Additionally, in English word-final stops (p, t, k) tend to be de-aspirated. The puff of air (aspiration) that characterizes those "voiceless" sounds isn't as audible word-finally, so p, t, k sound like , [d], [g], respectively, but they are not voiced. There is no vocal fold vibration. They are de-aspirated.
EX:
that he is pronounced, tha[d] he (Note, "t" de-aspirated)
Now, the pronoun "he" begins with an aspirated sound, [h], but when it follows a de-aspirated sound, it too becomes de-aspirated, which is why,
EX:
thathe is pronounced, tha[di:] (speech is a string of sounds)
Quote:
|
A foolish young fellow once astonished an old clergyman by boasting that he didn't believe in anything "he" couldn't see and understand
|
As for the second "he", notice its environment. Aspirated [h] doesn't follow a voiceless sound (p, t, k). It follows a "voiced" stop, g.
EX:
anything he is pronounced, anything [hi:] (Note, "g" is voiced)
Voiced [g] won't change the aspiration on [h], because the two don't share any similar characteristics. In fact, it's those differences which make them distinctive, and why speakers may make it a point to fully pronounce [h] in that environment.
Quote:
|
A foolish young fellow once astonished an old clergyman by boasting that he didn't believe in anything he couldn't see and understand
|
Voiceless [t] and voiceless [h] share voicelessness. So, if you have two voiceless sounds in a row, and the first one is de-aspirated, then it's very likely the second one will be de-aspirated, too. The reason being, it's inefficent for a system, be it an engine or the human speech organ, to de-aspirate then aspirate. It's more efficient to continue the de-aspiration if the environment is suitable.
All the best,