Higher order perceptions, silence

[From Rick Marken (930511.1000)]

Oded Maler (930510.1000 ET)--

First, to make it explicit, I'm not a life/behavioral scientist, and
when I criticize your extrapolations, I'm not doing it in the name of
an alternative *scientific* generative theory. It is based on many
out-of-duty thoughts and observations about political, cultural and
historical matters, and on a better (I think) intuition about the
complexity of prediction, that is, how things scale-up as the number
of variables grows.

That's what I thought. I like that kind of candor.

I do have a problem with your description of a controlled perceptual
variable. Your description doesn't seem to match my conception of
a controlled variable. Perhaps there is a language problem. You say:

Take a perception of muscle tension and
your perception of, say, the correctness of US economic politics. The
former has at least some hypothetical relation with an electrode one
can put in the appropriate place in your body. Can you suggest a
similar CEV whose measurement will have some relationship with your
higher-level perceptual variable? Can you poke some millions
electrodes in the hearts, stomachs and pockets of (even a sample of)
the American population and define a CEV as a function of those?
Could you find a resonable correlation between this hypothetical
signal and your percept?

In PCT, a controlled perceptual variable, p, is a neural current whose
magnitude varies as a function of sensory input variables, v. In
general, p = f (v1, v2,...vn). The function, f, as well as the sensory
source of the vi (retina, basilar membrane, muscles) determines "what"
is perceived. The magnitude of p is just the degree to which the per-
ception is "present", via f(), in the prevailing sensory input. The
actual environmental correlates of the sensory inputs correspond to
properties of the environment according to the current model of physics.
According to PCT (neurophysiologists take heed), when you record the
spike rate from an afferent neuron you are measuring p. This measure of
p is a measure of the degree to which the sensory inputs "satisfy"
f(). If f() is just a summation then spike rate is proportional to,
say, p = a1v1+a2v2. So increases in p are directly proportional to
variations in v1 and v2. If p = a1v1-a2v2 then incerases in p are the
result of increases in v1 OR decreases in v2. In the first case, p
just measures the size of the sum of v1 and v2; in the second case, p
measures the size of the difference. But in both cases p just tells you
about the result of the functional combination of v1 and v2 -- not about
v1 or v2 individually. p is a perception of "sum" in one case and
"difference" in the other. In both cases p is just a varying rate of
neural impulses. The "meaning" of p depends on f(). In the first case
p "means" sum; in the second, difference. When a control system sets
a reference for p it doesn't "know" whether it is controlling a sum or
a difference. If the reference is set at 10, then p will be brought
to 10 (approx.) impulses/sec. But in the first case, what is "really"
happening is that the system is controlling a sum of sensory values;
in the second it's controlling a difference.

Similarly, p could be the measure of tension in a muscle or of some
aspect of US economic policy (say, the "strictness" of the fed's
money supply policies). In the first case, p is a perception of
"muscle tension" because it is a function of a sensor located in a
muscle spindle. What defines p, in this case, as a muscle tension per-
ception is not f() per se but the source of the sensory input, v. In
the second case p is a perception of "monetary strictness" because of
the VERY complex f() that computes it from sensory (and lower level
perceptual, p) inputs AND because of the source of the inputs to f().

It is entirely plausible, according to PCT, that one could poke an
electrode into a cortical level afferent neuron and record changes
in firing rate (p) that depend on (what the experimenter also perceives
as) variations in "monetary strictness". This is all a higher order
perception is -- a neural signal whose variations in magnitude indicate
the degree to which a particular sensory experience exists in the
pattern of sensory inputs. Subjectively, p is an experience to which we
might give the name "muscle tension" or "monetary strictness".

I think there is quite a bit of neurophysiological work that is consistent
with this perceptual hypothesis. Neurons have been discovered whose rates
of firing vary depending on 1) the proximity of lines to a particular
retinal region 2) the orientation of lines on the retina 3) the direction
of movement of lines on the retina 4) the binocular disparity of lines
on the two retinae. Things may work differently when you are dealing
with higher order perceptions -- maybe the value of p is carried by the
state of networks of neurons. But I think it is a fairly strong (neuro-
physiological) prediction of PCT that, if a variable can be controlled,
then it must be represented neurally as a unidimensional magnitude, p.
NB. Ken Hacker -- A falsifiable prediction.


I do not consider it a LIE to take
interpreted experiences which a person calls "data" or "facts" and
reinterpreted them.

I don't seem to be making myself clear. Of course it's not a lie to
interprete data in PCT terms. This is what I called "just so" stories
and I think they can be great fun. What I called a "lie" was the claim
that PCT could account for 100% of the variance in some conventionally
collected set of data. It can't -- and I think it is important to
explain why to those who expect PCT to do better than their own theories.
The short answer is: because the data that they would have PCT account
for has been collected in the context of the wrong model -- the
cause-effect model.

There is NO BOSS REALITY and that is a fact!

I don't think I'm ready to sign up to that one. It may not be the BOSS
but I think there's some pretty good evidence that their is SOMETHING
out there consistently constraining the means I can use to control my

Of course if
you have very little concern whether anyone uses PCT or not then I
would guess (just a guess) the most efficient act to perform is

As I said in the post to which this is a response, I am interested in
trying to help people (including myself) understand PCT. I would prefer
that they (and I) understand it before they start using it. I think
people often start using PCT before they really understand it; this can
lead to real problems. I don't want to name names but I think something
like this happened with a fellow whose name rhymes with a fancy women's
college in the east (OK, it's Vasser). Basically, this fellow is using
PCT in what are sometimes quite non- PCT like ways. So I won't be silent
about PCT -- not because I want people to use PCT, but because I want
them (and myself) to understand it.