[From Bill Powers (980407.0926 MST)]
Bruce Nevin (980406.2212 EDT)--
Reorganization system, memory, and attention all seem contiguous to and
interdependent with aspects of the control hierarchy, but do not consist of
dependent variables or functions within it. There appear to be some
interdependencies between these three as well.
You have many interesting things to say about the phenomenon of attention,
but I still have no way to model it. Is it a variable? A function? Ectoplasm?
Best,
Bill P.
[From Bruce Gregory (980407.1128 EDT)]
Bruce Nevin (980406.2212 EDT)]
It does seem as though
* Attention goes where there is error. Error "gets" our attention.
That certainly is HPCT compatible.
* We ignore what we don't care about.
I think this would have to be stated as we ignore signals that
generate no errors. "Don't care about" is not an HPCT term.
* We ignore whatever we are controlling without difficulty (not only with
low error, but also without notable effort).
Again, error signals must be the key, not efforts since the
latter are not defined in HPCT.
You can control a perception to which you are not consciously attending.
Heartbeat and digestion are obvious examples, and most of the breathing
that we do. (Whether you are subconsciously attending to it I will leave
to the philosophers and angel counters. So controlling without conscious
awareness defines one aspect of whatever we mean by the subconscious,
subconscious control.)
Definitely.
Contrary to (b), you can also attend to a perception that you are not
controlling, with no consequent effect of changing it. This is done in
meditation practices where you pay attention to sensations in the body.
For example, in the practice called anapana you pay attention to
sensations associated with air moving in and out of the nostrils. (To
state the obvious, these are side effects of breathing, you're not
attending to the perceptions that are autonomically controlled by the
actions of breathing.)
To pay attention to a perception you are not controlling
requires that you _are_ controlling another perception, poosibly
a higher order perception.
While meditating it seems as though
* You do not control the chosen perception (e.g. sensations associated
with air moving in and out of the nostrils). Nor do you control the
breathing that affects the changing state of those sensations.
See above. You are controlling for perceptions associated with
"I am meditating."
* You do control a perception of attending to the chosen perception.
Exactly.
There really does seem to be a perception "I am paying attention to these
sensations" such that realizing I am drifting with a school of fish, or
having a fight with my brother in the 6th grade, or reliving a
conversation I wish had gone differently, or watching some nameless
display of intertransforming visual colors and shapes, or whatever, is
definitely a disturbance, and awakening to it I return to paying
attention to these sensations at this part of my body.
Yes. Just the point I was trying to make.
* You control perceptions amounting to posture.
Yes.
* You ignore error in a lot of perceptions that you normally don't notice
that you control (this itch, that cramp).
Again "ignoring error' is not defined. Either the gain on
certain loops is reduced or the reference values are reset by
higher level systems.
Ignoring does not mean gain is zero. If gain were zero I don't think you
would notice them. It seems to take effort to retain the chosen focus of
attention. Ignoring seems to be accepting error without reducing gain.
Then there must be a higher gain associated with the perceptions
that you are controlling.
* Most of what you ignore is stuff out of memory and imagination.
Controlling other perceptions associated with observering
breathing.
This last is interesting. It appears to me that most of the distractions
that dilute attention are not perceptual inputs from the environment but
perceptual inputs from memory and whatever process underlies fugues of
imagination. (In passing: imagination appears to me to be a lot wilder
and woolier than what the imagination loop seems to model. What artists,
composers, and poets do is considerably more complex than the appearance
of control by substituting a copy of the reference signal in place of the
perceptual input. This is in my view a major oversimplification or
misdirection in B:CP.)
Definitely. Much of the time people are controlling internal
perceptions--voices in your head. I agree B:CP does not address
this or artistic imagination. But what the hea... it's only one
book.
I believe that the quality (detail, articulation, clarity, specificity)
of your memory and imagination correlates directly to the quality of your
attention. Hence one of the immediate and more well recognized benefits
of meditation.
Meditation appears to focus on the control of internal
perceptions.
What does all this say about the place of attention in PCT? These are
naturalistic observations. One consequence, the contradiction to (b)
noted at the outset, is that the proposal that "attention is what
determines loop gain in a control system" cannot be true,
if we understand by this that
where attention is on a perception, loop gain on the control of that
perception is increased; and where loop gain is high, attention is on the
perception being controlled.
I don't follow this. Attention is a response to error. i don't
see why another response to the same error might not be a gain
increase in the control loop where the error is arising. In
fact, this seems like a reasonable step. If it fails,
reorganization might be called for.
One can attend closely to one's control of a perception while decreasing
the gain in that control loop: make a fist loosely; draw a more or less
straight line; pronounce words normally rather than carefully. Do these
things while paying close attention. It appears that attention does not
increase loop gain.
Different perception. You are controlling for the perception
that you are loosening your fist, drawing an "more or less"
straight line, etc.
Attention does seem to be necessary for certain phases of learning a
skill. ("Hold the racquet like this ... <<adjusting student's grip>. This
is the western grip.") Practice seems to have the purpose of changing
output functions in control loops (and perhaps input functions) such that
after the changes attention is no longer required for control with high
gain. We say then that the new skill has been learned. (Bruce Gregory we
presume now uses the western grip with high gain but without distracting
his attention from that oncoming ball.)
You will be amazed to learn that Bruce Gregory does not play
tennis. or perhaps you won't....
Attention doesn't cause the change (reorganization, learning). But the
quality, clarity, and specificity of memory, and thence of reference
perceptions, can be no better than the quality, clarity, and specificity
of one's observation of the desired state or process, or of the
observations on which one's imagining of the desired state or process is
based. You can teach yourself to do something in a certain way (western
grip, C-G7-F-C on a guitar) by doing it carefully over and over. It can
help to follow a closely observed example and compare one's own rendition
with the example; perhaps doing it much more slowly than the accomplished
person providing the example. It can help to have the teacher or coach
place one's hands, feet, etc. in appropriate positions; Japanese teachers
do this much more than we.That still doesn't tell us what this thing is, attention, in a PCT model.
It isn't a part of PCT according to Bill.
It seems almost to be a function in another system orthogonal to the
perceptual hierarchy,
This seems compatible with Bill's statements.
This has the hallmarks of attempting to perceive a thing from inside that
thing, attempting to understand a level of perception from within that
level, attempting to resolve a conflict from the locus of its results
rather than from its origins up a level. It does not seem possible to
step outside of awareness and then to report back.Attention (whatever it is) does seem to have a point of view on each
level suggested in B:CP, even the highest. It is tempting to think of it
as a primitive term, antecedent to the emergence of the lowest level of
control, able still to access it as well as all the levels developed
above it. Or (here's a wild card) as the highest level, which does not
control, not even with a reference of zero, it simply observes, and new
levels of control evolve "under" it.
This seems to me what Bill is saying.
Bruce
[From Bruce Nevin (980407.1354 EDT)]
Bill Powers (980407.0926 MST) --
You have many interesting things to say about the phenomenon of attention,
but I still have no way to model it. Is it a variable? A function? Ectoplasm?
I suggested a function in another system, affecting variables such as gain
in the control hierarchy. I suggested that the problem is with our poor
understanding of memory, reorganization, and attention, and that all three
of these are separate from the control hierarchy but affect its functions
or variables; and that they might be related to one another. I'm trying to
reach some clarity about subjective experiences of attention and control
and relate that to what we have said so far about attention, gain, etc.
Some of what has been said seemed inconsistent with experience. No grand
conclusions in sight, sorry, but I hoped it might be of some use to try.
Bruce
[From Bill Powers (980408.0241 MST)]
Bruce Gregory (980407.1128 EDT)--
Bruce Nevin (980406.2212 EDT)--
* You do control a perception of attending to the chosen perception.
Exactly.
You can attend, but how do you choose the perception to which you are going
to attend before you have attended to it?
Best,
Bill P.
[From Bill Powers (980408.0244 MST)]
Bruce Nevin (980407.1354 EDT)--
You have many interesting things to say about the phenomenon of attention,
but I still have no way to model it. Is it a variable? A function?
Ectoplasm?
I suggested a function in another system, affecting variables such as gain
in the control hierarchy.
What sort of function in the other system? An input function or an output
function? What are the variables at the input and output of this function?
What does the "other system" do? That is, what is its purpose?
Best,
Bill P.
[From Bruce Gregory (980408.0606 EDT)]
Bill Powers (980408.0241 MST)]
You can attend, but how do you choose the perception to which you are going
to attend before you have attended to it?
I've been struggling with this one for a long time. Ever since I started to
make sense of PCT. Is it just a matter of gain? What sets the gain? We have
lots of perceptions of which we are only aware of a few. What makes these
privileged?
Bruce
[From Bruce Nevin (980408.1212 EDT)]
Bill Powers (980408.0241 MST)
Bruce Gregory (980407.1128 EDT)--
Bruce Nevin (980406.2212 EDT)--* You do control a perception of attending to the chosen perception.
Exactly.
You can attend, but how do you choose the perception to which you are going
to attend before you have attended to it?
The example was of focussing on sensations at the nostrils. Speaking just
of that, the choice initially was to follow the instruction of a vipassana
meditation teacher, and observing sensations at the nostrils was the
evident and present means to do that. Thereafter, the choice was to start
"doing anapana" and observing sensations at the nostrils was the remembered
means to do that.
I think you are asking for something more general and not limited to the
practice of meditation. I don't think "control a perception of attending to
perceptions x" is a common choice. It does seem to me that there is always
a kind of witness point of view that is very much like this, but that
normally is not directed by a purpose that I could identify, other than
just witnessing. It witnesses me sitting and starting to "do anapana" for
example. And meditation practice does seem to make this point of view more
available -- available to awareness, whatever that is.
I'm working on it.
Bruce Nevin
[From Bruce Gregory (980408.1320 EDT)]
Bruce Nevin (980408.1212 EDT)
The example was of focussing on sensations at the nostrils. Speaking just
of that, the choice initially was to follow the instruction of a vipassana
meditation teacher, and observing sensations at the nostrils was the
evident and present means to do that. Thereafter, the choice was to start
"doing anapana" and observing sensations at the nostrils was the remembered
means to do that.I think you are asking for something more general and not limited to the
practice of meditation. I don't think "control a perception of attending to
perceptions x" is a common choice. It does seem to me that there is always
a kind of witness point of view that is very much like this, but that
normally is not directed by a purpose that I could identify, other than
just witnessing. It witnesses me sitting and starting to "do anapana" for
example. And meditation practice does seem to make this point of view more
available -- available to awareness, whatever that is.I'm working on it.
Good. I think it is very important. I was wondering before I saw
your comments about following the instructions of the vipasana
teacher and doing anapana whether we might always direct our
attention in order to reduce errors of higher-level perceptions.
This seems like a real possibility to me.
Bruce the Obscure
[From Bruce Nevin (980408.1527 EDT)]
Bruce Gregory (980408.1555 EDT)]
I was wondering [...] whether we might always direct our
attention in order to reduce errors of higher-level perceptions.
This seems like a real possibility to me.
Observing x can be means of controlling a perception "I am observing x". I
don't see that it has any other direct effect on any perceptions, such that
it can be means for controlling those perceptions.
What seems to happen with MOL, prescribing the symptom, and other ways of
shifting attention, is that we step out of a runaway feedback loop fed by
emotion associated with memories and imaginings. Perhaps it is something
like the oscillation of a buzzer, if that is a fair analogy of error
arising from conflict. Then control systems not caught in the error-buzz
seem to resolve their differences if left alone. Deliberate distraction is
another example. It is a commonplace conversational gambit in genteel
company to shift the subject when participants are getting argumentative.
(I'm trying to connect subjective experience with a wobbly sketch of a
model, your mileage may vary.)
Actions that we interpret as "paying attention" (eyes focussing and
tracking, etc.) are means for continuing to have perceptual inputs that we
are controlling. Mouse movements control alignment of cursor with target;
eye movements and movements that maintain posture control continuing to
have visual perceptions of the cursor, the target, and the relationship
between them. Is "continuing to have this perception that I am controlling"
a controlled perception, or is it part of controlling the perception along
with the mouse movements? If the former, it sounds like observing those
perceptions for the purpose of controlling them. And that is what sounds
like observing is a controlled perception.
BN
[From Bill Powers (980408.1409 MST)]
Bruce Nevin (980408.1212 EDT)--
I think you are asking for something more general and not limited to the
practice of meditation. I don't think "control a perception of attending to
perceptions x" is a common choice.
No. I think "attend to x" is the way it is experienced. Or even more
precisely, just "x".
It does seem to me that there is always
a kind of witness point of view that is very much like this, but that
normally is not directed by a purpose that I could identify, other than
just witnessing.
This is what I call "the Observer."
It witnesses me sitting and starting to "do anapana" for
example. And meditation practice does seem to make this point of view more
available -- available to awareness, whatever that is.
The witness _is_ the Observer _is_ awareness _is_ the point of view. There
is only the observer, observing. This viewpoint is often experienced as an
end-point of the method of levels; the state of observing while detached
from any object of observation (not unaware, but uninvolved and goalless).
This state is always present, but is hard to see when one is involved with
activities in the hierarchy of control systems. The MOL is a direct route
to this state, with no side-issues or fancy imagery or theories. Nothing
wrong with side-issues, imagery, or theories, of course -- you'll get there
anyway. But I doubt that you'll get there any faster.
Best,
Bill P.