With the tools from grammar and IPA etc., learning a language shouldn't require the old-school learn-by-example or learn-by-immersion approach. Instead, learning it "the hard way" - diving straight into theory - should be more efficient for people like me.

I need to point out early in this post, that surely pratical language usage exercise is necessary - just that the theory behind "why the sentences are the way they are" deserves explanation before rather than after they have been learnt.

What I mean by "analytic":

Go straight to questions like:

  • How does sentence structure change? How do I ask a WH-question sentence? How about a yes-no question? Rhetorical question?
  • How do verbs change? How many tenses are there? What do each of them mean?
  • How do nouns change? Is there a distinction between singular and plural? What signifies it?
  • How do pronouns change? How many different persons are distinguished? What does each of them mean?
  • How do adjectives change?

It even works for pronunciation:

  • What is a sound in a language that I already know that sounds exactly the same as the sound you want?
  • What is a word that has only the first vowel different from the word we were talking about?
  • How can I slightly modify one of my already-learnt sounds to get that sound? (e.g. round a /i/ into a /y/)

And when given a currently-intuitively-weird sentence, e.g. "Il n'y en a pas", "J'en ai marre", "Qu'est-ce que c'est", "Comment s'appelle-t-il", to care more about asking questions like "Why is this sentence structured this way? What does each part mean? Are there other sentences that are structured similarly (or demonstrate the same weird adjective behaviour)?" instead of caring to "simply learn that one sentence".

In short, this is about adopting a mathematician-like mindset of how to learn: find a general solution, not a specific one. Deal with the issue once and for all.