Yes, I realize that linguistics is largely about this sort of thing.
Nevertheless, it's interesting as to what actually can be rigorously defined as identical phonemes. It seems to be a largely arbitrary definition based on what people's culturally attuned ears have been educated to accept.
It is an immensely complex subject - of which I sadly know very little.
Speech recognition software still has considerable difficulty in simply distinguishing words from each other, for example.
http://vocaroo.com/i/s1Hp21xgOTg1
Gah!!!!
MakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstop
I need to find my microphone... I have always been told I have a weird accent (Being half German, half Guatemalan, and having grown up in the South)
Edit: I can't find it at the minute
Why, thanks!No you have a very agreeable voice, truly you do.
Oh, your voice is quite nice, Mr. B, if barely audible.Why, thanks!
But I was struck by the melodiousness of your voice. Otherwise I wouldn't have been curious to listen to other recordings.
http://vocaroo.com/i/s1Hp21xgOTg1
Gah!!!!
MakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstopMakeitstop
Yeah. I couldn't imagine trying to configure a speech recognition device to distinguish aspirated and unaspirated sounds, which are separate phonemes in Hindi. Designing an English recognition device capable of understanding the diverse dialectical differences in the language, including the elimination of a mid-low back vowel in American English (cot and caught are pronounced the same way here), and the heavy velarization of consonants in the Indian variant.