Acoustic Discreteness vs. Continuity in Production
|Submitter Email:||click here to access email|
Hello. I have two related questions:
1. No one doubts that phonemes are discrete. They are perceived
categorically, for example. But is it known whether their pronunciation is
discrete ACOUSTICALLY? That is, imagining an acoustic 'space' - I don't
know how many dimensions - maybe height of formant 1, height of formant 2,
amount of fricative noise, etc? - how much overlap is there? To keep it
simple, assume I'm asking about a single speaker:. I presume there is at
least some overlap, but is it substantial?
2. The above question was really to set the context for my main question,
which is about intonation. To many people, intonation at least seems to
vary continuously. I realize that there are theories (e.g. Pierrehumbert's)
which claim there are discrete tones (H, L, evidently M for some) and
discrete positions for them (H*, H-, H%, etc.) and further constellations
thereof ('surprise-redundancy', 'contrast-incredulity', etc.), but: do
their ACOUSTIC profiles 'clump' ( in the productions of a given speaker) to
the same extent as what I presume is found for segmental phonemes? Does it
do so at all?
References which show this?
Thanks for any help you can provide!
Sums main page