This is the Truth, and the Way, and the… sooner one stops rationalizing away the observable fact that the pathway to the good is through suffering, the better.
2 Likes
Hahhaa, when I’ve been feeling my worst at French, that’s when I’ve seen the best spurts in improvement. I have felt delusional about the R in Livre for months and just recently had it confirmed to me that French uses several different R’s as they’re represented in the IPA.
The amount of times I’ve been on the verge of barfing just trying to work my tongue and air flow down to my throat to get that ever elusive R. Eventually it will pay off.
You must certainly been past this book by now in difficulty level, but, la petite fille de monsieur Linh (very sweet book) took me months, and I honestly struggled so much with it in the beginning that I had to drop it, now I can read it in less than a week.
Same goes for le grand cahier which is what really propelled me in my French. I will say this. I have never struggled like that again in any language again. My Russian came faster to me after I had first learned the ropes of French.
À propos efficiency, I think that’s a learned trade in language learning, you either self doubt yourself too much or get too bogged down on where you should be. Practice gives experience, and perspective to show what the most efficient practice for you is.
1 Like
I’m curious what they are. I know that the r at the end of fleur, for example is generally hinted at, whereas the r at the start of rame is fully articulated. And the r in être is often just hinted at or even goes walkabout.
Of course each consonant and vowel varies according to accent, with some speakers articulating the r with a trill as per some English dialects.
In my opinion the IPA is only an approximation. I’m sure that in German the soft ch has several forms determined by the surroundings gubbins, to use the technical term. Of course it could be that the reference dialect does only have one form, I can’t comment on that. Or perhaps I am mistaken.
It’s a tightening of the throat while breathing out to voice a vowel that increases the flow of air over the dangly bit, to use another technical term. Thus the dangly bit starts vibrating to produce a trill. We need to build up the muscle memory associated with that action, so that it becomes second nature. I suspect I’m telling you the obvious. I find that listening to speakers often helps, as it gives me clues as to how they are shaping their mouths during the articulation of the sound in question. Sometimes when I can hear the sound, by which I mean that I have built a representation of it in my brain, articulation is trivial.
1 Like
Basically I agree, though one hopes for a way is not too inefficient.
From what I’ve seen of DuoLingo, I suspect there is not enough bang for the buck there to learn a foreign language beyond whetting one’s appetite for the Real Work.
Well, there certainly is no lack of those who do know the Right Way and are more than happy to tell you all about it!
I think I had unrealistic expectations based on various polyglot channels that I could reach B2 in a year. I could tell that I was making progress but after a year I knew I was, and still am, a long way from full B2.
Re: Maigret – Simenon goes above and beyond standard crime fiction in that there is always an interesting psychological layer. There’s also a lot of conversational French and expressions which I don’t find in news and articles.
2 Likes
I want to be able to understand native speakers when they’re talking to me. They usually make it a bit easier to compensate for me being a foreigner but that varies from person to person and you can only grade your language so much before it stops being a conversation. I want to be able to understand them well enough that they don’t feel tired after talking to me 
. When I was very seriously thinking of living in Japan then of course I wanted the whole deal. I wanted to understand natives all the time. This is super hard. Since I left Japan I also lowered my goals for my Japanese and basically it’s better than I need really. So I started studying Korean instead.
2 Likes
Reaching B2 can take a while even for easy languages. The key is if you enjoy studying it and getting there. I have been at Korean for 5 years and the progress is slow but I enjoy it regardless. It would be nice to be B2 by now though lol.
4 Likes
I agree.
Indeed. Most are pushing an application, or a product such as course books and flash cards. Others are chasing clicks i.e. YouTube income. Caveat emptor.
My German has progressed at roughly the rate predicated by a couple of experienced German teachers on YouTube. Probably not dissimilar to your French progress. My pronunciation is probably well above average, due to previous language learning experience, but learning words and grammar is a struggle.
There are people who do learn a language to a high level in what seems like a short time. Lýdia Machová learns a language every two years. I think there are ways to accelerate ones progress, through deliberate study. I haven’t figured them out yet !
Thanks. I’ve ordered a second hand Maigret book from eBay.
This is entirely subjective, I have no idea how true it is, except for a few YouTube videos. But I believe French has two R’s both produced in the back of your throat. However one is trilled, and one is, more throaty. This is the video I base it off of. But I say it because my R’s following consonants seem natural, but the R after livre is wrong.
This is the one I do along with audio file.
Second one is what I aim for. I might be completely wrong, but her video made me believe I am at least partially correct.
1 Like
i think the rolled r in french is not common in modern day metropolitan france unless the person is very old or speaking in a regional dialect . it was more common a long time ago a lot of the older french singers like edith piaf spoke and sang with that r
2 Likes
I am getting confused. My dictionary represents the French r with the R symbol. That is a trilled uvular r. However this site represents it with the upside down R:
To my ear the two sounds are similar, it’s just that one sets the uvular vibrating. The non vibrating one requires a bit less force, a degree of gentleness.
I found this interesting:
2 Likes
Incidentally, do you have correct timing? I don’t know your native language. In French it is syllabic. Get that wrong, and the r consonant can be hard or even impossible to get right.
The goal posts keep moving. First it was to have a very basic understanding; then to be able to understand to a level where I could have a have a conversation with someone; then it was to be able to understand TV series and movies; and now it’s to be able to understand the wide range of Italian accents with ease in pretty much all situations (this also includes the various dialects, which Italians also understand, such as Romano). It other words, scope creep.
In terms of difficulty, I wrote this list a while back:
It’s not in there, but also dubbed TV series and movies are easier than native ones.
Ideally, you want both enjoyment and efficiency. Low enjoyment is very demotivating, but so is low efficiency, because you can’t see progress.
If you put an hour of study in per day for two years, you come out with 730 hours. If your efficiency isn’t too low and you focus on the spoken language, you should be able to reach B2 in a related language to one you already know in this amount of time (such as a Romance or Germanic language for the hybrid language of English). This aligns not only with the experience of Lydia (who spends an hour per day) and other polyglots, but also language schools. If you do the maths for the classroom time and homework time for the language schools, you come out with a similar number to get to B2 (for Romance/Germanic languages at least). I also passed a free online Italian test for reading and listening comprehension with an estimated ~700 hours too. (Note: passing a language test is easier than fulfilling the criteria of the CEFR levels.)
It’s probably best to think that some of these polyglots learn a language for two years before moving on. That’s it. Not that they reach a certain level before moving on. Though, Lydia probably does reach a B2 in two years, depending on the language in question. Secondly, B2 is actually not that high. You will have reached what could be considered basic conversational fluency, but you are still lacking in many areas. C1 is a much more solid level of fluency, though, I’m still thinking it’s lacking. My guess is I might be able to pass an Italian C1 comprehension exam either now or very soon and I still consider myself quite lacking.
2 Likes
Very true.
I used to be very anti Duolingo for countless reasons including low efficiency. There are people who have used it for five or more years and cannot speak the language. However lots of people enjoy using it, and these days I just accept that there are many ways to deprive a feline of its outer layer.
You pointed out some while back that the FSI estimates for study time for various languages are class room hours, and I think you were right to say that they assumed additional out of class study, even though I could not find explicit information. Also that 700 hours is for an ‘easy’ language. Lýdia Machová does as you say spend an hour a day for two years on a new language. I assume she gets to a fluent level i.e. able to hold a conversation on general subjects.
She states on her company website that she is fluent in seven languages, and has basic knowledge in two others.
It is often said that language learning gets easier the more languages you learn and I suspect she has methods to progress more rapidly than many. For example, one common trick is to focus on necessary - commonplace - grammar, and ignore an in depth understanding. She’s also used to speaking and learning.
I know some quite obscure vocabulary in French, but only because I want to reach a very high level. So my comprehension is quite high, but my self expression is mediocre, despite good pronunciation and despite studying significantly more than one hour a day over two years.
What I note about Machová and others is a high level of practical attainment. Thus her English is pretty much native level with a slight accent or colouration to her voice. As another example, Richard Simcott speaks numerous languages with a near native accent and grammar. I’m sure if you pushed them in their second languages you would find holes.
Sorry. I wasn’t very clear. I was not referring to the FSI. They do not use the CEFR.
For example, let’s consider the Goethe Institut website (the most well-known language school for teaching German), but you can look at these numbers on any of the large German language schools, which have been around for years and they will have something similar written on their website. Note: ‘h’ is the symbol to mean 45 minute lessons, not hours!
The number of lessons that are usually required to reach a certain level and take the corresponding Goethe certificate are based on experience. Individual aspects such as language learning experience and native language also play an important role. We will grade you before the start of the course in order to ensure optimal learning success.
Their scheme is five 45 minute lessons per weekday, totaling 3.75 hours per weekday or 18.75 hours per week of coursework. Based on their experience, the “number of lessons that are usually required to reach […] and take the corresponding Goethe certificate” for B2 works out to be 600 of their 45-minute lessons. This totals 450 hours of classwork.
I’ve never been to the Goethe Institut, but I’ve been to other German language schools, and my homework hours were not massive by any means, nothing like what they expect at the FSI (you can imagine why). Even if you want to be extremely generous and say that they are asking you to spend the exact same amount of time doing homework as in class (which they are not), you are only looking at 900 hours. This is an hour per day for 2.5 years, so pretty much the same.
This is for German, a language similar to English from a historical, cultural, grammatical, vocabulary point of view, so it obviously takes less time than one which is not.
But it’s not very surprising that B2 for a similar language to one which you already know “only” takes two years to get to with an hour per day. You can do things with the language, sure, but it’s still a low level. I think people overestimate what reaching it actually means. I think they are just putting their own impression of basic fluency on it or they read the CEFR descriptions of it. In actual practice, passing a B2 exam can be achieved with a lower than that. If you are at the lowest end of B2 and just passed the B2 exam, you will struggle quite hard with a lot of things. I’ve met people who have just reached B2 and some of them seemingly struggle with standard conversation, not to mention they won’t be able to read native books at all.
The interesting thing for me is the amount of hours required for C1 and C2 (requiring the same amount of time to get through the C-level as the B-level). With numbers like that, I think I should take an Italian C1 or C2 comprehension exam and see if I can pass it.
2 Likes
You give an example of an easy language (for an English speaker) and the study time including homework is noticeably more than one hour a day for two years. And as you say, the end result, B2, is not particularly high.
However, the point I was making is that there are ways to accelerate one’s learning progress if one wants to achieve a working fluency, and many are used by well known polyglots. Obviously these tails of fluent in 3 months are nonsense. Research has shown that students who take control of their learning make more progress than students who take a passive approach. We also know that quantity of input is important, and most language learning takes place implicitly. My experience is that classes are mostly explicit teaching, and language requires significant practice - especially reading and listening - for that to sink in. When the aim is to achieve a working fluency, one can ignore the uncommon vocabulary and grammar, and focus on a broader knowledge.
There are other methods to improve learning, obviously spaced repetition is a well known method, but interleaving also leads to improvements. Thus different apsects of grammar are interleaved, rather than doing each one in turn, for example. Regular testing has also been shown to provide benefits. I have no idea how the CEFR courses teach a language, or even if there is any standardisation in the teaching methodology. Perhaps only the curriculum and tests are stndardised.
The key advantage of the CEFR is that many jobs require candidates to have a minimum CEFR level of the working language is not their native one, and CEFR provides a common standard. Thus many people will not do a CEFR course, but they will take the exam(s).
I don’t understand why a language is ‘easy’ if it takes mildly less time to learn than one which takes thousands of hours. I’ve never studied Japanese, so maybe that’s why, but I don’t consider Russian any harder than Italian. It is just slow gains. I don’t consider a book, which has more pages and hence takes longer to read, to be harder than one, which has less pages.
My extremely rough estimate of 900 hours for reaching B2 in German was the upper limit based on their “usually required” number of hours. The homework hours I used was extremely generous at 18.75 per week. From my experience with (other) language schools, I was doing more like an average of half an hour to an hour per day, so 3.5 to 7 hours per week. With the 18.75 hours per week for classwork and 7 hours per week of required homework, we are looking more at an estimate of a “usually required” 618 hours to get B2 through a German language school. This is less than the 730 hours of Lydia’s hour per day for two years. Is this “noticeably” less? Personally, I don’t consider plus or minus a few percent much, as the noise is larger than this.
My point is that even with an intensive language school, you can achieve B2 in the same amount of time as Lydia’s ‘hour per day for two years’. But if you’ve seen the below video from Lydia, you’ll notice: (1) she is predominately learning languages, which are similar to languages she already knows, (2) I don’t know if she actually takes B2 exams for all her languages afterwards?
I consider that the CEFR only has limited use, mainly limited to intitutions, agencies, and companies to make decisions on paper without ever having to meet the individual. Eg. If a company has 400 applications for a position, they can just weed out all those who do not have the certification, then do a more in-depth assessment of the applicants’ language abilities afterwards. Alternatively, it can be used by institutions, etc. as a way to outsource the work of assessing someone’s language level, as they don’t want to do it themselves.
As a general note, the downsides of using the CEFR out of these contexts are:
- The majority of the population are not aware of the framework at all, so don’t understand your references to it. Eg. Your friend’s aunty won’t understand you, if you use it.
- There are plenty of people who say they are at a certain CEFR level, but have never taken a relevant test for it. I.e. People are predominately self-assessing, which is the equivalent of self-diagnosing.
- People then judge other people’s claim of a CEFR level based on their own ‘self-diagnosis’ of the level.
- Pretty much only some European institutions use the model to refer to some European languages. Eg. English or Hindi wouldn’t have institutions whose job is to give exams to measure you according to the CEFR. So you are just guessing again.
- The implentation of the CEFR levels in the form of exams is quite different from the qualitative descriptions of the CEFR. I.e. You can pass an exam with a language school at a particular level, but it’s still debatable if you actually fulfill the CEFR qualitative descriptions.
There are much better ways of discussing an individual’s ability than to ask for their CEFR level, if you are actually interested in knowing. This is for individuals, not for the mass paper shuffling of institutions. It isn’t standardised in any way, but provides a better understanding of someone’s language ability than asking for their CEFR level based on their best guess.
- A native speaker will very quickly be able to discern someone’s conversational ability in a matter of a seconds.
- On LingQ, you can glance at their stats (obviously taking in mind it’s an underestimate, as not everything is recorded). Or ask how long they have been studying for (you want to convert the answer into hours).
- You can ask someone to describe what they can do with the language. For instance, this is how I describe my level to friends and aquaintances, who have never heard of the CEFR for the life of them:
a. For Italian, I can go to a party and spend the entire evening in Italian, no problem. I can watch TV series and movies and listen to podcasts. I don’t understand every single word, but I understand everything that’s said. I struggle a bit with reading books without a dictionary, because I lack the vocabulary. However, I suspect in another half a year of study, I’ll be at a level, where I would be able to university in Italian, if I so felt, with the exception being for law or literature.
b. My Russian is still reasonably beginner, but I can somehow scrape by talking to a native Russian speaker who speaks simply to me for several hours with a bit a difficulty. - Last resort, ask if they have done a CEFR test. And this is the bare minimum they are.
At the end of the day, Lydia’s and other polyglots’ claims are hard to judge and compare. How do you know she got to the goal faster than others? Was the goal even the same? Or did she reach a higher level with the same amount of time invested? It’s just assumed that she (and other polyglots) are more efficient, as a sort of general improvement of learning ability, or perhaps better said, due to the result of trial-and-error process, an evolution of sorts. How true is this assumption?
On the contrary, I think more people are taking language courses than doing CEFR exams. I’m not sure what CEFR courses are. Do you mean a language course of a European language, which decided to put B1 on the course title instead of saying lower intermediate?
Side note: I read a paper or two a while back by the fella, which Steve Kaufmann likes to quote in his videos about interleaving, and it was rubbish research. The sample sizes were so tiny, it wasn’t even worth paying the cent to host the server, which it was uploaded to. Interleaving sounds like a good theory though. Not sure, if there’s any legitimate research to support it though.
3 Likes
Just hopping on the discussion train here. Harder is usually meant exactly that way: requiring more time to achieve. If an instrument takes longer to learn then another one, I would also say it is harder to learn. You may use a different wording, but in the end we are all talking about the time needed.
Doesn’t that apply to all kind of certificates or degrees. I have a diploma in mathematics, but most people don’t know what exactly I had to accomplish in order to get that. The only thing they can assume is that I am probably quiet good at calculus 
The average friend’s aunty is probably not the target audience for this certifcates anyway.
And any unproven claims should always be taken with a grain of salt. If someone claims to have a certain level in a language and you judge that based on your assumptions on your own level, without any of you ever have taken your language skills to a test (of whatever kind), I wouldn’t expect reliable conclusions to be drawn from that. But that isn’t the problem of those who actually make those tests and those who use those tests to filter out applicants, for example.
- This heavely depends on the level of the native speaker. Beeing native doesn’t equal to beeing good at a language. I’ve met lots of L2 German speakers who speak much better German then many native speakers. The span of competence is enourmous. In addition, conversational ability is a very vague term, as the kind of conversation plays an important role.
- I am not sure to which degree you can judge on ones competency based on some stats. It gives you a rough impression, surely, but so does probably any test.
3 Likes
The American FSI class language difficulty on the basis of how long it takes students to reach a decent level, however they define that. Essentially the further the language from one’s native languages), the longer the time required.
You’re assuming they reach the same standard as Machová and other high achieving polyglots. The latter demonstrate a high level of achievement in grammar and pronunciation. I have no idea what the standard of students on CEFR courses is like when they are certified as B2. Perhaps someone who has done a course could comment. See also comments at the end.
Many universities, probably most, will require proof of language skills from foreign applicants, and some countries require, or favour candidates with, specific language skills. Quebec province in Canada, depending on the job, may require French skills. So regardless of your opinions, many people do require such certificates.
Absolutely not. Languages are far too complex.
I said many and not more. I don’t have any figures so cannot comment on numbers.
I mean a course and tests that comply with the CEFR levels:
There’s plenty of good research, some is referenced in Make It Stick written by two psychology professors from Washington University and a professional writer. Interleaving is used in non language teaching including by many sports coaches. I find it very effective for ice skating.
There’s also some good research that suggests that language learning and production uses similar neural mechanisms to activities such as playing an instrument and ice skating. See Declarative Procedural Theory for example.
I’ve met hundreds of people who learned English as a second language. One friend arrived from Korea aged 15, and lived here 30 years. He has a noticeable accent, and cannot understand films in English despite speaking well. A Russian speaking friend has lived here 19 years and is barely intelligible in English.
You have previously acknowledged that learning methods vary in their efficiency. They also vary greatly in efficacy.
Numerous polyglots have demonstrated an ability to reach high levels in multiple second languages, both in grammar and pronunciation. I don’t know if there is any solid evidence to show that they learn any quicker than most other people, or even if they do, but I do know they use numerous useful techniques which are not used in conventional classroom based language teaching.
When I went to evening classes with several respected organisations - Institut Français and Alliance Française - the teaching was very traditional. They certainly made no attempt to teach me how to learn.
My experience of formal full time education - a degree in physics - was atrocious. I attained a top grade but I had to work ridiculously hard. They used lectures, exactly as per 100 years earlier, maybe 1,000 years earlier, which are a very inefficient way to learn. Unfortunately physics researchers think that being an expert in physics makes them expert teachers. I’ve seen the same phenomenon with native French speakers who thought they were expert French teachers.
I have since a number of very useful learning techniques, some of which I learnt from polyglots, from their videos and books. Spaced repetition is an obvious one. I use it with phrases, not words. Regular testing is a similar technique to spaced repetition. Interleaving is another useful technique that was mentioned earlier. Then there is training the ear to recognise speech by listening while reading the transcript. I found this simple technique significantly improved my ability to follow speech. Another technique is listening like a child. Simply listen to speech, focussing on words, but not translating, not trying to decode, just listen, and let the brain work on it. Then there is mindset, a so-called growth mindset as described by Professor Carol Dweck can make a significant difference to the end result, as demonstrated by research. And as for accent, there are tehniques to improve that. For example, many languages have a different timing to English, and once the student is aware of that fact, they can work on their timing and make a noticeable improvement to their speech. I had to discover by trial and error that French is syllable timed. 30 years ago some French speakers were in effect trying to explain that to me, but they didn’t know how to explain it, and I didn’t understand. I’m sure there are many other useful techniques. All too often a learner will hear a difference, but they need explicit tuition to allow them to realise that difference for themselves.
1 Like
You don’t need to pinpoint their exact level down to the percentile, but rather have a rough judgement of their level. You talked about your experiences with your friends. You don’t think a native speaker could, for example, estimate your Russian-speaking friend’s conversational ability (listening comprehension and speaking ability in the conversational setting) quickly? How long would it take someone to have a rough judgement of their conversational ability?
EDIT: Moved this response to another thread instead of bombing this one. xD
I thought this is still debatable in the scientific community? I don’t keep up to date on these things though.
A lot of the level claims are indeed like this.
1 Like
My Russian speaking Lithuanian friend actually speaks well and has good comprehension, but his accent is so strong you’d think he was a beginner. He claims to speak Polish, his literal mother tongue, and Lithuanian, the language of the country he grew up in.
I think it could take a few minutes to get an impression of someone’s level, and even then it would just be the conversational level. The problem is that we could be dealing with someone who is very confidant but with a lowish level and can bluster their way through, or someone reserved who has a high level, but does not show it. And then they might know basics, but talk to them about current affair, and they dry up because they don’t know words for on parole, in custody, prosecuted, surrendered, retreated, sexually assaulted and so on. And a perfect accent can confuse people. I once met an Italian with native level American English pronunciation, but beginner level skills !
Then again, it comes down to how you define level. I knew a female French Canadian who spoke English, but essentially used French pronunciation, grammar and phrasing, and a male English Canadian who spoke French with English pronunciation, grammar and phrasing.
These standardised tests provide a more reliable assessment, though I don’t know if they are a measure of breadth, or just grammatical competency. Breadth of vocabulary is a huge part of competency.
I believe it is considered to be a valid concept, not that I am an academic psychologist. Wikipedia, which I do not consider a reliable source, lists a negative study, but Dweck in her book lists many studies showing results consistent with her theories. I speak as someone who used to have a fixed mindset. From a personal perspective I have found it to be a very valid and useful concept. When I was growing up, I was told that I was poor at languages and sports, and that people are either good at technical subjects (science, maths etc), or good at non technical subjects (literature, French etc) and sports. Complete nonsense of course.
In some ways this is comparable to job interviews. Most interviews I have been to asked a series of questions aiming to assess my knowledge of software development. I usually did badly. In some cases they gave a test whereby you had to write some software, either at home, or during the interview. I always excelled, in fact in one company I was the only person to have come close to the perfect solution of a difficult problem. I believe that the former style of interview relies too much on subjective impressions, tone of voice, self confidence and unconscious bias. Thus I was never hired by people significantly younger than me, which became a problem in my fifties.
1 Like