Re: ReTonogenesis

On Wed, Feb 02, 2005 at 10:21:40PM -0000, caeruleancentaur wrote:
> Would someone explain to me the difference between pitch-accent and
> tone-accent, if there is a difference?
IIRC, pitch-accent means that you have different pitch level phonemes
(e.g. high, mid, low). Tone-accent means that the phonemes in question
are pitch *contours* rather than mere pitch levels (e.g. Mandarin
mid/high level, low rising, low falling, high falling).


> I have made Senyecan a pitch- accent language, I think.  I envision
> a person "singing" on three whole steps on the piano keyboard, e.g.,
> C-D-E.  The base accent is "sung" on C, the secondary accent
> (indicated by a grave accent) on D, and the primary (indicated by an
> acute accent) on E.  The exact notes used will, of course, vary with
> each speaker.  As one listens to another speaker, he will discern
> what notes/pitches the other speaker uses in his idiolect.  I use
> the word "singing" in a loose sense.  There is no structure of whole
> notes, half notes, etc. Only the pitch, with comfortable duration on
> the syllables, is present.
[...]

That looks like a pitch-accent system, as opposed to a tone-accent
system.

Also, having the three pitches on C-D-E seems a bit too close to me. I
don't know if there are natlangs to prove me wrong, but IME a more
naturalistic layout might involve a mix of differing intervals rather
than two equal whole steps.

For example, my L1 Hokkien, although it's actually tonal, organizes
its tones around the pitches low F, low F#, C, D. The exact pitch
doesn't matter, of course, what matters is the fact that the two low
reference points are roughly a 5th to a 4th lower than the mid pitch C
(which is the level of the mid/high-level tone), whereas the high
rising tone centers around D, only a whole step above what is
perceived as "mid level". The difference between the low F and low F#
is indistinct to my ear, at least when it comes to language perception
(musically I can hear the difference, all right), and may be the same
for all practical purposes. I give F# as a rough estimate; it may be
F# and G instead of F and F# for some speakers. The point is that
they're at a distance of about a 4th to a 5th below "mid pitch".

Another natlang example is Classical Greek, which is said to be
pitch-accented. The distance between the high and low pitch (I'm not
sure if that refers to the distance between acute and grave, or in the
supposed low-high-low glide of the circumflex) is supposedly a perfect
5th. If this is correct, the Classical Greek grave and acute would be
C and G, respectively, with the circumflex being something like D-G-C
or C-G-D.

Other natlangs I have observed include Korean, where the high/low
pitch difference in fluent speech appears to be roughly the distance
of a perfect 5th or maybe a diminished 5th or a 4th (similar to the
difference between low and mid level in my L1 as I described above,
albeit here the difference is between high and low). So again,
something like F for mid-level, G for high pitch just before the end
of a sentence, and C for the end of the sentence. (I don't know Korean
grammar, so this is just a rough qualitative description of a typical
Korean utterance that I hear.) Some Korean phrases I picked up also
involve a pitch change of about a 3rd, perhaps something like E-flat,
G, and C. Roughly speaking, of course. The actual pitch in speech will
certainly vary.

So while it is certainly not inconceivable to have a natlang with
pitch accent centered around C-D-E (and its transpositions thereof),
the small natlang sample I have seems to suggest a more probable
distribution involving a mix of 2nd's (whole tones) and 4th's or
5ths, perhaps with some 3rd's thrown in for good measure.

Just my slightly-more-than-2-cents worth from my non-professional
observations. Hope you find this helpful in some way.


T

--
"No, John. I want formats that are actually useful, rather than over-featured
megaliths that address all questions by piling on ridiculous internal links
in forms which are hideously over-complex." -- Simon St. Laurent on xml-dev