Skip to content

Instantly share code, notes, and snippets.

@t9md
Created June 19, 2019 06:50
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save t9md/b639917c4086b04f0a916cec6921a5e9 to your computer and use it in GitHub Desktop.
Save t9md/b639917c4086b04f0a916cec6921a5e9 to your computer and use it in GitHub Desktop.
- リスニング能力は総合力。ここが本当の実力
- スピーキングは大分後回しで良い、勝手にレベルアップする(という理論、事例が多数あり)
- 何故かと言うと、自然に喋れるためには Grammar intuition と パターン蓄積が必須だが、これは大量Inputによって得られるから。
- 常にスピーキング能力はリスニング能力より下になる。これは当たり前の話。日本語でもそうだよね?
- パターン蓄積とは、特定の事言いたい場合の言い方の例を溜めること。 例: お腹空いた vs I am hungry
- なので、大量のリアルな音声インプットが必須
- このとき、50%しか、あるいは20%しか分からなかったとき、それで完璧主義/潔癖主義に陥らないこと。
- 曇った眼鏡でも楽しめるはず。そして続けていれば20%→21%になる。それはレンズの曇りが晴れてより楽しめるようになる。
- こういう姿勢で臨めば自分が興味がある分野の動画/音声を、"楽しんで毎日聞き続ける"(お勉強ではなく)ことが可能になる。
- 楽しめる→継続できる→慣れる→解像度が上がる(キャッチできる部分が増える)→より楽しめる→勝手に少しずつレベルアップする
- 毎日、自分に強いなくても継続できる、というのはメチャ重要。言語習得はめちゃ時間を要するので。近道はない。
- そのすべてのキーになる態度が、眼鏡が曇っていても景色を楽しめること。悪いスピーカーでも、音楽を楽しめること。
- ただ発音は個別にやったほうがいい。
Listening: Throughout 1 to 5, we extract abstract "concept/meaning" from pure wave of air vibration.
1. Ability to capture richness of raw sound
2. Parse by word based on vocabulary: tokenize
3. Semantic Analysis by grammar
4. Evaluate Meaning lead to understand
5. Whole process should be happened by intuitions
Speaking: Clarity/purity of information is reduced gradually from 1 to 3.
1. Having what you want to convey
2. Based on grammar intuition + vocab, express sentences
3. With sound using your articulator as musical instrument.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment