[From: Bruce Nevin (Fri 931029 12:06:30 EDT)]
( Bill Powers (931026.1230 MDT) ) --
I am still trying to get in touch with Keith Johnson. I have just
written a query to John Ohala. I think these guys probably studied with
him--it's his kind of work. In any case, he'll know who they are for
sure and may know how to reach them.
In my note to John (who may remember me from 20 years ago at UC Berkeley)
I quoted your excellent suggestions for tightening the results by
introducing a disturbance to continuous control of F1/F2 coordinates.
I sketched some basics of PCT briefly, as pertinent to the immediate
issues. Mentioned his article on hypercorrection and hypocorrection as
origins of sound change (which Martin had sent me)--I think you will like
this one very much too, did Martin send it to you?
We'll see what comes of it.
I think you're on target (so to speak) with kinesthetic perception. I
had persuaded myself that acoustic perceptions predominate in vowels,
kinesthetic in consonants. But recently I tried something else. I can
imagine producing a vowel sound as I move my mouth appropriately for that
sound. If I do this for some time, for various vowels all over the map,
I have very clear kinesthetic impressions, even though my attention as it
were is on the imagined sound. Then when at intervals I vocalize so that
the vowel is audible, the vowel is always right on the intended target,
with no apparent shift of articulators and no audible error being
corrected. This suggests the odd circumstance of controlling one kind of
perception (kinesthetic) even while attention is focussed on another kind
(acoustic).
I wonder if I could persuade my dentist to anesthetize my tongue for the
sake of a psychoacoustic experiment. My recollection of novocaine is
that the affected parts of the mouth feel puffy and stiffened. That may
be trauma related to the dental work, or it may be genuinely a reaction
to the novocaine. To be appropriate for the experiment it would have to
cut off kinesthetic perception without disturbing muscle capacity, tongue
flexibility, etc. I'll bet that I would not be able to articulate
at all clearly in the absence kinesthetic and touch feedback, whereas my
(flawed) experiments reducing acoustic feedback had only slight effect on
control of articulation. This would fit with my experience with
articulating imagined sound interspersed with real sound.
Could do the interspersing of real sound by using an artificial
larynx--placing a buzzer on the throat. With a moderate amount of white
noise in headphones, the vowel formants of egressive air "whispering"
through the mouth would be inaudible, but the formants with the buzzer
switched (unpredictably) on would be audible. Maybe I can set that up
and try it out when I get back next week. (Off to help my dad cut wood
and put stuff away for the winter--he's not been feeling well.)
The attention being focussed on acoustic perception makes good sense,
actually, given the social function of language. When I hear what you
say I have to go from the acoustic information, which is ambiguous in the
specific sense that more than one articulatory configuration can have the
same acoustic consequences. (This is the basis for the hypocorrection
described in the Ohala article.)
Sorry I can't expand that, but I've got to quit and re-do three volumes I
printed yesterday.
···
-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-
Rick Marken
( Rick Marken (931028.1500) ) --
You might be right. In any case, I'm glad I haven't said anything
challenging. That would be upsetting.
-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-
Just got a reply from John Ohala, as follows:
Date: Fri, 29 Oct 93 08:41:53 -0700
From: John Ohala <ohala@cogsci.berkeley.edu>
Message-Id: <9310291541.AA28058@cogsci.Berkeley.EDU>
To: bnevin@BBN.COM, userohal@mts.ucs.ualberta.ca
Subject: Re: hyprarticulation data
Cc: bn@BBN.COM
Dear Bruce,
Thanks for the interesting angle on the Johnson et al. paper -- which
I also enjoyed. Keith Johnson is currently at Ohio State Univ. (or
should be by now). I don't have an e-mail address for him but
you could try something like 'postermaster@linguistics.ohio-state.edu'
and see what happens (or leave off 'linguistics'). Mary Beckman
is there (ASA has her addr as beckman@shs.ohio-state.edu; possibly
the 'shs' has been replaced by 'linguistics' now).
Richard Wright is a graduate student at UCLA. I don't have his
address handy but I will try to find it and send it to you.
John
+ John Ohala, Professor +
+ OFFICE HOME +
+ Dept of Linguistics 1149 Hillview Road +
+ Univ of California Berkeley, CA 94708 +
+ Berkeley, CA 94720 Tel: 510 841 7134 +
+ Tel: 510 643 7703 FAX: 510 649 0776 +
+ FAX: 510 643 5688 +
+ e-mail: ohala@cogsci.berkeley.edu +
I'll follow this up.
Bruce