One can read with only 3000 characters? the half blood prince (6th book of harry potter) has 4200 characters in it. I know these will be mostly one time occurrences but that is 1200 unknown words which will definitely impede comprehension.
I just don’t like exemplars for progress to be unreal for example if barry bonds wasn’t caught for steroids that is now a benchmark people are compared to which is unfair. If i study chinese 5 hours a day for 1.5 years and don’t have the progress of will (which is so impressive im questioning the type he should take this as a compliment if it is true) but discouraging others is almost worse than finding the small improvements one would find with the interview. He seems to study pretty regularly (although a few minor details)
it’s like when i began studying spanish the first 6 months i don’t really count cause it was dabbling but it definitely counts like when he stated he dabbled with 3 languages and ended up with chinese cause he liked it he definitely didn’t count this time so its not really lying its just the criteria for what gets counted and not counted or what he can remember it sounds like he has had an eventful 1.5 years with a bunch of memorable chinese encounters.
Mattvjapan has an example of this he states i could read books after 2 years in japanese and easily knew 15,000-20,000 words which is true but he doesnt state where this 2 years was. It was after his japan trip so technically it was year 4 so it is very deceiving. I have the time stamp of the video to quote him if you don’t believe me.
just a P.S. I don’t care about the input vs output arguement that has been happening in this thread (interaction with the language= progress). A bunch of methods seem to work and the constant seems to be engage with the language.
PS. Great podcast it seems to be creating a lot of buzz and chinese is very interesting especially since it is so stereotyped/mysterious etc.
I arrived in Japan in 2002 and got the JLPT 1 in 2006. I didn`t use Anki or SRS, though I used an electronic dictionary to save words that I looked up during the day and revised them when I had time, maybe just once, since I knew I would have massive input anyway and was surrounded by “teachers” all the time, so the words would keep coming.
Will is great, I wish I could have that pool of Chinese friends. There are many Chinese here in Japan and I often find some kids in the schools I teach, though usually there isn’t much time to practice with them and though I have HSK 3 my speaking ability is quite poor and I don`t have any Chinese friends at the moment.
Everybody is so serious about the discussion that I just laugh at the girlfriend factor :-), massive immersion amped by super motivation, love!
How about adding emotional inteligence and the importance of that factor for more easily memorizing/learning things that you have a great interest?
I had a Hungarian girlfriend a long time ago at university , she gave me a dictionary and I learned a lot very quickly, to the point she asked me if I had memorized the dictionary 
However, after we broke up, the information was quickly deleted from my Hard Disk, lol.
“extract wisdom from his story which we all might be able to apply to our own language learning.” (Michilini)
Exactly, that’s the point!
For me personally, it’s about the enhanced version of the “(ultra)reading while listening” approach. That is, I would never rely for years only on “reading and / while listening” like Matt to learn an L2 without becoming more (pro)active.
Or to put it another way, a prolonged “silent period” of many months or even years seems completely “unnatural” to me as an adult learner (iMichilini is absolutely right on this point: adult learners shouldn’t behave like babies because that’s just BS).
@kimojima
“feel euphoric”
The problem with personal “breakthrough” stories is that
- there is too much ego, esp. the amalgamation of “competence” and “self-worth”, involved,
- there are too many biases (Dunning-Kruger bias, confirmation bias, etc.),
- there is often no tracking / there are no stats,
- there is too much hyberbole à la “reaching perfect fluency in days, weeks, months” at play here - even though the concept of “perfect” fluency is utterly absurd
etc.
And even worse, most learners don’t understand even the most basic things when it comes to acquiring practical skills (math, languages, programming, etc.), i.e., they don’t understand
- how “deliberate practice loops” work,
- what SMART goals are,
- that relying on volatile states of the mind (“feeling like it”, “fun”, etc.) compared to habit forming is utterly foolish,
- that running blindly forward and not knowing what you’re doing is usually a recipe for failure, or at least leads to wasting many months or years of precious life time without accomplishing much
etc.
Therefore, at least from a professional perspective (i.e. as a teacher, coach, course designer, app / platform developer, etc. ), we need more science and (tracking) data in this context not more hyberbolic breakthrough stories based on vagueness, mysteries, feelings, biases, etc.!
Or to put it differently, the language learning industry suffers from “overpromising and underdelivering” because everything must be super-super easy, super-super fun, super-super fast, yada yada, yada.
So, highly emotional breakthrough stories plus excessive marketese are not the solutions, but parts of the problem!
On the other hand, most learners will probably never study a science related to skill acquisition processes (psychology, SLA, etc.), but will simply be inspired by “personal success stories”. So there is a place for personal storytelling, too!
However, the role of critical thinking (whether based on science or not) isn’t simply to blindly believe such success stories, but to analyze, deconstruct, etc. them.
Just my 2 cents
“There is evidence from a number of case studies now that learning to speak fluently first helps immensley with learning to read later.”
I’d say it’s much more complex than you claim it is:
- First language or second language?
To quote literacy expert Timothy Shanahan (“Does Oral Language Instruction Improve Literacy?”, Nov. 28, 2017. URL:
Does Oral Language Instruction Improve Literacy? | Reading Rockets):
I don’t include oral language in my framework because there aren’t studies showing that direct and explicit teaching of oral language results in improved literacy. At least not with first-language learners. (It’s different with ELLs — teach them oral English and their English literacy improves. For those students I’m a big fan of Claude Goldenberg’s research findings that providing explicit oral English lessons to ELLs is beneficial.)
Without evidence that teaching oral language improves literacy, I’m unwilling to include it in my framework.
The reason I agonize over my choice is because so many respected experts include it in their theories of literacy development (at least for young children); see The Prevention of Reading Difficulties (1998) , for example. Also, there is extensive correlational evidence showing a close connection between young children’s oral language development and their later reading comprehension success (National Early Literacy Panel, 2008). That is, the kids who are best with language tend to learn to comprehend better than other kids (similar to the ELL finding that kids with the most proficient literacy gain the most from reading comprehension instruction).
But that doesn’t mean giving kids explicit language teaching would necessarily improve their reading significantly. The kids with the best language tend to have the most educated parents and other helpful resources. Perhaps, it is greater parenting support that matters rather than language that we’re picking up on.
What we need are studies showing that explicit teaching of oral language results in some kind of improvement in literacy. Unfortunately, such studies don’t yet exist." (highlighting by me)
- (Non-)native speaker parents and their kids
From my teaching experience, it is a common problem in Germany (and probably in other countries with high immigrant populations, too) that children of non-native German speaking parents are often fluent in German (at least if they grew up here), but have tremendous difficulties reading and writing in German.
This is also the case if the parents are native speakers of German, but aren’t very educated, don’t read and write much, etc.
*** Nota bene (for American readers of this comment):
“1. How many people in the U.S. are illiterate?
Approximately 32 million adults in America are considered to be illiterate; about 14% of the entire adult population cannot read.”
I guess most of them are native speakers of English, aren’t they?
“2. How does that compare to the rest of the world?
On a global scale, illiteracy affects 774 million adults aged 15 or older. Among developed nations, the U.S. ranks 16th for adult reading skills.”
URL: Shocking Facts: 23 Statistics on Illiteracy in America ***
My hypothesis is therefore:
The decisive factor for developing good or even excellent literacy skills isn’t “fluency first”, but rather the social milieu and the educational level of the parents →
- How much importance do they attach to education?
- How much do they read and write themselves?
- Are they good role models for their children?
- What are their economic resources?
etc.
- The gap between the spoken and written dimensions in the L1(s) and the L2(s):
For instance, in the Romance language family, the gap between speaking (including listening) and writing (including reading) is greater in (Brazilian/European) Portuguese and French than in Spanish.
Because of the many contractions and complicated pronunciation, one has to learn “two” languages in both Brazil and France (e.g., there is a big difference between spoken French “sche pa” and the written form “je ne sais pas”). This isn’t the case in Spanish, for example.
- Final thoughts:
Good writers are always good readers. That is, being fluent in a language isn’t good enough when it comes to literacy skills - neither in the L1(s) nor in the L2(s)!
This is wisdom shared by the professional American author Scott Sigler in one of his podcasts (but I don’t remember his exact words):
- You want to be a good writer? Read a lot and write a lot!
- You want to be a great writer? Read some more and write some more!
This is probably a credo shared by all pros, i.e. fiction writers, non-fiction writers, journalists, technical writers, etc.
@Hagowingchun
“Yeah I’m just not believing this story 1.5 years for all this? I think the time is getting stretched a bit or a lot.”
Time frames in “months/years/decades” are in general completely useless for assessing processes of practical skill acquisition (beyond language learning).
What is relevant in this context, however, is the number of hours required to achieve a high level of fluency in particular and proficiency in general in the L2, esp. based on deliberate practice.
In not so distant L2s (let’s say an English native speaker wants to learn Dutch, French, Spanish, etc.) achieving a B2-C1 / C1 speaking-listening and reading level in 1000-1500 hours is definitely possible.
And if learners spent 3-4 hours per day on language acquisition, they can achieve this within a year, esp. if they’re experienced language learners.
Achieving this in “group 5 languages” such as Mandarin, Cantonese, Japanese, Arabic, etc. demands much more quality time (at least 3000- 4000 hours - or more).
And with all the activities involved, i.e.
- making, learning, and reviewing Anki cards,
- “input” acitivities
- a lot of speaking / corrections
- shadowing
- learning several thousands of Chinese characters
- and extensive reading
Will’s numbers simply don’t add up! - He would still be extremely fast to achieve a high level of proficiency in speaking, listening, and reading in ca. 2500 hours! My guess is, this takes at least between 2500 - 3000 hours. And that’s ca. 5-6 hours a day for 1.5 years = ca. 547 days.
- In contrast, achieving this feat in less than 1000 hours (i.e. 4 hours a day for 120 days = 480 hours + 2 hours a day for 245 days = 930 hours in one year) is next to impossible with a group 5 language!
For comparison, see Stevi’s MIA approach in Japanese (7000 hours in 3 years):
“This week we talked to Stevijs3 AKA Stevi, who has studied a staggering 7,000 hours of Japanese in just a period of 3 years. He tracked every minute he spent studying the language in an extensive excel sheet, including Anki reps, pages read, and even a level up system he designed to gamify learning Japanese. In part one of this conversation we dive in to his exact Japanese study methods, philosophy, and progress.”
Learning Japanese: 7,000 Hours Later (ft. Stevijs3) - YouTube
Nice day to y’all
Peter
PS -
Experienced learners track their progress, but don’t rely simply on feelings!
Peter isn’t there some kind of diminishing returns for learning so tracking hours is obviously more effective but will be off by up to 5-20%?
Most other skills have a physical component so there is a bodily limitation on it for example free throws, golf, or shotput.
Doesn’t the same thing exist for language learning one has less concentration the first hour vs 3rd vs the 7th.
Also watching anime (low coefficient) is not the same as watching a book (high coefficient) vs writing an essay (no idea) so where do we get the productivity coefficients?
Also, I’m glad that I’m not the only one that seems this example as outlandish and a bit unrealistic the 3 years 7000 hour guy is a lot more realistic because 3 years is a lot longer than 1 year. Also people aren’t going to count their contemplation studying before they went hardcore so he could be at 3.5 or even 4 years but the first year was dabbling etc (but still counts).
Personally how big is the divide for spelling pronuncation from french from spanish you said there are 2 types im curious of your experience because you know all the romance languages haha?
Also korean is a cat 5 language but used to use chinese characters is this part of that consideration and korean should be lower than cat 5 since it doesnt use characters anymore or is still justifably just as difficult as chinese/arabic/japanese?
Thanks Peter! Very analytical you are!
@Michilini
“I don’t know why you got the impression that it wasn’t.”
Will said in Rita’s interview that he knew only ca. 100 Chinese characters by heart - if I remember correctly.
Apart from that, I don’t know what books you’re refering to (the number of words, the level of difficulty, graded / non-graded, etc.) and what “good” means to you.
Therefore, similar to my “bar / restaurant test” for assessing a learner’s fluency, here are the following experience-based reading tests (in all L2s that I know - apart from Latin):
- A learner can read non-fiction texts such as Harari’s trilogy (“Homo Sapiens”, etc.) with ease? → B1 / B1-B2
- A learner can read undemanding contemporary fiction texts (Stephen King’s “It”, etc.) with ease? → B2.
- When learners have read approximately 3-4 million words of fictional and non-fictional texts in a group 5 language, they reach the advanced level (C1 upwards in reading comprehension). So if you want to know what the reading stats of an advanced reader of Mandarin look like, just check the LingQ profile of Chase Bodiford: Anmelden - LingQ
That’s all the time I have for today.
Nice day to y’all
Peter out
On the other hand, most learners will probably never study a science related to skill acquisition processes (psychology, SLA, etc.), but will simply be inspired by “personal success stories”. So there is a place for personal storytelling, too!
The reason there’s a place for storytelling is not merely because “most learners will never study a science”. It’s because this is not an area that lends itself particularly well to scientific study.
This isn’t physics. When studying how people learn leanguages, there are a huge number of variables involved. It is extremely difficult to conduct reliable expereiments or measure outcomes with precision.
That doesn’t mean we shouldn’t try and study it scientifically but we should be humble and recognise that our tools are so blunt that we’d often probably get more benefit from sitting around a campfire and telling each other stories.
To put this in a language that might resonate with you: I think “ghost talk” is often of more value than pseudo science.
Since I read your post I’ve started to experiment with sentence mining, and I feel it is already paying off! It’s a really interesting topic to look into.
That’s great to hear Essie! That’s what it’s all about. I’ve also been sentence mining and it is an absolute game changer.
Was will systematic about mining which sentences should be added and which not?
Did they mainly come from TV shows?
@PeterBormann:
In contrast, achieving this feat in less than 1000 hours (i.e. 4 hours a day for 120 days = 480 hours + 2 hours a day for 245 days = 930 hours in one year) is next to impossible with a group 5 language!
In general I would agree if you’re going full spectrum across all the focus areas: reading, writing, listening, speaking.
My hypothesis, however, is that group 4 languages (you have a typo: you mean group 4) are overall group 4 but may be lesser or greater than the quoted number of hours if you target only one or two focus areas.
It gets interesting if you ask the question: would the language in question still be of equivalent difficulty if e.g. it had an easy writing system? or none?
My own theory is that dropping the attempt to learn to read cuts down the effort required. That combined with the unprovable (but logical) hypothesis that our brains are more clearly evolved for speaking/listening than they are for reading/writing (circa 100,000 years of speaking vs 4,000 years of writing).
The components of whether a language is group 3 or group 4 are: substantial or extreme cultural differences and linguistic differences from English.
“linguistic” means vocabulary, grammar, writing system and other features (like tones).
Mandarin Chinese has tones, it has almost no cognates with English, it has significantly different consonant pronunciation, it has a different but not very complicated grammar system and it has a substantially different writing system.
Russian has no tones, it has maybe 10% cognates with English, somewhat different consonant pronunciation, extremely complicated case system and a somewhat different writing system.
In both of these cases, dropping reading and writing should make it quicker to achieve speaking and listening only.
My theory was that dropping speaking should bring it down further.
The interesting question is by how much. In the case of Chinese vs Russian, is most of the perceived difficulty in the writing system? I’d guess a large chunk of it is.
It’s also clear that the tones and difference in consonantal pronunciation makes it difficult to learn to speak.
For listening, however, I suspect Chinese’s much more simplified grammar is an advantage. If therefore, you could focus on the tones and pronunciation first, I think it is possible that you could drop the difficulty level of Chinese into group 3 instead of group 4.
Idle speculation, though.
I won’t know until I tackle it myself, which is some time off for me.
@Atlan
Everybody is so serious about the discussion that I just laugh at the girlfriend factor
LOL. Correct. It would be disappointing (though I suspect has heavy weight) if the two most important features were the immersion through friends plus the girlfriend, because those are the two things that cannot be easily replicated in this method:
“The girlfriend plus hang out with native friends method” sounds a lot less impressive LOL.
By hyperfocusing on a specific area, you will definitely see more gains in that area than you would by splitting your focus in any way, but I also think there is a lot of cross pollination that happens with language acquisition.
By getting some reading, listening, speaking, writing and cultural exposure all at once (often literally samtidigt), you get more efficient gains overall. And possibly, in the long run, better results in the specific areas in less time (duration and hours), but that is speculation from my end.
That is, I think if the goal is overall mastery, I think it is likely a mistake to completely forego the other areas for a medium/long term time. We can never go backwards, but if I spend just 5 minutes a day studying some kanji, that is 30 hours in a year.
It does not take much, but just that little bit of intensive focus every day will pay dividends in the long run (I believe). Even when they are not a focus now.
I’m coming to the conclusion that is true. Particularly speaking plus listening. It definitely seems that pronunciation practice improves listening. It is likely true that speaking practice also improves listening.
What’s also interesting from the use of anki in particular is that the word/sentences/logograms/whatever if combined with audio, you eventually get some of the reading for free just by repeated exposure, even if you don’t do much reading.
I find the whole concept of language learning super interesting and also the tweaks you have to make to the approach when tackling languages from different language groups.
It was a combination of TV shows, conversations and corrections from native speakers. He mined sentence he thought would be useful in conversations on topics of interest.
“recognise that our tools are so blunt that we’d often probably get more benefit from sitting around a campfire and telling each other stories.”
Unfortunately, I don’t have enough time, therefore only this reply for the next few days.
No human being is able to understand all the scientific tools (i.e. concepts, methods, theories, statistical approaches, algorithms / math, etc.) any more that exist today to study languages (language development, language plus other media, hybridization of media, nonverbal, oral, written, etc. communication, and so forth).
That is: when it comes to first and second languages, there are so many scientific disciplines involved (e.g. communication / complexity research, psychology / cognitive sciences, media studies, SLA, linguistics, neuroscience, etc.) with sub-(sub- etc.) disiplines that we can speak of a hypercomplex epistemic network that overwhelms everyone!
It’s therefore almost impossible to draw a clear line between science and pseudoscience based on a profound understanding of things. The simplest criterion is usually:
- If XY is used in scientific communication, it’s science(-related).
- If XY is used in other communication sytems (e.g. mass media, politics, etc.) it’s not science(-related).
“I think “ghost talk” is often of more value than pseudo science.,”
There may be occasions where the “ghost talk” of lay people has more value than that of (pseudo / real) science, esp. when people want to have
specific tips that can be applied to their individual situation.
Science, in contrast, tends to be more abstract, general, complex / complicated, etc. so it’s sometimes difficult to apply to specific circumstances.
On the other hand, lay people without a scientific background are often so epistemologically behind that they don’t even know how far behind they are!
A case in point is “communication”:
I doubt that a person without a scientific background can contribute much any more to how nonverbal / oral / written etc. human communication needs to be conceptualized today.
Lay people usually repeat just obsolete metaphors (sender / receiver, input / output, active / passive etc.) without understanding that they’re only using metaphors.
So should you, Michilini, decide to write blog posts about communication processes, active / passive listening, etc. without any scientific backing, I doubt that there’s much value in it.
In short, the explanatory potential of your “ghost talk” would probably be on an (epistemic) zero line in this case.
Of course, you might “feel” different, but “feelings” don’t mean much in this context (see also below)
To give you two other examples from your own recent writing in this forum and on your blog:
- “total fluency”
There’s no such thing as “total” or “perfect” or “complete” fluency, not even for the most educated adult native speakers. Why?
- There’s so much specialized (legal, medical, tech, economic, etc.) knowledge in our hypercomplex society that no human being is able to master / understand it all. Instead, we’re often speechless in many of these domains because we aren’t able to understand or use the domain-specific lingo!
- Even with everyday things, we’re sometimes completely speechless- For example, in my case, reg. the use of makeup, American football, baseball, etc… I’m not able to appreciate these things because my understanding and vocabulary in these domains are more or less zero!
- According to the linguist Eugenio Coseriu, language is a network of networks of varieties (idiolects, sociolects, dialects, etc.) . As soon as a language grows bigger, it’s impossible to master all these varieties!
In short: Language is not a simple unity with only one pronunciation. It’s rather a complex and associative network of subnetworks of… where the vocabulary, the pronunciation, etc. are constantly changing!
And if no native speaker can master all this complexity in a larger language community like Mandarin, then neither can non-native speakers (including Will!).
In short: yes, there are varying degrees of fluency, but there’s no total (perfect or complete) fluency!
- Fluency and proficiency as (vague) feelings
“Fluency” (in the oral dimension) and “proficiency” (which includes several language skills: speaking, listening, reading, writing, and cultural “fluency”, sensu: cultural background knowledge) can be interpreted as “competencies”.
What you seem to be saying is that these two expressions can’t be determined once and for all, therefore they must be vague and you can rely only on your feelings. That seems to be in your case:
- “If I feel fluent / proficient in Mandarin, then I’m fluent / proficient.”
- “If I feel that other non-native speakers of Mandarin such as Will are fluent / proficient, then they’re fluent / proficient.”
a) First of all, that seems to be a complete misunderstanding of how language works.
We know (at least since Derrida’s writings of the late 1960s) that the meaning of words can’t be fixated once and for all (esp. not for all contexts - keywords here: “the collapse of the static structuralist model”).
That is, words depend radically on their contexts. And when contexts change, the meaning of words also changes (more or less subtly). Thus, words can - always - be interpreted in various ways.
And these interpretations can be criticized by weighing the pros and cons of each and every interpretation that is mentioned in a specific context.
If that wasn’t possible, human language would never exist.
Therefore, it’s possible to determine expressions such as “proficiency” (in a language), but It’s only one definition among others that are possible.
The question is then: how useful is this definition in specific (but never in all!) contexts?
b) Your solution seems to be, instead:
No one knows what (language) “competence” really really means (in the sense of an absolute truth / reality?), therefore I can only rely on my feelings.
Unfortunately, feelings (even from experts and even more so from beginners and intermediate users) are so unreliable without data, statistics, etc. to assess one’s own skills (or the skills of others) that such assessments are more often wrong than not.
I could write much more here, such as why your absolute statements about the myth of the danger of early outputting or the use of SRS-based “sentences” are wrong.
And it’s quite obvious to me that you lack both the knowledge and the (learning / teaching) experience in this context.
But that’s ok, we all were beginners on our SLA jouneys…
However, your (over)confidence in the superiority of sharing personal “ghost stories”
- without any data
- without any stats
- without any scientific knowledge / methods
- without experiments
- without enough experience both in learning and teaching languages
etc.
has no substance at all, because it only leads to obscurantism (vague feelings, the mysteries due to all the intangibles in the linguistic world, etc.).
Be that as it may,
- Inexperience is one thing.
- Overconfidence (due to inexperience) is another thing.
- However, obscurantism is something else again - and I doubt that the language community needs more of it. But, of course, feel free to do what your heart desires
Nice day to y’all.
It would be really fun to participate in this discussion a little longer today, but my boss is going to rip my head off if I’m not ready on Monday
Anywhere, here’s a recent video from the great Korekara podcast that could be interesting in this context:
And there was me thinking you were a fan of my work…
But seriously, I wasn’t pitting science against stories. I was actually trying to make the same point as you and say we should all be humble and learn from others with an open mind, whether they are a ghost talker or a scientist.
My background is philosophy. So I agree with you that words like “fluency” and “proficiency” aren’t scientifically precise. But that doesn’t mean we should never use them, especially outside a sicentific context.
If we adopt the approach of not using words unless there is a consensus over their precise scientific meaning, then we basically have almost no language left to work with.
The examples you cite in my case are taken from the headlines. Unfortunately, in today’s online communications climate, bombastic headlines are necessary in order to get anybody to hear what you have to say or write.
By the way, I’m actually slightly stunned that you can on the one hand raise obscurantism and ghost talk as a problem and in the same breath cite Derrida as a credible source.
Anyway, let’s not fall out over it. I have some content coming out over the next few weeks that I think will be of huge interest to you. One is with a world leading phonetics professor with flawless Mandarin. Another is with perhaps the world’s most famous language learner…
Will’s progress in only one year is far above what I’ve seen from many other people after studying Mandarin for many years. He has tested the boundaries of what I thought was even possible in terms of speed of acquisition for Mandarin. And all without even using any pre-made decks of flashcards like the HSK decks which I previously thought were the foundation of learning Mandarin.
He’s made me totally rethink my strategy for learning Chinese. Sentence mining seems to be a strategy that not may people talk about and yet Will has demonstrated tremendous progress from using it. I had previously not heard of this method at all, but I plan to use it heavily over the next few months and I can’t wait to see what my results will be.