Re: How does singing interact with voiceless phonemes?
- You can get tones with no vocal folds/cords by tapping your own cheek
while opening your mouth with differente apertures. The sound is more
audible with rounded lips.
2013/7/25 Matthew George <matt.msg@...>:
> I am considering making my conlang's phonemes either all voiceless or all
> voiced - it seems nicely alien to lack vocal chords. Further, I remember
> reading suggestions that heavy emphasis on voicing makes words seem
> 'nastier', with examples given of how Tolkien made Orcish seem harsh and
> barbaric. So I was leaning towards voicelessness. But no chords means no
> singing, too, which is rather a shame; I like the idea of a language that
> is always sung.
> Which then made me wonder: in languages where phonemes are distinguished
> by voicing, how do we recognize them in sung speech? Human singing
> necessarily involves voicing to create a tone. Do we simply perceive the
> phonemes that make sense in context (the way we don't thinking about the
> last sound in "dogs" being /z/), or are degrees of volume or similar
> features permitted to vary by singers to distinguish between phonemes?
> Matt G.
- On Thu, 25 Jul 2013 15:55:16 -0400, Matthew George <matt.msg@...> wrote:
>Which then made me wonder: in languages where phonemes are distinguishedOthers have already answered "exactly the same way as we do it in speech", but one basic fact that I think it's worth making totally explicit is that speech also has tones. You know this: even English sentences have intonation, and tonal languages exist. Yet we can recognise phonation in ordinary speech. Singing is a red herring.
>by voicing, how do we recognize them in sung speech? Human singing
>necessarily involves voicing to create a tone. Do we simply perceive the
>phonemes that make sense in context (the way we don't thinking about the
>last sound in "dogs" being /z/), or are degrees of volume or similar
>features permitted to vary by singers to distinguish between phonemes?
On the level of phonetics, the pitch (i.e. tone) of a speech sound is the frequency with which the vocal cords are vibrating -- that is, the frequency of the voicing. So, at the basic implementational level, voiceless sounds have _no_ pitch, whereas voiced ones have a pitch, and thus they are distinguished.
Of course, the human perceptual system patches over the little pitchless gaps when e.g. listening to a solo vocalist and extracting the melody. So the fact that a good percentage of the speech stream is voiced is enough to carry the song.
(Any sound that occurs during periods of non-voicing will just be noise which is all smeared out in the frequency domain. Things like place of articulation of voiceless stops, which are in significant terms silent in and of themselves, are principally perceived by means of the frequency bending they cause to the formants of nearby voiced sounds.)
In any given language there may well be other cues, e.g. English stop phonation being mostly aspiration; and context will help; but that doesn't change the basic form of the answer.