Not that I am good at languages. Plenty of people are better. What I have is something slightly unusual: I did it deliberately differently each time, so I have some idea what actually changed the outcome and what did not.
Spanish
I did what everybody does: classes, exercises, vocabulary lists, tests. It went nowhere in particular. Then I started again on my own and did the opposite — I read, I listened, I consumed as much Spanish as I could stand.
That worked, and it worked far faster than the three years before it. It was the first time I suspected the problem had never been effort.
Chinese
I learned Chinese the traditional way, in a class — but at the same time I was immersing myself heavily, hours of listening, well beyond what the course asked for.
The difference was strange. Grammar explanations that should have been hard simply landed. I had heard the pattern hundreds of times before anyone gave it a name, so the explanation was a label for something I already recognised rather than a rule to memorise. I only had to be told once.
That is the clearest thing I know about grammar: it is cheap to learn after immersion and expensive before it.
French
For the first nine months of French I only read and listened. At the end of it I could not produce a sentence. Not my name, not a number, nothing. On paper, nine months of daily work for no visible output at all.
Then I started speaking, about an hour a day with a tutor, while continuing to listen and read. It moved very quickly — quickly enough that my tutor assumed I was shy rather than unable, because he had never met the version of me from a month earlier.
Over roughly eighteen months at around two hours a day, that is about 1,100 hours. Which is worth stating plainly, because it is the least impressive and most useful fact on this page: 1,100 hours is roughly what the US State Department budgets to train a diplomat in French. I did not beat the system. I hit the normal number, part‑time, while doing other things.
In total I have spent more of my life on Spanish than on French. My French is better. Same person, same stubbornness — different method.
Why I am writing any of this down
Every teacher in this field describes the method that worked for them, with total confidence, and is telling the truth. What almost nobody does is say which parts of their advice are supported by evidence, which parts are contested, and which parts are just what they personally did.
I am an engineer by training, which mostly means I find it uncomfortable to state something confidently without knowing how well it is measured. So the book has a chapter listing my own claims that the research does not support — including my favourite one, the nine months of silence.
And you cannot check any of the above. Every story on this page is unverifiable, which is exactly the problem with this entire field. I have chosen not to solve it with testimonials, because testimonials can now be manufactured by anyone in an afternoon.
So please do not weight my history heavily. Weigh the arguments and the numbers — those you can check yourself, and where I have got them wrong, you will be able to tell.