[Martin Taylor 2018.05.30.17.28]
philip 5/30 1:16
Again I think we are talking about different things that have the
same name. To start at the end of your comment, You are correct that
Powers drew and talked about the imagination connection as a
retrieval of a past perception. If you go back far enough through
reorganization in the individual and evolution through the
ancestors, everything about the organism is the result of past
effects of the environment on the values of variables inside the
organism. The question is at what point it become sensible to talk
about that as an example of perception being due to sensory input.
If I imagine myself as being able to survive without a space suit in
the vacuum of space, dancing on the low-gravity surface of Ceres,
every bit of that perception is a replay of some past sensory input,
but I doubt I have ever perceived that combination in that way in my
life up to this moment. Likewise with Bill’s imagination loop. The
actual value produced by that either is or is part of the input to a
perceptual function, but has that imagined perceptual replay ever
been grouped with other inputs, some from the current environment by
way of the senses, some from other imagination loops, in just that
same way? Is the combination a retrieval of sensory input, or all
due to sensory input in any useful way of thinking about it?
Now if you think of Newton and F=ma, the blow-you-off answer would
be that Newton created a new perceptual function at the program
level. But that’s not really a blow-you-off answer, because the same
reorganization process can create new perceptual functions at any
level of the hierarchy, if patterns of data consistently go
together. That’s pretty well what led Hebb to propose that neurons
that fire together connect together to make neural cell assemblies.
As for Newton, he presumably had been thinking of Galileo and of his
own imagined observations of what would happen in a frictionless
world. Could Aristotle have been right that moving things come to
rest of their own accord, or did they perhaps come to rest because
of some force applied to them? When he came up with F=ma, suddenly
he didn’t have to think of special private properties of each
object, but the new program perception gave the same answer for any
object at all. That’s the kind of thing that a new perceptual
function does. Instead of thinking about all the individual things
that go together, you get just one perception for all the different
ways those things have gone together instead of all going off in
their own directions.
Yes, it's true philosophically that everything we perceive (or do)
depends on some pattern of sensory input received by us or our
ancestors over a few billion years, but for practical purposes it is
useful to distinguish imagined variables from variables responsive
to sensory input in the here and now.
Now you say that if the source of the perception is by way
of the imagination loop that perception is not affected by the
sensory input. The imagination connection is a recording of
past sensory input, so the source is still sensory input.
Further along you say, “when I control a prediction I am
bringing the prediction to line with what has been observed in
previous situations where the same prediction was used”.
Consider the equation F = ma, for instance, which was never
perceived by Newton as coming from the environment. This
equation was coming from Newton’s imagination, however, it is
certainly not a recording of sensory input. I suppose my
memory fails me. Was the imagination connection not simply
the retrieval of a past sensory input?  Â
Martin
PS. This is probably my last contribution for several days, as we
leave for a month tomorrow and I haven’t started packing.
···
On Wed, May 30, 2018 at 12:39 PM,
Martin Taylor mmt-csg@mmtaylor.net
wrote:
[Martin Taylor 2018.05.30.15.32]
philip 5/30 12:07
Thank you for the response, Martin. Before I
read this paper I want to discuss your response
for a moment. Specifically, the statement:   “Prediction”
here is the thing perceived.Â
I
feel it’s important to not refer to a prediction
as a perception because it is “in the mind” so
to speak. I don’t think it’s proper to refer to
anything in the mind as a perception, whether
the thing is in your mind or in someone else’s.Â
 I
prefer to reserve the word perception for
physical phenomena found in the environment.  Thus,
the color of a chemical reaction or the
trajectory of a particle are perceptions.
But our predictions are not. What are your
thoughts about this?  Â
I tend to go along with Bill Powers when that
makes sense to me, and i this case it does. I have two
versions of the meaning of “perception”. One is a signal
based in whole or part on input from the senses that
produces a variable that may potentially be controlled by
influencing some other variable, quite possibly in the
external environment, but certainly in the environment of
the control unit. That variable is a “perception” in the
hierarchy. If the source of the perception is by way of
the “imagination loop” that perception is not affected by
the sensory input. The other meaning is of conscious
perception – that of which I am aware, some of which
comes from the senses, some from imagination, and some
from memory. When I perceive (and control) a prediction I
am bringing the prediction to line with what has been
observed in previous situation where the same prediction
process or its ancestor was used. It is not the “blue”
that is the controlled perception, but the predictive
process that failed to produce “blue” as the prediction.
Martin
On Wed, May 30, 2018 at
11:09 AM, Martin Taylor mmt-csg@mmtaylor.net
wrote:
Philip,
You may be interested in "Predictive
Coding Theory". Try this paper by Seth and
Friston, which Warren Mansell introduced
to CSGnet a while back: https://scholar.google.ca/scholar?output=instlink&q=info:dRUJ4E-HQwEJ:scholar.google.com/&hl=en&as_sdt=0,5&as_ylo=2014&scillfp=16177073628382108007&oi=lle .
I don’t know where on the web this is
discussed, but if you are interested, I
don’t imagine it would be too hard to find
out.
Your example of the chemistry experiment
suggests that you are controlling a
perception either of a prediction or of
the process by which that prediction is
made, but you are not controlling a
perception of the colour of the chemical
product. Your are asking “why did I not
understand the chemical process, and what
is the correct understanding?” or “I have
a reference to perceive my predictions as
being correct, and this one isn’t, so I
must act to correct it.” “Prediction” here
is the thing perceived, not a method of
control.
Your comments on "theory control theory"
are apposite, but I think what you observe
is control by different people of a
perception of what they want to be the
same theory, for which different different
people have different reference values.
The theory perception that they control is
a perception of what might be in someone
else’s mind, which might be a correct
perception or might be far off the mark.
But their perception of what is in someone
else’s mind differs from their reference
value for what they want to be in that
person’s mind, so they act to influence
it, and the mode of action is the post a
message on CSGnet. I don’t see what that
has to do with prediction.
Martin
On
2018/05/30 1:38 PM, PHILIP JERAIR
YERANOSIAN wrote:
philip 5/30 10:10
I will express a criticism
about why this theory perhaps
should not be called perceptual
control theory. We begin by
observing what is constant and
what is variable in a science
experiment. In general, the
results of an experiment are
constant and our theory is varied.
The experimental method dictates
that we initially intuit a
prediction based on our past
experiences. This prediction
usually takes the form of some
equation, but not always. For
example, we may predict that the
color of the chemical reaction
will be blue. Then an experiment
is conducted, yielding an
observation which is, by and
large, a perception. If our
prediction does not match the
perception, we vary our prediction
by some process of rationalizing
or reevaluating our theory.
Naturally, it comes as no surprise
therefore that  we  control our
prediction and not our perception.
This argument, as well as the fact
that the discussions on csg have
largely revolved around what seems
to be “theory control theory” lead
me to believe that perceptual
control theory might better be
referred to as prediction control
theory.Â
There is the counterargument
that someone chasing a helicopter
or hammering a nail is obviously
doing perceptual control and not
prediction control. However, I
think that if a person acts only
to control a perception, but
cannot predict or simply ignores
why or how the environment behaves
as it does, then there is some
open loop aspect of his behavior
which prevents it from being
“true”.
On Wed, May
30, 2018 at 7:13 AM, Martin Taylor
csgnet@lists.illinois.edu
wrote:
[Martin Taylor
2018.05.30.10.03]
philip 5/29
15:19
1. What is a guess
in PCT? Is it a
perception, is it a
reference?
I think that is one
you could answer by asking
yourself the question: “Is the
guess about what you want to
observe or about what you may
observe?”
2. What Rick seems
to refer to as q.i is
an observable from the
point of view of the
experimenter. And
what seems to be the
business of control
theory (or perceptual
control theory) is to
understand if q.i has
anything to do with
the behavior of the
organism. And if it
does, i.e. behavior
varies with q.i, we
understand that our
perception of q.i has
some analogue as a
perception in the
organism.
I don't agree with the
premise, but I think this is
more in Rick’s ballpark than
mine, since the question seems
to depend on the meaning of
q.i.
However, for almost all
meanings of q.i that I have
seen, your conclusion seems to
be the same: that perceptual
control is irrelevant, and
that behaviour is a simple
stimulus-response process.
3. Why are we
trying to discover
what perceptual
variables are being
controlled by
organisms?
Sometimes we want to
know why people did what they
did or why what they do is
what they do. If PCT is a good
theory, to answer such
questions requires knowing at
least an approximation to a
dominant controlled perception
that is disturbed in such a
way that the error-reduction
action is what is or was
observed (and sometime also is
likely to be observed if such
a disturbance were to occur).
That’s very useful, but I
think your question might get
a variety of answers if you
changed “Why” to “When”.
Martin
On
Mon, May 28, 2018 at
8:52 AM, “Boris
Hartman” csgnet@lists.illinois.edu wrote:
Â
Bruce
Â
From:
Bruce Nevin (bnhpct@gmail.com
via csgnet
Mailing List)
<csgnet@lists.illinois.edu >
Sent:
Sunday, May
27, 2018 1:53
AM
To:
Bruce Abbott
bbabbott@frontier.com
Cc:
CSGnet csgnet@lists.illinois.edu
Subject:
Re: Re:
Controlled
Quantity
Â
[Bruce
Nevin
2018-05-26_19:52:28
ET]
Â
Fred
Nickols
(2018.05.26.0533
ET)
Â
I
will agree
that “cups� is
a variable.Â
They can vary
in size,
shape, color,
composition,
etc.Â
Â
That
said, I stand
by my earlier
assertion that
the cup in
front of me is
not. It is
what it is; it
does not vary.
Â
BN
: It is a
controlled
variable with
an unchanging
reference
value for its
physical
integrity as a
cup.
Â
HB : Cup of coffee is "controlled variable" in
outer
environment ?
No kidding.Â
And where do
you see
“unchanging
reference
value for it’s
physical
integrity”.
Â
In PCT the only "controlled variable" is
perception
which is
matched to
references.
Â
BN
: Collectively
controlled, if
someone else
in your
household also
does the
dishes and
endeavors not
to break any.
Â
HB : Does this mean that all in household
collectvelly
control
“controlled
variables”
(cup, dishes)
in outer
environment ?
Â
Boris
Â
Â
Â
BN
: If this
particular cup
evokes
memories that
you enjoy
perceiving in
imagination
(‘sentimental
value’), then
you probably
control it
with
especially
high gain.
Â
BN
: Control
occurs over a
wide range of
time scales.
Â
Â
Boris
Â
Â
/Bruce
Â
Â
On
Sat, May 26,
2018 at 11:42
AM, “Bruce
Abbott” <csgnet@lists.illinois.edu >
wrote:
[Bruce
Abbott
(2018.05.26.1140
EDT)]
Â
Fred Nickols
(2018.05.26.0533
ET)
Â
I will agree that “cups�
is a
variable.Â
They can vary
in size,
shape, color,
composition,
etc.Â
Â
That said, I stand by my
earlier
assertion that
the cup in
front of me is
not. It is
what it is; it
does not vary.
Â
Exactly
so!
Â
Bruce
Â
From:
Bruce Nevin (bnhpct@gmail.com
via csgnet
Mailing List)
<csgnet@lists.illinois.edu >
Sent:
Friday, May
25, 2018 9:07
PM
To:
Erling
Jorgensen EJorgensen@riverbendcmhc.org
Cc: CSG
csgnet@lists.illinois.edu
Subject:
Re: Re:
Controlled
Quantity
Â
[Bruce Nevin
2018-05-25_21:06:16
ET]
Â
Fred proposed that 'amount
of coffee in
the cup’ is a
variable but
the coffee cup
is not (Fred
Nickols
2018.05.22.1446
ET).
Â
Erling said no, the cup is
a variable
because you
could have
used a glass
to drink your
coffee (Erling
Jorgensen
201805.22 1550
EDT).
Â
Rick said the cup is a
state of
variable that
he called
‘type of
drinking
vessel’ (Rick
Marken
2018-05-22_17:48:57).
Â
If I wanted to point out
some slippery
ground here, I
could
paraphrase
‘type of
drinking
vessel’ as
‘affordance
for drinking’.
Â
Imagine a collection of
drinking
vessels of
many kinds.
Over here are
coffee cups,
with their
more delicate
kin, the
teacups,
shrinking back
from those
ruffians, the
mugs, each of
these with its
handle. Except
for those
styrofoam
coffee cups,
and some of
the paper
cups, which
have no
handle. Over
there the
glasses,
eponymously
made of glass,
have no
handles, nor
do their
heftier kin,
the goblets.
But oh,
there’s a
stein with a
handle; and a
flagon with
two. And by
Jove, here’s a
coffee cup
made of glass.
Â
This seems rather
categorial–canonical
exemplar with
more or less
deviant
similars,
criterial
features, some
but not all of
which may be
absent, and
judgement
influenced by
prior set.
Among the
glassware in
my kitchen is
a tapered
glass 8 oz.
tumbler that
happens to
have a handle.
Or so it
appears as it
stands among
the tumblers
and smaller
glasses. Were
I to serve a
guest coffee
or tea in it,
alongside the
smaller glass
coffeecup that
I mentioned
above, they’d
probably
remark on its
size, but I
suspect they’d
still call it
a cup, not a
glass. I don’t
currently own
a glass beer
stein, but if
I did, and if
I brought it
out in that
company, it
might be a
glass again.
Â
My sense of this is that
we have
different
perceptual
input
functions for
these
different
kinds of
drinking
vessels. All
of them fire
to some degree
given
perceptual
inputs
associated
with drinking.
Each of them
fires to some
degree given
input of a
certain word
or words – at
the least cup
vs. glass vs.
mug, etc.,
naming the
types. One
fires more
strongly given
perceptions
associated
with coffee,
and so on
Ordinarily,
there are more
and stronger
inputs to one
of these than
to any of the
others, and as
far as we are
concerned that
is what we
perceive.
Â
I come to this view from
puzzling out
how in heck we
sort out the
ambiguities
rampant in
language. But
an important
point here is
that you don’t
need language
to experience
and sort out
ambiguity.
Â
A configuration can be
perceived
(present) to a
greater or
lesser degree,
depending on
how well
lower-level
signals fulfil
the array of
inputs that
the input
function for
the
configuration
is constructed
to receive.
Some, of
course, may be
supplied out
of
imagination.
Â
A cup is a perceptual
variable that
may be
perceived as
present to a
greater degree
(as
alternative
variables are
less vividly
perceived) or
to a lesser
degree (as
additional
perceptual
inputs
contribute to
the perception
of some other
variable but
not to the
‘cup’
perception).
Â
/Bruce
Â
On Fri, May 25, 2018 at
7:58 AM,
“Erling
Jorgensen”
<csgnet@lists.illinois.edu >
wrote:
Forgot to 'Reply All' on
this
message…
Â
Confidentiality:*
This message
is intended
only for the
addressee, and
may contain
information
that is
privileged and
confidential
under HIPAA,
42CFR Part 2,
and/or other
applicable
State and
Federal laws.
If you are not
the addressee,
or the
employer or
agent
responsible
for delivering
the message to
the addressee,
any
dissemination,
distribution
or copying of
this
communication
is strictly
prohibited. ** If
you have
received this
in error,
please notify
the sender
immediately
and delete the
material from
your computer.
Thank you for
your
cooperation.***
*** Please
also note:****
Under 42 CFR
part 2 you are
prohibited
from making
any further
disclosure of
information
that
identifies an
individual as
having or
having had a
substance use
disorder
unless it is
expressly
permitted by
the written
consent of the
individual
whose
information is
being
disclosed or
as otherwise
permitted by
42 CFR Part 2.*
>>>
Erling
Jorgensen
5/25/2018 7:56
AM
[From Erling Jorgensen
(2018.05.25
0748 EDT)]Â
>[Rick Marken
2018-05-24_18:21:54]
Â
EJ:Â Just a brief comment,
because it’s a
busy morning.Â
Â
>RM: But your theorist
is free to
define p as
not equivalent
to q.i (and
apparently he
or she has) so
you (or
someone
familiar with
your theorist)
should be able
to demonstrate
the empirical
basis for this
definition.Â
Â
EJ:Â But what about
research that
indicates p as
a log function
of q.i. from
the
environment?Â
I think
there’s
acoustic
research in
that vein and
I suspect
other
examples. Why
can’t we say
that
perceptual
input
functions
create
functions of
the
environmental
input, and so
they are not
strictly
equivalent?Â
‘Strongly
correlated’ is
sufficient for
the pragmatics
of the Test
for Controlled
Variables to
succeed.Â
Â
EJ:Â Useful discussion,
minus some of
the snark or
ad hominem
from one side
or the other.Â
Â
All the best,Â
Erling
Â
Â
Â
Â
Â
Â