beginning. I have no interest in reading a simplified or dumbed down version of Moby Dick. I would rather read the original on LingQ.
Since I do a lot of my reading on LingQ, I do not need 95% vocabulary coverage in order to read material of interest. Since I can also listen to the same content, I can handle material with 50% known words. If I only read material with 95% known words, it would take me forever to accumulate the vocabulary that I feel I need.
Yes to acquire a large vocabulary you have to read a lot, probably over 1 million words. My statistics at LingQ tell me that I have done that in both Russian and Czech for example.
Personally I have not found it more difficult to learn words or remember words as I grow older.
Yes activation is always difficult. Personally I find it uninteresting to try to speak when I have very few words and simply don't understand what the native speakers are saying. I am quite convinced that speaking early is not a condition of becoming fluent. If the opportunity presents itself certainly one should speak. But these awkward conversations were I stumble to speak and don't understand what other people are saying are not something that I pursue in my language study.
I don't have the impression that traditional language instruction tries to keep the student quiet, as you suggest. Rather classroom teachers try to get the students to engage in role playing and talking to each other and other useless activities.
Since I like my language learning activities to be meaningful, I do not engage in talking to myself in front of a mirror. However if other people enjoy doing that I don't see any harm in that activity.
To me fluency means being able to speak on a number of subjects and understand what other people are saying. This implies a very large vocabulary. There are no shortcuts.
Hey, Steve. I just wanted to add another perspective. I used to teach English as a second language in Taiwan for years. And for the final four years, I was running my own school and had complete control over the curriculum and teaching methodologies.
There were two pieces of my curriculum that were fairly innovative and unusual for the Taiwan buxiban scene. One was a aural phonics training system for beginning level students and the other was a strong emphasis on extensive reading, using the Oxford Bookworms series mentioned in the presentation, in fact. My primary inspiration for this decision was Krashen's research as well as the "book flood" experiment Beniko Mason did in Japan.
The benefits I saw my students gain from the graded readers were dramatic. I realize it wasn't a well controlled experiment like those of the linguists mentioned above, but I saw clear improvements pretty much across the board—not only in skills but in terms of affect towards and interest in English. High quality graded readers give students a window into the target culture, exposure at an easy comfortable level, reinforcement of previously learned material and at least for most adults, more interesting material than children's books (which I also like learning from). Most importantly, when using graded readers, students are actually *reading* at a reasonable rate instead of picking through unfamiliar words and struggling to understand. In general, I saw students struggle with the first couple of level one readers (partially because of how early I gave it to them), but then read them pretty comfortably after that. After about 10 readers of a level, most students were not only ready but wanted to move up to a harder level.
And this brings us back to the question of what you can do with only 400 words. You can talk with a sympathetic native speaker, such as a teacher or conversation partner and you can read an Oxford bookworm such as The Coldest Place on Earth, Pocahontas, or The Monkey's Paw! Get hooked on that and you won't be stuck with only 400 words for long.
I'm interested in the idea of Lingq, and tried it at one point. I think the inclusion of high quality premium content such as extensive readers and children's books would be a huge asset. I remember asking you on youtube a few months ago for good Spanish reading resources for a beginner. Now I have a good answer to my question—Blaine Ray Novels, the Read It! graded readers, the Lola Lago detective series and maybe the Penguin Parallel Texts. Get that kind of content, start working on integrations so that users know when they're ready for a given level of text and Lingq will be much stronger for beginning students. I know this is a huge amount of work. Until very recently I was one of just two engineers working at Verbling and I have experienced on a deep level just how much work it takes to push new features.
Another, more philosophical problem is how to handle phrasal verbs and other multi-word semantic chunks. When does a learner actually "know" a word like go? I had to explicitly mention to my students as we were reading, that "go on" was 繼續, "go over" was 重溫 or maybe 受歡迎, while "go through with" was 實行, etc... These multi-word chunks are every bit as important as many medium frequency words and they make binary counting of words being known or not very difficult. Ideally, I think they should be treated as separate vocabulary items and included in any sort of collection of high-medium frequency words for learners who are trying to get to a basic level of competence in a language. Letting Lingq users see what items are remaining before they have covered the high or high+medium frequency words in a language would probably be a useful feature, too.
I intend these comments to be friendly and constructive to the discussion and am in no way complaining about what you've made so far. Thanks again for all the videos and content you've shared over the years!
Thanks for your comment. To me, interest in the subject matter is important to ensure that I stay with it. The only graded material that I enjoy is non-fiction, history and the like. If the content is literature I prefer the original. Thus I have only used graded material in Chinese, where simplified histories etc. were available, and I only used those at the very beginning, first few months or so. I couldn't wait to get to authentic material. I am not interested in children's stories, and don't necessarily find them easier.
Learners can save phrases at LingQ but we have no way of assigning them a frequency coefficient.I think the concern about learning high frequency words or phrases is overdone. If we save phrases at LingQ and they pop up again and again as yellow saved LingQs, we will get the reinforcement we need.
As for speaking to people when I have only 400 hundred words, my experience tells me that the conversation will be very limited and I won't understand much of what the other person says unless he/she really limits the subject matter. Just not something I like to do.
But then that is just me. Cheers.
Anthony you are a very enthusiastic and entertaining presenter. You won't be surprised to hear that I disagree with much of what you had to say. I sat through the presentation and took some notes. I will present these here once I have collected my thoughts.
In your paragraph one, just so I understand, are you saying that I am not telling the truth?
In your paragraph two, the 1000 words referred to by Anthony are probably word families. They refer to the acquisition of words in one's native-language through the course of one's life. This has nothing to do with learning another language. The rest of this paragraph is unintelligible to me.
Just to set the record straight, after a year of learning Czech I could read the newspaper and understand radio programs, not entirely but well over 85% for familiar subject. This would not be possible with the vocabulary of 1000 words. I achieved close to the same in Romania because so many of the words are similar to Italian and French, or Slavic languages.
You're no longer based in China (as I see from your recent post on another thread.)
So what's up now? You planning to learn Japanese or something?
finally I got to see your presentation. I was in Budapest but missed your presentation while I was there. Could you elaborate a bit more on those tests quantifying the number of known words? Why is it such a complicated science? I am particularly interested in that topic because it regularly comes up here at Lingq where some suggest a rate of voocabulary acquisition of some 400 new words per day for languages related to one's native language. Steve for example reported some 57.000 words in Russian after a couple of years of studying Russian or some 12.000 words (or was it even 20.000?) after some months of studying Czech. From my own learning experience I find those word counts hard to believe.
In your presentation you suggest a rate of some 1000 words per year as an accumulation rate which would sound about right to me. Can you say something more about the methodology of those tests? Wouldn't the most obvious approach be to expose people to content of some 50.000 word and just count the number of unknown or known words? How long would that take? As I understand it, this is how Lingq generates the estimate of known words based on the number of known words in the Lingq system.
Interesting talk - thanks for posting this, Anthony.
But, in a way, you are too relaxed and too nice. You need to shave and polish your head, wear a green muscle vest and sneer and snarl like our friend, Mr C. That way, it wouldn't matter matter whether you are a "polynot" - you would be a linguisssst! :-D