[From Bill Powers (940216.1015 MST)]
Bill Leach (940214.1902) --
Our "subject" can not change "SO's" perception (short of
assault). All that our subject can do is act to change his own
perception.
This is a common misapprehension about what "control of
perception" means in PCT. To control my perception of my big toe,
I do not choose to perceive it as my thumb or as an apple. I
control some _variable aspect_ of my big toe, some attribute such
as its position, its pressure against the floor, the length of
its nail. To do this I must exert some kind of motor action on
the physical world -- we're not just talking about passively
changing an interpretation.
My perceptions are derived by perceptual processes from sensory
signals. It is the nature of the perceptual process -- the
weightings or other computational processes applied to the
signals -- that generates a new level of perceptual signal. The
relationship of the derived signal to the world is determined by
the form of the perceptual function. But the _behavior_ of the
perceptual signal, once the perceptual function is fixed, is
determined by what happens to my sensory endings, and that is
determined by what happens in the outside world. I may see a
certain object -- an ashtray on a table. I see this ashtray as
existing in a certain position and orientation, in certain
relationships to other objects. Anything that can act on the part
of the external world that I see as an ashtray can alter its
position or orientation or relations to other things. If I want
to control any one of those aspects of the ashtray, I must act on
the world. But other people or other influences can also act on
the ashtray to influence the same variables, at the same time I
do. Those other influences we call disturbances; my own
influences we call actions. When I am controlling the ashtray, I
am not controlling its ashtray-ness, but its variable attributes:
it remains an ashtray as long as my perceptual functions retain
the same form.
Whether I or some other influence affects the part of the world
that I see as an ashtray, the result is to change my perception
of the ashtray -- to change one of the variable aspects of it
that I perceive. So it is not true that one person can't change
another person's perceptions. Merely by moving into your field of
view I change your perceptions. By adopting postures and
expressions, by speaking, by touching you, I can cause drastic
changes in your perceptions. I don't change the _nature_ of what
you are perceiving, but only the _state_ of what you are
perceiving.
One way in which people do alter the form and not just the state
of a perception is by switching from using one control system to
using a different one. Every control system has its own
specialized perceptual function in PCT -- this is a
"pendaeomonium" model in which all control systems control just
one scalar variable in a single dimension. When we speak of "a
control system" we mean a representative control system at a
given level; all behavior really involves many control systems
operating at the same time and at the same level. It is what they
call nowadays a "massively parallel" model.
This means that if I want to change from perceiving the object in
my visual field from an ashtray to a hammer, I may turn off the
control system concerned with controlling the ashtray as an
ashtray and turn on a different one that perceives objects as
hammers, as things to be used to bash on other things. But having
made this switch, I still control only in the terms in which the
new control system perceives: I still control the position and
orientation of this "hammer," and its relationship to other
objects. My control action (which is now in part the action of a
different control subsystem) still controls only the variable
aspects of the new perception, not its identity. There is
nothing, of course, to prevent me from perceiving and controlling
the same object both as an ashtray and as a hammer, as long as no
conflict results. If I hammer very carefully, I need not spill
the ashes.
There is one final way in which the form rather than the state of
a perception can be changed: through reorganization. This is
largely what a baby is engaged in: creating objects and other
perceptions by forming and reforming perceptual functions that
produce new levels of perception that are amenable to control.
The forming and reforming of new perceptions continues through
life, but never again at the furious pace of infanthood. This is
a very slow process in an adult, because for the most part the
perceptual functions that exist are the most appropriate and
mutually-consistent set that can be found. Even if an adult
experiments with new forms of perception, it becomes harder and
harder to find a new form that is as easy to control, without
conflict with other control processes, as the perceptual entities
that already exist.
Reorganization is not a rational process; it is the build-in
process that does the constructing of the rational mind, of the
hierarchy of control systems. So it is not a part of ordinary
behavior, except in times of great difficulty, pain, or
confusion. It is not fast enough to deal with the moment-by-
moment demands of control in the normal world.
To put this discussion in a nutshell, control of perception means
controlling the world we see, hear, feel, smell, taste, and so
forth. We do not experience perceptions as being inside
ourselves; we experience them as the world in which we live, and
the body in which we live. The form of the world we experience is
created while we are very young, as we reorganize to bring order
into the mass of sensory intensities with which we begin. And at
the same time, we learn how to "do" things -- that is, to exert
forces which cause the experienced world to change, and then to
change toward the states we come to prefer. We learn to control
this world, which according to PCT is a world of perceptions that
are really neural signals in a brain, although they don't look
that way.
So we can and do affect other people's perceptions, whether or
not we know what those perceptions are. Whether we succeed in
controlling what we are interested in depends on whether the
other person is also controlling some of the perceptions that are
altered by our actions. Strictly speaking, we can't really
control other people's perceptions because we don't know what
perceptual functions are in the other person and don't experience
that person's neural signals. But we can control parts of the
environment (as we perceive them) on which the other person's
perceptions, whatever they are, depend.
To say that behavior is the control of perception is not to say
that all perceptions are controlled by behavior. Most perceptions
are not. When you reach out to pick up a glass of water, you
control the perception of your hand in relation to the perception
of the glass, but at the same time you _affect_ perceptions of
your forearm, your elbow, the rim of the glass, the lighting on
the glass where the shadow of your hand falls, reflections of
your hand in the glass, and so on. Yet none of those other
perceptions is under explicit control. They change because they
are physically dependent on the perception you intend to control,
but you have set no reference condition for them. You perceive
them, but not having any goal for those perceptions you do not
control them. There is nothing to say that you couldn't try to
achieve independent control of these other perceptions, but you
would not succeed, because you can't achieve independent control
of perceptions that are tightly coupled through physical
dependencies in the outside world.
To illustrate this latter (and I promise, final) point, hold your
fist two feet in front of your eyes and wave your elbow out and
in. You can control the position of your hand while you vary the
position of your elbow. But now look at the midpoint of your
forearm and make your forearm rotate about that point by flapping
your elbow. Now a different point on your arm is under control,
while the other points all change as your elbow moves. You can
control only point, one specific perception, at a time: the
others must necessarily be uncontrolled.
ยทยทยท
--------------------------------------------------------------
Bill P.