I am using natural language processing/speech recognition techniques so that I can provide better tools to learn English pronounciation.
While research on relevant topics, I found this fact:
Certain phonemes already well-articulated to reflect its pos position. For example, for the case of the suffix -ate
, there's a pattern like this
For Noun and adjective:
IPA(key): [...kət]
intricate, predicate, delicate
For Verb:
IPA(key): [...keɪt]
complicate, sophisticate
This pattern actually makes the computation process a lot faster because of this simple fact. I really glad to know this, and need to know any theoretical/academic names for this phenomenon.