[Martin Taylor 941007 17:30]
(Bill Powers 931007.1115)
Tomorrow Mary and I will be off on a four-day camping trip into
Canyonlands in Utah, so between tomorrow and Tuesday don't start
imagining things when you don't hear from me.
I wish I were going with you. A brief glimpse showed that it must be a
wonderful place. For me, I will be off on a five-day trip up to the
cottage on Manitoulin Island, returning Tuesday evening, so I'm not
really TOO jealous (unless there is a lot of wet snow). And don't
start imagining things because I don't reply either.
ยทยทยท
===============
Bill to Bruce Nevin (93107.1125 EDT)
There's apparently a bigger issue in this question of ECS insertion
than I had realized. I can't say whether Bill correctly interprets
Bruce, but he doesn't interpret Bruce the way I do, and his discussion
of "breaking down" perception versus "building up" perception seems
to have a lot of holes. I agree with Bruce, the way I read his postings,
but I reject being classed as one who breaks down perceptions into
lower-level entities that were already being used. As I see it, the
lower-level entities were not controlled perceptions, but were constructed
later in a form of reorganization.
The fact that a relationship between lower perceptions and higher ones
is LOGICALLY necessary does not mean that the lower perception ever
exists as an object of control. It may be only a hidden component of a
one-level PIF in an ECS that controls the "higher" perception.
Consider the relation between the sensation level and the
intensity level. If you look at a sensation like an edge, you can
"break it down" into different intensities on either side of the
edge. The intensities are of a different class from edgeness. But
this "breaking down" is a matter of redirecting your attention
among existing perceptions, not of creating perceptions.
Logically, an edge is based on an abrupt difference between the intensities
in two adjacent regions. Factually, the perception of an edge does not
require any difference in intensity between regions perceived to be
separated by a sharp edge. There are lots of illusions to attest to
this fact. However, let's ignore them and assume that the logical
statement is true.
We know that the retina contains at the very least cells that differentiate
the intensity field, providing to higher levels (still in the retina)
information only about the local gradient in a particular direction.
They are sometimes called "edge detectors." Now if "perceptual signal"
is defined as ANY signal in ANY axon, then Bill's statement is trivially
true (ignoring illusions). It would be logically possible to construct
an eye in which only signals from such edge detectors were available to
higher levels. A creature with such an eye could never control intensity.
It might infer that some such variable existed, and say/see that there
was more of it in one place than another. But it could not perceive
intensity. It could perceive edges with no problem.
If there were no intensities, there could be no way that a
perception of edgeness could exist.
An engineer's viewpoint truth, not a truth from the viewpoint of
the observer.
It is the sensation of
edgeness that is synthesized from a collection of intensity
signals, not the other way around.
Signals, yes. Controlled perceptual signals, not necessarily.
The same relation holds at every level.
And the same argument applies at every level.
Configurations depend for
their existence on a set of sensation-signals and perhaps
intensity signals.
But not on those signals being controlled perceptions.
Transition perceptions depend on a set of
configuration signals and lower. And _event perceptions depend
for their existence on the existence of signals representing
transitions, configurations, etc._.
Etc.
What I (and I think Bruce) are saying is that the controlled perceptions
of a baby's language start out with fairly large units based (an analyst's
view) on a complex perceptual input function, within which there may or
may not be modules that react to patterns the analyst might call "phonemes"
"syllables" "transitions" "configurations." The claim is that none of
those subsidiary signals are initially controlled, if they actually are
formed within the complex PIF, not that the logically necessary information
is unavailable.
The baby's early control of speech is very vague. Phonemes, as an adult
knows them, get all mixed up. Words have only a distant acoustic relation
to the words as an adult might say them. There is no indication that
there is good control of anything much between the gross "template" at an
event level and the vocal musculature. (My parents tell me that at some
stage I used to say what they report as "Einabein dutter" for "Bread and
Butter"; some similarity of overall pattern, but not a lot of phoneme
relationship. Formal studies have lots of better examples.) I'm talking
now about the stage at which babies are using voice to get differentiated
things, called the "holophrase" stage. That means that the baby uses
single acoustic groups, whether to an adult the group represents a part
of a word (ba for bath or ball) or many strung together words ("allgone" is
a common example). At this stage, there is no evidence that the baby uses
part or all of one holophrase in constructing another.
At the next major stage, the baby puts together two words, which can be
determined from the fact that one at least of the two words tends to
appear in a variety of vocal contexts with what seems to the adult to
be the same external referent (allgone ba, pussy ba, daddy ba). Even
then, the "same" vocalization can "mean" wildly different things (daddy
took my ball, daddy's home and I want to play with the ball, "daddy, that's
a ball") if we can reasonably infer the baby's intentions from its other
actions (a PCT no-no, but something parents do a lot).
What is happening here, I think, is that the baby has reorganized to control
for the perception of recurring different holophrases or their larger parts,
probably working both top-down, splitting out words and putting them
together in different ways, controlling word-level perceptions as well
as controlling phoneme or syllable-level perceptions (not using PCT
levels, but perceptions that formed a "logical"--as perceived by
an analyst--part of the holophrase). It is the control of the components
that is new, not the existence of the signals corresponding to the components.
HPCT says that the perceptual process begins with detailed
sensory stimulation that leads to a large, maximally
differentiated, set of intensity signals. A second level of input
functions receives subsets of these intensity signals, weights
them and adds the results together, and produces a new kind of
signals, sensation signals, that each represent the collective
state of intensity signals in some specific respect.
Nothing I or Bruce said should be seen as contradicting anything here.
So the perception of an event is _synthesized_ from sets of
lower-level signals, which are _synthesized_ from still lower
signals, until we reach the level of sensory stimulation. Without
the lower signals, the event perception could not exist. And only
those event perceptions exist which are derived from lower-level
signals by specific input functions.
Or this.
When we analyze a perception into "its components" we are really
backtracking the perceptual process to see where a specific
perception is coming from. As we attend to lower and lower levels
of perception, more and more detail, we are moving the locus of
awareness against the flow of information, toward the source. We
are looking toward the causes of perceptions, not the effects.
Higher perceptions are caused by lower ones, in a manner dictated
by the organization of the intervening input functions.
But if one takes "perception" here to mean "controlled perception,"
we can disagree. Backtracking and attending to lower and lower levels
of signal may allow the baby learn to control those signals, and thereby
turn them into perceptions.
...it is not possible that a higher-level perception could
exist before lower-level perceptions exist. It is not possible
that the perception of a "whole word event" could exist before
perceptions of "its components" exist, because the event is a
function of the components.
Again, don't mix "perception" with "signal." The argument is that
the "whole word event" may be (badly) controlled before the signals
that the analyst sees as its components can be controlled and turned
into perceptions.
You are having difficulties in reconciling contrast and other
phenomena with the idea that words are derived from syllables
and/or phonemes. This is because you are looking at the
perceptual process as one of analysis instead of synthesis.
I don't see that at all. What does it mean, to "reconcile" contrast
with anything? The analyst sees words as derived from syllables
(true in most or all languages) and/or phonemes (apparently not true
for speakers of languages written in a syllabary, or for illiterates).
What the speaker perceives is a subject for test, and I don't see
what it has to do with analysis or synthesis.
Martin