Hacker News new | past | comments | ask | show | jobs | submit login

> Ideally there is a one-to-one mapping between text and speech.

Ideally, yes, in the sense that if you were to create an artificial language, you would strive for that (at the beginning at least).

In practice, the speech->text direction isn't nearly as important as text->speech when learning a second language. If you aren't immersed in the environment of the target language, you have to rely on texts in order to acquire vocabulary. And a messy script like English makes that very hard (although not as hard as Japanese). I've been learning English for years and can pass an English proficiency test with a perfect score, but I still make basic mistakes in pronunciation.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: