Re: A gripping language, and a question about suprasegmental analysis (WAS: re: conlanging partners)

Below L> is Leland; S> is Sai.

L>Looking below, it seems to me that the dominant person always has the more
L>flexible position, phonologically speaking, and so my temptation is to build
L>in the asymmetry as a discourse element: The person with the dominant (thumb
L>outside) grip speaks, and turns are taken. (I find this slightly
L>unsatisfying, but it seems most practical.)

Exactly my initial thoughts too.  The asymmetry really bugs me, and this
would be a nice way to throw it out.  But as Sai points out, it's a whole
lot of trouble to switch dominance every time the turn switches.

From the point of view of usability, incidentally, I expect it would be very
difficult to learn to use this language in both dominances, subject to
dominance not being for turntaking.  It's impossible to keep both which
finger is used to convey foo and which spot on the hand is pressed to mean
foo constant as dominance varies, which ouch brain breaky.

L>Of course, each of these
L>places it's own limitations on the movements possible at that point:

Unless the thumb dispositions are subtle enough that you can temporarily
leave them to make a motion with the thumb.  Which they may perhaps be, if
they're signalled by just setting a dom finger at the tip of the sub thumb.

S> * disposition transitions: short-short, short-stroke, or stroke-*  (I
S> found stroke-stroke and stroke-short to be too hard to reliably do
S> differently)

No, no.  It was the distinction between stroking and not stroking _on the
inside_ which you couldn't reliably distinguish.  Stroking along the outside
of a thumb is easily distinguishable, and could be a phone by itself without
switch of thumb disposition, modulo its being visible.

L>It strikes me that the first three of these have a markedness order: Knuckle
L>  > short gap > gap, in that knuckle press requires lifting and twisting the
L>finger and short gap requires lifting.

S>To me knuckle press only requires shifting the finger slightly, no
S>retraction; short gap requires full retraction; and gap is default
S>(depending on degree of grip torsion).

Yeah, I too find knuckle and short gap presses to be not quite orthogonal
motions, but close, so that the markedness order isn't so clear.  (It's the
lowest of the three articulations of the finger we mean for the knuckle
press, not the middle, in case that was some confusion.)

L> Also, what about lifts of various fingers, as movements unto themselves?

S>Possible, but seems to be liable to cause perception issues, in that what is
S>perceived is a *lack* of baseline sensation, rather than a *sensation* per
S>se. Also, this may be confused w/ movements that are needed as part of
S>executing something else (e.g. a short gap press).

I'm not sure; it may be that if you lift your fingers I can feel the motion
in the muscles and tendons of your hand.  Something to try, I think.

S>Hm. Palm press / lift seems likely to rely significantly on finger-join
S>leverage. Would this cause noise?

Noise as in interference?  With what?

L> In cases where no disposition change
L> would be necessary, you could add a movement derived from the disposition
L> change: So a tap on the tip of the sub thumb in "thumb-up" and "index-up"
L> positions, etc., though differentiating short and stroke changes then
L> becomes tricky.)

S>This part I don't understand. Elaborate?

Let me give this a shot.  Suppose capital letters are state 1 motions and
lowercase letters are state 2 motions.  Further suppose the end of the
transition from state 1 to state 2 feels something like a |z|, and the end
of the transition from state 2 to state 1 feels like a |Z|.  Then one could
declare that all words in state 1, resp. 2, gain an initial epenthetic |Z|,
resp. |z|, to mark the word boundary.

This would work, if the phonetics allow it.  At this point I have the
feeling pause for word boundary is likely to work fairly well, though.
(Whether this is efficacious enough I suppose would be determined by the
shape of a word.  Is there inflection?  &c.)

L>Certain of the movements listed above can be superimposed, while others
L>can't. So a press and a squeeze cannot (fluently) be done at once; I also
L>feel as though (for instance) 3rd and 4th gap presses at once would be hard
L>to distinguisgh from just one or the other, while 1st and 4th together would
L>be highly distinct.

I can only guess which distinctions it's possible to learn.  Presumably it's
very possible to learn the difference between all the knuckles and all the
gaps when pressed individually, for instance, but even this I can't reliably
do at present.

L>I see two routes here: Either we should ennumerate the
L>possible movement combinations as phonemes unto themselves, or allow them as
L>allophonic variations conditioned by quick speech (contractions, more or
L>less). I see benefits and disadvantages to both sides.

Or take a middle way.  Or, as Sai points out, factor things even further; or
swing the other way and group things even coarser (I don't know, pretend
that explicit phonetic sequences are one phoneme?  Probably not sensible,
but I wonder if it couldn't help the listener in certain ways.  Compare the
fact that identifying exact pitches is hard, but discrimination of the
interval between successive notes isn't so much).

But certainly a good question to gnaw on.

Alex

Reply