[From: Bruce Nevin (Thu 931028 14:03:33 EDT)]
( Tom Bourbon [931027.1739] )
Yes. And your modelling is without a doubt infinitely faster than mine.
( Bill Powers (931027.1100 MDT) ) --
you have defined in a systematic way
the observations that require a model to explain them. A diagram
like the one above, with any amount of added detail required to
make it fit the facts, is a statement of the problem, not of the
solution. The problem is, "What kind of perceptual function(s)
would produce phenomena that fit this description?"
Yes. This is precisely what I have been trying to do. I started out
here asking, hm, I wonder how all this applies to language. Then I got
answers like oh, that's easy. You have phoneme detectors, and word
detectors perceive word-events where each event is a short, familiar
sequence of phonemes--the phonemes are like letters, so this is like
recognizing word strings in ASCII. Each word is associated with a
category perception, and then the words are just understood or produced
naturally in the order that follows from the way our perceptions are
organized. Whoa! was my response. It's a bit more complicated than
that. Look, here are some descriptive facts about language. Now, what
kind of perceptual function(s) would produce phenomena that fit this
description?
I currently don't have the equipment to construct and test models, nor
the considerable time that it would require to do so if I did, though I
persist in maintaining the crazy, error-prone goal of acquiring both. I
don't have the experience with modelling to wrestle descriptive
generalizations into modellable metaphors, but I have kept trying. And I
keep bringing up descriptive generalizations because that's where it
starts and where science always must recur: "what kind of perceptual
function(s) would produce phenomena that fit this description?" I keep
restating the problem because hitherto it had not been noticed that there
was a problem.
I'm just showing how the metaphor changes when
you approach the same phenomenon from a different angle, and how
a different metaphor requires a different physical model. The
choice of metaphor is more or less arbitrary; all that is
required is that it contain valid analogies with the observed
phenomenon. The choice of models is dictated by the choice of
metaphors.
First, the observed phenomenon. Then a metaphor which might be apt or
inept, no way of knowing in advance. Then the attempt to model the
phenomenon according to the metaphor. Then recurrence to the observed
phenomenon. I offered a description of phenomena which had not been
observed here (and an explanation of why they had not). To communicate
the description I used a metaphor that it turns out is inept for
modelling. I freely give it up, so long as the descriptive data get
across. Then those in a position to do the modelling can create more apt
metaphors. Sounds great to me. I keep pushing on that goal of doing
some modelling myself, but I'm not there yet. (Thanks for the mouse
tip.)
I like the vector metaphor. Gotta bring in the time dimension for
diphthongs, however. The positions of I (pit) and eI (pate) in the
formant space are so close together because the latter exploits a third
dimension, a contour through time. One might propose that this could be
taken as the product of an E vector (pet) plus an I vector with lower
gain. The raising of the syllable nucleus (the vowel with high gain)
from E to phonetic e would follow from the "trajectory" between them, per
Martin's account of coarticulation effects. A problem with that is that
exaggeratedly slow speech does not lower the syllable nucleus (the first,
high-gain vowel) from e to E: we say "I said leeeeeiid" to clarify,
rather than "I said lEEEEiid" ("I said laid"). One would expect the
target to be attainable under those circumstances. So there really do
seem to be separate vectors for I and eI, which speakers and hearers are
able to control so close to one another because one of them is
diphthongized.
ยทยทยท
-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-
Now, a separate topic, not to be confused with the above.
When we observe how language
or other social phenomena come into existence, we see that people
adjust their ways of speaking, dancing, or whatever to the ways
that are already established in their societies.
The appearance is that social conventions are
active things that have some sort of formative influence on
individuals. But when we try to model social conventions -- that
is, to propose an organization of matter and energy that could
produce this phenomenon -- we find that there is no organization
of matter and energy at the social level that could have such an
effect.
Here are two notions:
(a) the ways of [behaving] that are already established in a society
(b) an organization of matter and energy at the social level
The first, (a), is said to exist, and people are said to adjust their
ways of acting to it. The second, (b), is said not to exist. But all
that I have ever claimed for (b) is stated in (a). An individual
learning the ways of her society perceives them as ways of behaving that
are established in her society: she perceives others' behavior as
instances of those established ways; she perceives cooperation with and
approval of herself when she perceives her own behavior as instances; she
perceives confusion, failure of cooperation, or disapproval of herself
when when she perceives her behavior as not instances; she perceives
confusion, failure of cooperation, or disapproval of others when when she
perceives their behavior as not instances, and indeed may perceive
herself as confused, or may find it difficult to cooperate with such
others, and may express disapproval of them. The individual reifies
these perceptions as complex environmental variables. They are artifacts
of human activity in her environment just as buildings, highways, and
traffic jams are artifacts of human activity in her environment. They do
not govern her, but she governs herself in respect to them--buildings,
traffic jams, and social norms alike. Social norms may be studied as
objects in the world just as buildings may be studied in architecture.
But unlike buildings, and more like traffic jams (and rings and arcs),
social norms exist *only* in the controlled perceptions and consequent
behavioral outputs of people. It is this that makes them suitable for
study in PCT. Nonetheless, they do have this dual aspect, which is not
found in tracking behavior (except in the developments that Tom is
pursuing): they exist not only as perceptions but also as CEVs.
For an individual in a society,
(a) the ways of [behaving] that are already established in that society
are perceived as
(b) an organization of matter and energy at the social level
even though in fact (and in a PCT account) they are constituted only of
the behavioral outputs of other individuals controlling like perceptions.
Their perception of (b) as a CEV is comparable, so far as they are
concerned, to their perception of a building or a traffic jam or a circle
of spectators (I wonder what's going on that they're so interested
in--think I'll go look too). The news from PCT that it is illusory is
as irrelevant to them as the news from physics that what they perceive as
the solid wall of a building is in fact mostly empty space. In the
mirror world of perceptions projected onto CEVs, a social norm exists, is
real, and cannot be ignored indefinitely while still controlling other
perceptions in the course of living a normal life.
Just to underscore a possible difference of view: I am strengthening your
(a) statement to say not only that social norms are involved in how
language comes into being (in evolution, and in the learning infant) but
also in how language is maintained in ever changing daily use.
-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-=+=-
( Rick Marken (931027.1100) ) --
Language is, indeed, a socially conventionalized set of controlled
perceptions
Hooray! We agree about that at last. Modelling that is another question.
<Gobs deleted here. Too much dust in the air.>
By the way, you might enjoy _The Great Eskimo Snow Hoax_, by Geoffrey
Pullum. He is a linguist, but don't let that put you off.
Bruce
bn@bbn.com