If you wanted to design a better beginner course than anything out there, what would you try?

[garyb]

I think that the cobble-together approach is here to stay, since “beginner to B1” covers a lot of different stages and the needs and best methods at these stages can be different. I agree on all the points about why Assimil is great for preparing learners for B levels, however I also think it's a terrible choice for an absolute beginner since it introduces so much new language so quickly with only brief explanations, and I dislike how it pretends to be a beginner course by starting with a week of slow lessons.

I'm a big fan of the Michel Thomas style for a true beginner: focusing on teaching basic structure first, introducing new vocabulary gradually, repeating and testing new knowledge frequently, and giving the satisfaction of producing language right from the start. DuoLingo also does that sort of thing, but IMO with too much new vocab and not enough explanation of structure.

But I also think that that style doesn't necessarily “scale” well: after these first eight hours of the Foundation course, it either starts to feel tedious (see: Pimsleur for “easier” languages), or keeps introducing more advanced structures at a point where the learner would be better off focusing on more important things like vocabulary (see: some of the MT advanced courses), or it becomes too hard to follow and retain (see: Language Transfer, or Pimsleur for “harder” languages). At that point I'd say it's still far too early for Assimil, but it can be a good time to start a more traditional textbook which will feel much less overwhelming with that base.

Perhaps machine learning and speech recognition could improve on the MT template, by responding to the learner’s mistakes, avoiding the annoying-student factor, and maybe even making it useful for a longer period of time.

For reaching B1, I'd love an Assimil "With Ease part 2" that just continues in the same way, rather than the "Using"/"Perfectionnement" books which go too much into advanced/niche language like journalism, literature, and politics. And more grammar and exercises, and some more technology use wouldn't hurt. So getting like what Cavesa describes.

[Cainntear]

In high school, my first French teacher asked for class feedback. I became aware that I couldn't really remember everything we'd done, and the only things I could remember were the things that weren't all that effective. My classmates would give feedback about how much they enjoyed the only things I could really remember... which were games that didn't really teach much -- they remembered the things that didn't help.

How did you know what wasn't effective? It's the flip side of knowing what is effective. I do get the point though, that any given learner might not really know what 'worked', but the reason I'd ask them is to find out what they think worked. It might possibly be what kept them motivated and that's half the battle.

A fair point to challenge on, but I can answer it!

I believe my thinking was that I could do the end-of-lesson "fun" activities very easily, because I'd already learned the target language pretty well in the bulk-of-lesson "boring" activities. The way I saw it, the other kids were enjoying the task structure and therefore commenting on it, but they weren't being actually taught anything new during the session.

Of course, you've made me stop and question my preconception, and now I'm thinking that those tasks may have been useful in that it put correction into a fun thing -- a pupil who lost a point on their error could well be more motivated to listen to the next pupil's answer to "make them better at the game". I don't imagine this would have been too strong really, though, as they were never really going to play the same game again.

So thinking about it has made me realise that it may well be useful if the lesson before it was too hard, so basically it was a sticking plaster at best.

But yes, being challenged on it made me stop and think, which is always a good thing!

Is there such a thing as a learning style? There's never really been any good quantifiable research that actually found anything valuable.

My personal view of the "it works for this group A but didn't work for that group B" is that group A is tolerant to something being missing that group B genuinely needs in order to learn, but generally, group A won't do worse if that hole is filled...

I think there are different ways people prefer to learn. Which doesn't mean they aren't doing the same thing at bottom - which they obviously are because the basic ingredients are the same for anyone learning a language - but there are different ways people are motivated and how they want to approach the task. Some people are happy with a fleeting grammar explanation and want to get right into use. Some want more grammar and time to process what is going on. They're doing the same thing in a different way.

But the way I see it, if they're all doing the same thing at the bottom, that means that there's something at the top that is superfluous to some or all of them -- hence my favourite word recently being "superficial".

Take for example the grammar thing. You compare "fleeting grammar explanation" to "more grammar and time". To me, this points towards differing problems in understanding the grammar explanations being given. Grammar explanations are often far too technical and can be as confusing as clarifying, so we should be striving to eliminate the difficulty.

I'm sure I've said before that I agree with MT that the common ways of describing what "nouns", "verbs" and "adjectives" are overly technical and of no real, immediate help; but that I don't really agree with the alternative descriptions he came with because they are technically far wrong, and only worked for his students because they evoked the right images when given, and they couldn't look back at them to fixate on the explanation. Unfortunately, the CD format of the home courses means people can relisten to them and memorise his explanations.

Instead, when I'm teaching the concepts, I'll write a sentence like "I have a ____" on the board, ask for a number of words to fill in the gap and say "we call them nouns," then write "I have a ____ house"... "we call them adjectives" and "I ____ it". This is a particularly good one, because you might get verbs in various tenses, but they're all verbs anyway.
My point with this is that everybody already has well-formed unconscious structures dealing with grammatical categories in their own language, and we've just got to evoke that structure and put a name on it -- grammar explanations shouldn't be presented in a way that they're teaching a learner something they don't already know if they genuinely already do know it!

[grayson]

I'd say the best combination would be: [...]

What you described there brought Rocket Languages straight to mind. I really like their approach -- except for the fact the audio is SO ANNOYING. Boring audio is a problem I've had with pretty much every learning course; maybe I'm unusual, but I'd rather be learning something interesting about science or history or whatever than listening to people run through canned dialogues about everyday things or shoot the breeze podcast-style. But Rocket Languages has inadvertently made this even worse for me through their audio combo of a non-native-but-fluent host (usually British or American) plus a native speaker. I really just abhor the hosts; everything they say feels fakely enthusiastic and their very voices grate on me. (That said, I've only tried Mandarin and Spanish, so maybe those are just particularly egregious.)

[grayson]

In an ideal world, where I had the money and the time, I'd learn languages by taking intensive immersive courses -- the kind where you spend every hour of every day learning and practicing in an environment that is not your home/work/daily grind. Where to get anything -- a spoon, more toilet paper, directions to the center of town, an answer to your question -- you have to ask in the target language. Where using your shitty forty-two words plus hand gestures is all you have, until you learn more; combined with hours and hours of input (grammar, vocabulary, explanations, cultural and historical info) from one or more knowledgeable guides (who are not just "fluent" but well-spoken, educated, and worth emulating in style and pronunciation) in the company of other people at a similar starting level and, ideally, with enough in common that you enjoy having them on the ride with you.

[Cainntear]

My personal view of the "it works for this group A but didn't work for that group B" is that group A is tolerant to something being missing that group B genuinely needs in order to learn, but generally, group A won't do worse if that hole is filled...

I think personal peculiarities do play some role. For example, I personally hate group audio-lingual drills with the passion of a thousand burning suns. They're not necessarily bad! But I won't personally go near them.

Ah but here's the way I see it: those pecularities are less about what type of good teaching works for us and more about what sort of suboptimal teaching we can put up with. A heck of a lot of audiolingual drills can be done mechanically, and I have to fight with my own brain to force it to engage when it knows that there's an easier way of accomplishing the task that has been set.

And similarly, I'm personally not too fond of textbooks with pages of theory and a few sentences in the target language. I've actually seen this done well! But I want generous amounts of the target language, and I'm happy post-poning grammar until I have thoroughly internalized a bunch of examples.

As I said in response to Le Baron: this is really about how the grammar explanations are suboptimal. Quick explanations should be enough, because we already know grammar, in reality. I can often work through the problems of clarity vs ambiguity in grammar descriptions, but many can't. In fact, many will get so confused by it that they just freeze up and stop learning.

I think these are differences in personal taste, and nothing as formal as a "learning style." But there's value in actually enjoying the process. So it might help to have a gut feeling for common preferences, so that an author could design a course that appealed to a relatively wide audience.

My point is unproven hypothesis, but it still stands: learning is enjoyable, and where learners are not enjoying the process, that's an indicator that there's a barrier to learning in there somewhere. But the opposite is not true: enjoyment doesn't, by itself, prove there's no barrier to learning -- it can instead simply be exercises acting like drugs.

Compare MT to Duolingo. Many people find MT enjoyable because it teaches a lot of language with little effort; but the many more people who seem to find Duolingo enjoyable are engaging in trying to achieve certain game goals which only interact incidentally with language learning goals.

One specific area where I've seen huge variation between students is how much grammar they want to see up front. Some people prefer a tidy framework to organize knowledge, and other people prefer to spend more time looking at concrete examples. Within reason, I suspect it's possible to accommodate both tastes.

I was pretty massively in favour of "serious" grammar learning before I used MT, but now I see the slow laborious thing of painful attempts at precision to be utterly unnecessary. As such, I know see it as potentially as much of a distraction from language learning goals as Duolingo's leagues and achievements.

Assimil, in particular, is too grammar-light for some people, and it just annoys them unnecessarily. Toss in something like a Dover Essential Grammar book and they're happy.

... which is pretty mad, because Assimil only pretends to be light on grammar. Just look at any Assimil book and answer me this:

• Is more paper devoted to dialogues or to grammar notes and explanations?

Then look at the size of type used for each of them and answer me this:

• Are the dialogues in bigger type than the notes or the other way round?

So then we can go back to the first question and reword it:

• In terms of word count and letter count, are the dialogues larger that the notes or smaller?

So, yeah... I've always thought that Assimil gives more space to grammar than to dialogues, but I've never measured it, so I might be wrong; but I just glanced at an Assimil book off my bookshelf and I think I'm probably right -- in particular, note that every pair of page in the dialogues has a parallel translation on the facing page, but the explanatory notes continue across the two pages. This means that if you have a dialogue that covers two-thirds of the target language page and two-thirds of the L1 translation page, the explanatory notes aren't just a third of a page, because a third of two pages totals two thirds of a page; and that's two-thirds of a page printed in smaller print and with smaller spacing.

[raoulhjo]

What I have complained about in the beginner courses is the choice of words. Too many words don't belong to the basic vocabulary but instead to specialised domains. I had to make my own frequency list of words.

[Granrey]

In my books the furure of this is on AI. A computer tutor plus an MMORPG game experience.

Since we are not quite there yet. These are the things I would recomend.

If money is a problem, I would start with free aps, free teaching podcasts and free books out there.

Duolingo with unlimited hearts is free if you create or join a classroom.

There are lots of books available for free in the internet. you can even use the internet public library to get books for free.

There are free podcasts that resemble a classroom.

There are free aps out there that allows you to have double closed captioning in videos.

Now, if money is not a problem. The only thing I would add to above is the tutor. If you can have a native tutor to help you along the way, everything will be quicker and easier. In my opinion, if big cash is to be spent, it should be with tutors.

[tangleweeds]

My ideal course would have what might look like massive redundancy of resources, and some kind of self-testing to help the user decide when to move on. I know that I enjoy learning languages very differently than many other people, and maybe even everyone learns very differently from everyone else (I, for example, need very explicit and rather detailed grammar, but I also read music theory and programming language specifications for fun, so go figure).

I definitely work better when there’s an ongoing plot to the dialogues, but given my theory that people should be able to skip around what they hate, it grows into something akin to a choose your own adventure. There would need to be main plotline dialogues (that work both as audio and video) but with side quests or subplots (with humor! and easter eggs!) that appear, say, in the reading section, or the optional additional comprehensible video sections, that might tempt the user into exploring those, but not break the main plotline if skipped.

To overcomplicate things more, there would need to be sufficient audio so that someone could do the course listening in their car (or lying on their sickbed in a darkened room like me), or but also tons of video for video-centric people. That would have to encompass both dialogue and grammar—note how there’s an entire Youtube sub-industry of videos explaining grammar from the Japanese textbook Genki!

And then some of us do well with exercises, while others would simply do better with more of the comprehensible storyline audio/video/reading. And some like self-correcting online exercises, others use pencil and paper exercises as part of their explicitly offline hours, and some of us even enjoy audio-lingual drills (good ones keep me on my toes and get my blood pumping!).

This is all becoming so complicated that I think I will need to become a billionaire to fund its creation as a charitable enterprise. And I hadn’t even made it to the multi-option SRS system yet!

(initially posted in the wrong place! there are downsides to doing forums with morning coffee)

[rdearman]

In an ideal world, where I had the money and the time, I'd learn languages by taking intensive immersive courses -- the kind where you spend every hour of every day learning and practicing in an environment that is not your home/work/daily grind. Where to get anything -- a spoon, more toilet paper, directions to the center of town, an answer to your question -- you have to ask in the target language. Where using your shitty forty-two words plus hand gestures is all you have, until you learn more; combined with hours and hours of input (grammar, vocabulary, explanations, cultural and historical info) from one or more knowledgeable guides (who are not just "fluent" but well-spoken, educated, and worth emulating in style and pronunciation) in the company of other people at a similar starting level and, ideally, with enough in common that you enjoy having them on the ride with you.

You've just described the French Foreign Legion.

[Cainntear]

There are free podcasts that resemble a classroom.

You don't even need a podcast for that -- you can get a big piece of paper and draw a picture that resembles a classroom. It'll miss all the important stuff that goes with actually being in an actual classroom, mind, but then so do podcasts!