I like Paul Noble as an SLA researcher a lot. He’s also written more practical books for learning / acquiring L2s. Here’s his Audio course for French for beginners (esp. for native speakers of English): https://www.amazon.com/Collins-French-Paul-Noble-Natural/dp/B004TLS0IU
You should give it a shot!
Today, of course, I wouldn’t buy a CD anymore.
I’d use Audible, for example, instead.
Cheers
Peter
I don’t believe in this “fossilization” theory. You could just as well “fossilize” bad grammar or poor vocabulary. I reality, you begin with a poor level and you improve on it gradually. All areas of language learning are similar in this. If the only way not to “be defined” by your bad habits is not making any mistakes, learning a language would be an impossible task. I also don’t believe that trying to repeat sounds that you can’t even differentiate clearly is the best way to practice pronunciation. It certainly is not for me.
I agree that there’s no need for a “silent period” but, on the other hand, there’s no problem in having one and it is beneficial for many learners. Again, there are a lot of different skills involved in mastering a language and you must improve in all of them, with occasional periods of particular focus in a particular one.
Agreed!
Non sequitur
Improvement in all areas require conscious decisions. It is for each learner to decide what their priorities are in a given moment. You have to decide to improve, acquire knowledge about (conscious or subconscious) about each skill and then practice. To “master” the language you need to get all of them “kind of” right.
For me, it all boils down to “good habits”: it is much more difficult to change “bad” habits once they are established than to establish good habits - right from the start!
I’ve experienced this many times not only in language learning, but also in mathematics, programming - and also in sports.
“Improvement in all areas require conscious decisions.”
Yes.
Having worked for 10+ years as a language coach and as a private tutor for math (and some programming) parallel to my second degree in computer science, I can tell you that most teenagers and many adults I have taught do “not” have enough knowledge when it comes to their learning processes.
Unfortunately, schools in Germany seem to think that learning is something that just happens “naturally”. But, this is only true at a low level. It no longer applies at intermediate or advanced levels of learning.
That is why many people struggle not only with language learning, but also with math and programming.
In summary, “conscious decisions” require that you know what you are doing. Too many people - or at least 80-90 per cent of the people I had taught - lack this (profound) knowledge.
In other words: They have only very “superficial” ideas about learning. And this often leads to bad decisions and bad habits when it comes to their learning processes - sometimes with disastrous results.
I agree with all of the above. My disagreement comes from your insistance that we must go through a period of sentence repetition before learning anything else or your bad pronunciation will fossilize and you’ll never pronounce correctly. This is, IMO, an example of the worst flaw of most language learning advice: overstating a point. I know you think you haven’t done that, because you value flexibility and so on, but in practice you have:
Your words, and that in a thread about grammar problems!! of all places, so much for flexibility! For all you know the OP may have the most perfect accent in the world because he only talks about his troubles with grammar, but you still go out of your way to explain how beginning with Lingq would be a terrible thing from which he’ll never recover. If that’s not an overstatement, I don’t know what is.
The problem with overstatements is that someone else will try this and discover it is not a good fit for them (even if it is for you) and that person will go on to totally overstate their point, in the name of the same principles you defend: avoiding bad habits and, hey, “I know someone who repeated Chinese sentences for years and never got to speak and they still can’t pronounce all that well” or something like that… So, that person says that you must never utter a single word until you understand the language well or you’ll never speak, and so on… Or someone else may talk about how so many people freeze out of fear of mispronouncing so they never learn, and. thus it’s better to speak however you want but at least not freeze and you must avoid sentence repetition like the plague…
Then someone sees the problem of this approach and goes back to case number one,…
I can find you examples of all those positions and the same goes for grammar, advice going from “having a solid understandig from the beginning” to “don’t learn any grammar or you’ll never progress” and that again, based on the same ideas of “avoiding bad habits” and anecdotal evidence of a friend of a friend who did whatever and never got to learn.
And all of that just because people insist on taking good points and overstating them to the point of absurdity.
Your main contention about being mindful of pronunciation from the beginning is sound (pun intended) advice but there are different ways to go about it and each has advantages and drawbacks. Pounding Pimsleur worked for you (and I’m glad it did) because:
a) You don’t mind repeating a lot of nonsensical sentences. Ok but it’ll bore most people to death. This is partly personality (I guess), partly confidence because you’ve learned other languages (this is huge to get you past this phase), partly money (you don’t mind paying an additional amount for a temporary resource) and, most important, partly because those sentences are not so nonsensical to you, afer all you already speak (at least) two closely related languages and have been exposed to some French words since childhood. Most learners are quite different: Try repeating Russian sentences composed of words whose syllables you can’t repeat even immediately after hearing them, for months on end until you get all the sounds right. Hey! and don’t even touch that Assimil before you sound like a native or you’ll never speak properly.
b) Because of the aforementioned pre-exposure and the fact that German pronunciation is not very different from the French one (except nasals, it has most “strange” phonemes, including u, eu, r, …), and there are even quite a few words in German that use nasal sounds (“Balance”, e.g.), you stand a solid chance of learning French pronunciation in a short time simply by repeating. This is not the case for most people, and this is the main failure of your argument. If an English/French speaker had to master the trilled R before learning Spanish, not many people would learn, whereas I’d be a billionaire if I got one cent for each foreign person who communicates well around here with less-than-perfect Rs. Another example: if you had to master all soft/hard consonant distinctions in Russian before beginning to learn you’d probably never learn (added to the difficulty of merely repeating words). How many Chinese words do you have to repeat before mastering the tones and their varitions? When I began learning French (decades ago, it was my first foreign language) I could not pronouce the “z” sound to save my life, not to mention the more difficult sounds (“an” vs “in” anyone?). I can swear that I can pronounce rather well now, even getting mistaken by a native on occasion. There was no Pimsleur available here then or anything remotely simillar, but my mistakes did not “fossilize”, whatever that means.
c) Because it is relatively easy for you to pronounce French (I know it doesn’t feel so but, trust me, it is), you haven’t even tried other pronunciation learning methods, which are in fact way more effective in many (not all) cases. Here “no size fits all” for sure, but a combination of careful listening and knowledge of how to move the articulators is my own favorite method. For French this video series is a great example of the latter: frenchsounds - YouTube
This method takes time and it’s much better pracising as you learn the language, not as a prerequisite. For me, this method is much better. Some people call it “hearing with your mouth”, it is based on the fact that learners often can pronounce phonemes much earlier than they can identify them in others.
However, and this is my main point, I will never go around saying that you must learn phonetics and articulator movement or else you’ll never learn and so on. That would be overstating my point. This is just a method that works for me. Maybe Pimsleur works for you and, again, I’m hapy for you. I’m sure it can help other people in a situation similar to yours. NO, it’s not a pre-requisite. No, it’s not a good fit for everyone.
Bottomline: YMMV
Others have mentioned Assimil here as well, but I just wanted to add to this: The big advantage of Assmil, is that grammar points are introduced and explained within the context of the conversation / story that makes up the lesson text. So, you’re learning grammar from context, unlike a grammar book that presents the rules and then tries to provide example sentences without a context. This makes a huge difference in my experience.
One can achieve a similar effect by reading books and articles and then looking up the grammar points encountered – or just absorbing the patters. But of course that requires a more comfort with uncertainties etc. But for a beginner trying to get a comprehensive grasp on a language and its grammar, Assimil is hard to beat.
Also good to know. Thanks.
It’s very likely that this works for little children because of the phonetic plasticity of their brains. It could also work for “some” adult L2 learners. But it’s “not” an automatic result for “all” (or even “most”) adult L2 learners. So, it’s a good idea for the adult target group to focus on pronunciation.
How you want to deal with pronunciation (by using AIs like “Elsa” (for American English), by paying accent trainers, by using Pimsleur, whatever) is entirely up to you as an L2 learner. Choose what works best for you and discard the rest.
So, pronunciation issues in an L2,3,4, etc. often don’t disappear “by magic”!
Nevertheless, as this isn’t a “law” - it might happen.
But as a science-oriented guy I dont trust magic, religion or wishful thinking - even though the “placebo” effect can be quite powerful! 
The “overstatement” problem: “And all of that just because people insist on taking good points and overstating them to the point of absurdity.”
Yes - I agree. The problem starts as soon as someone tries to defend really bad learning habits (e.g. a “cramming” style for developing skills), which just don’t work very well.
So it makes sense for me to distinguish between rather “good” and rather “bad” practices - even though there can be “no” law, no “one-size-fits-all” approach and no “single best” practice in SLA.
Focusing on pronunciation right from the start of your L2 journey (as developing a good habit) is such a rather “good” practice. But, again, “how” you implement that is completely up to you as an L2 learner.
In sum: In theory anything seems possible. In practice, it’s not.
"the fact that German pronunciation is not very different from the French "
.Well, as a German native speaker who studied French at university and also lived/studied in France, I can assure you that the pronunciation of German and French is quite “different”!
And, in this respect, many Germans simply tend to “butcher” French often.
On the other hand, you could say that the French who learn German also tend to butcher German. - Yes, but when the French do that, it sounds rather “mignon” to German ears
Three last questions:
I teach Psychology of Language at the university. One of the topics is acquisition of a second language. I think I know the research rather well but, as a researcher myself, I feel there’s not so much that is usable. Research nowadays has lots of issues. A part is useful, though. But personal experience is important as well. Notice that your argument in this point is mostly about your own students, based on anecdotal
I have helped people learn languages and even taught lessons back in the day, but most of my teaching experience is as a university professor in Psychology.
My level in French is clearly C2. I even passed the appropriate exam at the Académie Française long time ago but I have improved ever since.
I got a grant to study German at C1 level at the Goethe Institut Berlin long time ago. I didn’t take an exam (I had a previous one of a lower level) and I have only practiced sporadially ever since then but I do travel to Germany from time to time and I do read and listen to German.
You absolutely lost me here
Of course there are differences and of course it is hard to learn, but they arevery close compared to most languages. Again, most of the phonemes typical learners struggle with exist in both languages, to some extent. I already gave you examples.
Agreed!!!
Main point:
But you said so!!! Again, your own words:
Here, you’re clearly telling a learner, who’s not German, who doesn’t attend your classes, that he should not use what you call “grammar oriented” courses (I don’t think this is a good characterization of Lingq, but whatever!), implying that, instead, he should use “pronunciation oriented” ones first, which use sentence repetition. No hedging, no qualification, nothing! This is what I consider “overstatement”
“I teach Psychology of Language at the university. One of the topics is acquisition of a second language. I think I know the research rather well.”
There’s much more to SLA than just the “psychology” of language. Linguistics, communication science, media studies, social complexity science, intercultural communication, SLA research per se, neuroscience, etc. - not to speak of Natural Language Processing and Machine Learning approaches (that are relevant for my startup reg. SLA).
It’s hard to know even “one” of these disciplines very well. And I’ve spent many years on working on social complexity science variants as part of my PhD project alone…
“But personal experience is important.”
Definitely! IMO, a SLA researcher or linguist should not only be a teacher / researcher, but also always be a language student so he or she can better understand the concrete problems L2 learners face!
“Notice that your argument in this point is mostly about your own students, based on anecdotal”
Yes. After 10+ years of teaching the pronunciation issues of Germans learning French were a “recurrent” pattern.
But reg. “fossilization” we don’t have to trust on (my) anecdotal evidence, Wikipedia or some obscure blog posts.
If you use “Google Scholar” and type in “fossilization” and “second language acquisition”, you’ll get more than 11000 scientific search results.
So, “fossilization” in SLA is “not” a pure matter of “my” anecdotal evidence or “your” personal belief.
Ad “French / German”:
OK. So you should be aware that German and French are quite “different” when it comes to pronunciation. And both French and German learners have many problems in this regard.
Of course, it’s more difficult for Indo-European speakers to learn tonal languages (Vietnamese, Chinese, etc.). But, that’s a topic for another day.
And even if French and German are part of the Indo-European language family, the “real” pronunciation issues for French / German learners usually don’t simply disappear - if you don’t work on them.
“I already gave you examples”. I’m sorry., but your examples aren’t helpful in this context because they don’t make the “practical” pronunciation issues go away.
“But as a science-oriented guy I dont trust magic, religion or wishful thinking…”
What I wanted to say is that fossilization, pronunciation, etc. issues usually don’t just “disappear” if you don’t work on them.
For instance, if you don’t work on “pitch accent” in Japanese explicitly, you’ll probably never acquire it and you’ll always be recognized as a foreigner - immediately.
It seems that many Japanese learners don’t even understand that their accents are just “off”. Anf if they do, they don’t know why and / or how to fix it.
I’m pro (mass) immersion, but immersion alone might simply not be enough. Dito for concrete interactions. So, corrective feedback reg. pronunciation, etc. is “useful”. And that’s not just “my” personal opinion, but that’s a position in SLA research, too.
b) If focusing on pronunciation (and this includes "speaking!) is part of a “good” habit from the start (see our whole discussion above), then a pure “input”-/ immersion-oriented approach might not be sufficient, because it does “not” automatically prevent all pronunciation or fossilization problems!
And this is a SLA research insight (against Krashen), too: Some corrective feedback does have its place in L2 learning / acquisition.
And this means further that “learning” and “acquisition” are intertwined. This is “not” an either-or-relationship!
That’s why I opt for a flexible “mix” of approaches and tools. (see my examples for both Portuguese and Japanese!).
“How” an L2 learner implements his focus on “pronunciation”, on “grammar”, on “speaking”, on “writing”, etc. is his or her choice.
For example reg. “pronunciation” alone I’ve checked many resources: books, the Pimsleur and other smartphone apps, YT vids, podcasts by accent coaches and AIs such als ELSA.
In sum, the key message for the implementation (i.e. the “how”) aspect is:
OK, there are also other maxims. For example:
But, these are all subjects for another day!
The fact that I specialize in Psychology doesn’t mean I am not interested in other aspects. For example I learn languages and I also have extensive experience in machine learning techniques. One of my teams recently won a hackathon. This is a very disingenuous argument on your part. I could just as well assume that as a SL specialist you have no idea of Psychology and so on.
The fact that there are a lot of papers about one topic doesn’t mean that the underlying research is of quality.
You seem to base your arguments on credentials and number of papers. Those are not acceptable arguments. If you know of a piece of good research that supports your point, you have to reference it. That’s how this works if you want to use research as a basis for your argument.
Notice that the fact that “fossilization” is found is not enough to support your argument. Of course it happens. The merit, or lack thereof, of your argument depends on showing:
Also, I suppose you are aware that it is necesssary to weigh all available evidence, in favor and against it. The mere fact that some researchers have found evidence in support of your argument doesn’t mean much if better research has found the opposite.
For example Krashen’s research (that you mention, and which I think is among the best) certainly does not fit in well with your idea of beginning language learning by drilling pronunciation, quite the opposite. In fact your advice in this thread runs contrary to Krashen’s conclusion and against his research, which should coun against your argument.
re: Japanese tones
I’ll be recognized as a foreigner in Japan just by looking at me =)
Frankly, there’s nothing wrong with having a foreign accent. I work with many people with all kinds of accents (I’m in IT). Many do not have perfect grammar either. But I can understand them and we can all communicate at a deep level…so really it’s not important to get perfect and it doesn’t bother me in the least that they have an accent.
Even people who are at an extremely high level in a second langague are going to be identified as foreign. Even in the English language. There is always going to be something a little off. Always…I don’t care how good they are.
Now that doesn’t mean one shouldn’t try to improve, but I think there are diminishing returns. i.e. the amount of time to get it perfect is not worth the effort beyond a certain point. One may simply never quite get it right. But that’s ok.
@PeterBormann. An example of research:
From A “dramatic” approach to improving the intelligibility of ITAs - ScienceDirect
This goes frontally against your proposal of going through Pimsleur or similar before a course. That paper gives a very interesting alternative: using dramatization to improve intelligibility. Notice that you need to know a bit of the language to start using this method and that it does teach pronunciation (which I think it’s very important) but through more organic methods, more compatible with comprehensible-input approaches and much more connected to the overall learning experience. Not as a separate pre-requisite based on minding repetition. This piece of research suggests that it is also much more effective.
@ericb100
Exactly!!!
My problem with “fossilization” is that it is a very ambiguous term. In principle, it was used to highlight the fact that adult language learners often attain near-native grammar and syntax abilities but seldom do they speak “like natives”.
As you point out, in principle this is not a real problem. It is interesting to know the causes (which are very much unclear at this point) but in practice it is no big deal.
A different problem altogether is if learners speak so badly that commuication suffers. Peter Bormann seems to have used “fossilization” in this sense, which is what I dispute. Adult learners, if taught appropriately, can achieve good working pronunciation, sometimes even easier to understand than some native accents, and you don’t need to go through an extended period of pronunciation drills prior to learning the language to achieve it. Steve, for example, is easy and even pleasant to understand when he speaks Russian. He’s no native but he can communicate. He has never gone through any pronunciation course.
Do you need a conscious commitment to improve pronunciation? Yes!
Can explicit pronunciation training / accent reduction help? Probably but the exact best method is unclear, although repetition is probably not it.
Your Japanese pitch accent example is a good illustration of this phenomenon. Again, Steve doesn’t even care for them but he can communicate very well in Japanese.
Of course, working on learning them also makes sense, depending on your goals. You need, IMO, a good understanding of how they work, paying attention to them in the speech of natives and lots of practice.
Thi series of videos, IMO, provide a good base for the first step: Learn Japanese with Yas - YouTube
I would definitely advise against trying to learn them as one of the first steps in your Japanese learning, unless, of course, that you already master a very similar system (maybe Norwegians can transfer their own pitch accent to Japanese relatively easily?)
“The fact that I specialize in Psychology doesn’t mean I am not interested in other aspects. […] This is a very disingenuous argument on your part.”
When it comes to SLA research (and for 14 months I’ve researched this stuff almost full-time because I’m in the middle of writing a book about it) you give the impression that you’re “not” up-to-date and that you have only “cursory” knowledge. But, I might be wrong…
However: If SLA is one of your academic specialities then it should be easy for you to point me to some of your current research projects about SLA or Psychology of Language (research papers, books, whatever).
Your are a specialist in ML, too?
Nice. I have also an academic background in computer science (besides French and the Social Sciences, esp. Deconstruction and Social Complexity Research).
And I’ve been working on an ML-based startup regarding SLA for intermediate / advanced learners for some time now.
Again, just show me your ML research, please.
“Fossilization”:
Well, as I’m writing a book about SLA (as a part of my startup project). I’ll see what I can do in this regard.
“Krashen”:
He is great, but his research has been discussed very “controversially” in the SLA community… And “corrective” feedback is more and more seen as an integral part of SLA So, it’s “not” acquisition alone without any learning / feedback aspects anymore.
And the (comprehensible) “input” thesis is implausible after the transmission (sender-receiver) model has collapsed since the late 1970s / 1980s (dito for concepts such as “social mind”, “collective consciousness”, “interpersonal” communication or “intersubjectivity”. These are just metaphors or, in a literal sense, “nonsense”. In other words. No one knows what this “in-between”-dimension means!).
But, if you have never heard neither of Jacques Derrida / Ernesto Laclau nor of Niklas Luhmann and other social systems research, well…
Anyway, this doesn’t mean that “immersion” is useless. It just means that we have to conceptualize it in a different way. (with different concepts, theories, methods, etc.).
OK, I’ve just checked your website on Google myself:; SetShift - Grupo de investigación de la Universidad de Granada
As far as I can see, you’ve / your research unit has specialized in “cognitive psychology”.
Yes, you might have an interest in SLA, communication, ML, etc., but this seems to be more like your hobbies, not some serious “research” projects from your part.
Thanks for the discussion!
It was interesting - in more than one way…
Oh, my god! You’ve gone from a semblance of argumentation to bringing up credentials and hand-waving research to full “to hell with arguments, I’m the official expert and that is!!!”. Sure you’re an academic at heart. And that just when I presented
I shouldn’t comply with such a rude and unreasoned demand but, just to show that not everybody is like you, I’ll give you some info:
a) I don’t research in SLA, I research in Psychology, mostly cognitive processes and emotion. As I already explained my relation with SLA is in teaching, as part of the subject of Psychology of Language. I don’t claim to be a specialist but I am familiar with the literature, what you believe or not is of no concern to me. My research is easy to find by my name: Francisco Tornay.
Anyway, if you really have a leg to stand on, you don’t need someone who knows every paper and much less fall back to the “I am an expert” card. Real experts argument and provide concrete data, they don’t content themselves with waving around their credentials. Confession for confession, my impression about you is that you’re the typical cargo-cult academic (Cargo cult science - Wikipedia), obsessed with heaping up papers and followig ivory-tower theories, with zero argumentative skill.
b) Again, I don’t research in ML, nor did I claim it. I use it as part of my research, to analyze Psychological data, these days mostly thermographical imaging. Again, you can find my research easily on line. I also use it to help in socially-relevant projects, which I’m sure you disapprove of, since all you care for seems to be diplomas and empty publications. Again, you can also find my current project easily.
Oh! And the comprehensible input method that many here are using with great success to learn languages is “implausible” because some post-modernist guru put together a lot of half-comprehensible sentences a couple decades ago!!!
Derrida is your great argument, how is that for empirical science? He deconstructs with words and to hell with the rest.
I thought that the Sokal affair had had put this type of nonsense to rest years ago.
@ericb100
I’ll be recognized as a foreigner in Japan just by looking at me =)
Ha ha ha, good point! That’s why you should use an avatar à la “MattvsJapan”
Frankly, there’s nothing wrong with having a foreign accent.
I agree. And even natives have different kinds of accents (so a natural language is “not” homogeneous"). There’s nothing wrong with that.
The “real” problem starts when a bad pronunciation makes communication breakdowns more likely.
I think there are diminishing returns. i.e. the amount of time to get it perfect is not worth the effort beyond a certain point.
Of course. But “focusing” on pronunciation and trying to “perfect” it so that you’re completely indistinguishable from a native are two different ball games.
Have a nice day in IT!
@ftornay
I’m “the typical cargo-cult academic”?
Hm, it’s really strange when an “established” researcher says that about (most of?) his colleagues, don’t you think?
(Social) complexity research:
I’m refering mainly to “social systems research” (Niklas Luhmann, Dirk Baecker, Peter Fuchs, Armin Nassehi, etc.) and Complexity Adaptive / Dynamics Systems Theory in this context.
See for example:
http://chkjournal.com/
See esp. Diane Larsen-Freeman reg. SLA.
Or, from a linguistics perspective, the new book by H.-J. Schmid (2020), “The Dynamics of the Linguistic System, Usage, Conventionalization, and Entrenchment” (for a short review see:
https://global.oup.com/academic/product/the-dynamics-of-the-linguistic-system-9780198814771?cc=de&lang=en&)
But, I suppose you will never read that because you either know already “all” the relevant SLA research or you don’t need to read it because it’s just some “academic cargo-cult” to you.
Postmodernism:
Derrida has developed “strategies” for deconstructing distinction-based meaning structures. This is neither a scientific method nor a scientific theory - and definitely “not” empirical research.
But, it’s still relevant because of his emphasis on “distinctions” to construct something as something.
If you prefer a more “mathematical” approach to distinctions, see “The Laws of Form” (George Spencer Brown, Dirk Baecker / Niklas Luhmann, Louis H. Kauffman, etc.) Laws of Form - Wikipedia).
But, again: I suppose you will never read that either because you already “know” all that (-> Sokal).
Experts vs “real” experts:
From my academic experience “real” experts know their “limits”. And for researchers in computer science that doesn’t seem to be a problem at all.
But it seems to be sometimes a problem for researchers in the humanities and the social sciences…
Your social projects
Why should I “disapprove” of your social engagement? Ah, I forgot: This is what “cargo-cult academics” seem to do - at least in your academic environment.
One last point: I “really” like your humor!
When developing an ML-based startup “from scratch” (and I’m talking about “programming” here, not some hackathon or Wordpress-clicking) is a characteristic of a typical “cargo-cult academic”… well, you made my day!
Have a great time in Granada…
After a year of learning German here at LingQ, I started Duolingo. I know all the vocab and what I essentially train there - is a grammar, basic structures that Duolingo helps me to internalize.