How to properly do L-R method, any help for my attempt appreciated.

[Cainntear]

When you refer to "perceive and disting sounds of English" what should be a method to improve it?

There is a type of practical phonetic training called ear training. I'm still learning about it myself. Both the original descriptions of L-R and the book, Fluent Forever, talk about listening to audio recordings of minimal pairs, but there's more to it than that.

Indeed. Minimal pairs have always been particularly popular in Japan, but I vaguely remember reading a study carried out in Japan with first year university students where the students who got minimal pair training in the first semester weren't any better at phoneme discrimination than people who hadn't... except when given the same type of minimal pair phoneme descrimination that they'd been practicing doing. (I believe the other half were given the minimal pairs practice in the second semester so that they weren't missing out even if there was any advantage.)

I.e. minimal pairs is something you can learn to do as an isolated task even without modifying your language model, by directing specific attention to the expected difference.

I've long held that sound discrimination has to start with making your brain understand that a certain sound distinction is meaningful and important. I believe that the best way to do this is to start trying to consciously make your pronunciation a reasonable approximation of a native voice. Once your brain knows roughly where all the phonemes are, it'll know that they're important and start to hear them.

[Le Baron]

Sound discrimination and even knowing the differences between words via the writing system isn't even foolproof for native speakers. It's what also causes etymological pollution or hybrid etymologies over time. A Spanish person once told me to look up a word (I can't remember the word) and I asked how to spell it. He didn't know if it started with a B or a V.

A good deal of the differences can be assisted by knowing the written language as well as the spoken. The only time I've ever gone through minimal pairs was in the Dutch course Levend Nederlands and truth be told some of them really seemed like they had no real sense of confusion with one another. So I think they also differ per person by perception.

Minimal pairs 'training' seems to me a bit overrated.

[Cainntear]

Sound discrimination and even knowing the differences between words via the writing system isn't even foolproof for native speakers. It's what also causes etymological pollution or hybrid etymologies over time. A Spanish person once told me to look up a word (I can't remember the word) and I asked how to spell it. He didn't know if it started with a B or a V.

Except that's because (AIUI) the two letters represent a single morpheme and only exist etymologically. The fact that you need to explicitly know the spelling in such a circumstance is kind of obvious. I believe Lev Vygotsky described writing as a "second-order" approximation of language, or something like that. Language is phonological, first and foremost, and writing is just a drawing of language.

[Le Baron]

Except that's because (AIUI) the two letters represent a single morpheme and only exist etymologically. The fact that you need to explicitly know the spelling in such a circumstance is kind of obvious. I believe Lev Vygotsky described writing as a "second-order" approximation of language, or something like that. Language is phonological, first and foremost, and writing is just a drawing of language.

That's why I said it only 'assists'. In any case it clears up sound confusions in many instances. Instead of relying on just aural differentiation, which is quite difficult for L2 learners.