Take nobody's word for it: a case study

So last week I gave a few reasons why you shouldn't pay too much attention to other people's advice.  Of course, you do have to pay some attention to other people's advice, so you have to evaluate the soundness of that advice.

A few weeks ago I was alerted to a perfect example of bad advice via a post on the How To Learn Any Language forums.

What makes this such a good example is the fact that the guy giving the advice has kept a blog, which allows us to critically evaluate what he says.

Keith Lucas believes you can learn a language just by watching TV.  This flies in the face of a lot of opinion, experience and most theoretical models of human language.

A month ago, Keith completed a 2000 hour "silent period" of Chinese.  That is to say that over the course of 2 years, he has watched 2000 hours of Chinese TV without looking up any grammar or vocabulary.  He will only now start to speak the language.

Keith's inspiration is the Automatic Language Growth (ALG) method.  ALG is in essence a variation on the Direct and Natural methods of the late 19th century, which suggests that we can only learn a language through that language itself.  I was talking about "one all-important concept" last week, or in Decoo's words "a concept that is stressed above all others" -- this is ALG's all-important concept.  Decoo says that this concept has to appeal to the imagination, and this certainly does appeal to the imagination.

The earliest post I can find on his blog referring to ALG is from October 2008, but it doesn't tell us how he discovered ALG and the exact nature of his experience with it.  He then went on to talk very enthusiastically about the principles of ALG in several more posts over the course of that month.

Then we get the leap of logic.  That very same month, inspired by ALG, he decided that he would try to replicate the ALG classes by watching internet TV.  Of course while the principle of ALG is learn a language through itself, in practice, ALG uses a structured course to introduce grammar and vocabulary in a controlled manner.  As such, simply watching TV doesn't approximate the ALG method at all.  Well, for the beginner at least.  A well-experienced ALG teacher has said that TV is only a replacement for ALG when you can understand 55-70% of the language to start off with.  To quote the article exactly: "If you are a complete beginner, it won’t work. The TV would just become more noise."

There is more to a method than the all-important concept!

Anyhow, so why does Keith seem to have succeeded?  Is he lying?  Is the rest of the world wrong?  What?

Well, the clues are scattered around his blog.  On the day he started his simulated ALG he said: "I'm no longer going to try".  "No longer", he says.  OK...

Then it gets interesting:
"As I have already learned a little bit of the language, I hear many of the words that I have learned. I am experiencing first-hand the crippling effect of my learning. Whenever I hear something familiar, the meaning just won't come. I have to associate it with the English and then I understand the meaning. There is some kind of barrier."
The implication in here is that the previous learning is a hindrance, rather than a help.

Keith seems to be denying the usefulness of his previous study, and it's a pretty common thing to do.  I've lost count of the number of times I've seen a review for course X, Y or Z that says "I learned more in 5 days with course X than in 5 years of school/evening classes/whatever".

It is impossible to go back and start from zero with a blank slate.  Everything you've studied previously has some effect, and it seems to me that most immersive courses and methods are most successful for false beginners.  In general, it seems to me that most courses are incomplete, and immersion fills in the gaps, filling in the need for meaningful practice that a lot of non-immersive courses fail to provide.

People will tend to ascribe their success to one thing, but in reality, their success is the cumulative effect of everything they've done.

But, in fact, Keith actually knows this himself.  In 2009 he wrote about how to write a language method, and in that post he says that on the internet "you can find many examples of what people are doing or what they think they did when they successfully learned a language".  Notice that "what they think they did" -- the tacit implication is that (as I said last week) we don't always know what we're doing.  But despite knowing this, despite knowing that people are never fully aware of their own actions and ways of working, he keeps slipping into making definitive statements about how things work and about what he does.  And to be fair, so do I, which is what the title of last week's post was all about!  Being aware of common human failings does not make us immune to them.

So a general request: when giving advice on learning, don't only tell people the stuff you think worked.  Tell them everything you did, and tell them which bits you think worked and which bits you think didn't work, but tell them that you can't be sure that you're right.  Then they can evaluate the whole thing critically and decide what they want to do for themselves.