[Martin Taylor 2018.07.15.23.21]
[Bruce Nevin 2018-07-15_18:22:27ET]
OK. I guess we can say it's asuggestive but unproven “hypothesis that conflict most of the
time is due to ignorance”, where ‘ignorance’ ranges from lack
of input function to ignoring, with points in between.
Ignoring seems to be a controlprocess worth investigating. One reason for ignoring a
variable might be because to recognize it would conflict with
control of some other variable(s), as in the freier dynamic,
or in a different way as in denial of white privilege.
I believe that there are controlsystems specifically for ignoring.
This all seems reasonable to me. So far as I know, I can't see it as
a natural property of the Powers hierarchy, but if I were a betting
person, I would lay my bet on your side even at unfavourable odds.
For sure not all controllable perceptions are being actively
controlled at any one moment, so there is ignoring on the output
side. Your examples and others I think I could dig up with a little
searching on Google Scholar suggest that it also happens on the
perceptual side. What, for example, accounts for the inability of
most people to see a person in a gorilla suit walking through a
group of people tossing a ball when the viewer has been asked to
count the number of catches (o some such task)?
A low-levelexample: the primary function of efferent innervation
of the outer hair cells in the cochlea is understood to be to
amplify the signal picked up by the inner hair cells, but they
seem also to have a selective input-cancellation function. I
speculate that they are involved in selectively hearing one
voice amid many conversations, or following one instrument in
an orchestral sound; perhaps also the muscles attached to the
bones in the middle ear (tensor tympani and stapedius) whose
primary function is thought to be to protect the eardrum from
loud sounds can have an ‘ignoring’ function. (In myoclonus
they can actually cause the eardrum to emanate sound out from
the ear.) But there seem to be many ways to ignore perceptual
input and higher-level perceptual constructs. The remarkable
effects in studies of hypnosis indicate that we have very
powerful capacities to ignore as well as to imagine
(hallucinate). It would be hard to justify an assumption that
the mechanisms operative in the special circumstances of
hypnosis or other altered states have no role in ordinary
experience.
I was peripherally involved in a hypnosis study many years ago, when
I had a summer student who during the school term worked with a
psychologist experimentally studying hypnosis. Using a detection
theory analysis, I was able to show that in their experiment they
could use hypnosis to inhibit perceptual ignorance that happens when
people listening to something on which they had to concentrate in
one ear missed clearly audible events in the other ear. How would
PCT deal with such observations as you mention and with results such
as this? I don’t have any answer.
Martin
···
/Bruce
On Sun, Jul 15, 2018 at 5:17 PM Martin Taylor > <mmt-csg@mmtaylor.net> wrote:
[Martin Taylor 2018.07.15.16.46]
On2018/07/15 4:14 PM, Bruce Nevin wrote:
MMT: my
question to you was how you might ease the df
bottleneck, which might occur anywhere in the
conflicted loops, by learning new techniques for
controlling low-level perceptions. The new
techniques not only need to bypass the immediate
bottleneck, but also need to circumvent the
absolute limits imposed by our limited
musculature. So the “always” in your original
comment (which I suspect was a little off-hand)
remains suspect.
Yes, an off-hand hunchprovocative of thought, labelled hypothesis and
hedged with “I’m inclined toward”.
But help me to understandhow your observation about degrees of freedom
applies to this matter. Clearly, I am missing
something. A DF bottleneck at some particular
juncture in a particular control loop certainly
limits the control that can be effected by that
loop.
This may be true, but I think this is the start of amisunderstanding, I was intruding myself into a discussion
of conflict, and that inherently means that more than one
control loop is involved.Let's go back to the three-abreast columns ofpeople/degrees-of-freedom trying to get through a door that
allows only two per second to go through. Now suppose that
in each triad, the left one wears blue, the middle one
green, and the right-hand one red. If they arrive at two
triads per second, it would be possible for all the
red-shirts to go through, leaving the blue and green piling
up in the ante-chamber. If they arrive at one triad per
second, any two coloured sets could bet through, leaving the
third. Or in either case, a mix of colours could get
through, leaving an inverse mixture in the ante-chamber.Think of each person as representing successive samples ofthe output of one of three perceptions being controlled, say
the magnitudes of the red, blue, and green intensities of a
pixel. The rate of arrival is the rate at which independent
perceptual samples arrive in sensors for the three
primaries. In order to control any one of them, the output
sample rate for that one must be at least as fast as the
input sample rate. The “door” perhaps represents the two
independent directions of movement for a mouse, which we
assume can be done at a rate of one df/sec for each
direction of movement. Using that mouse, one could perfectly
well control any two of the three, but not all three at the
same time.The "classic" conflict is a simplification of this, in whichthe “door” is the value of some external variable that two
controllers want to perceive each with a different reference
value. That value is the bottleneck where the two control
loops converge. The triadic equivalent would be a case in
which there are two variables, X and Y, and three
competitors have different references for X+Y and X-Y. Any
two of them can bring their perceptions to their reference
values, but all three cannot do so simultaneously.In a canonical conflict, arestriction on the efficacy of q.o is imposed by
opposing action produced by the conflicting control
loop. Are you considering the “degrees of freedom”
here to be the range of alternative control loops
and environmental feedback paths that can be means
of control at the higher level?
That's certainly a possibility, but I was actually thinkingof limitations such as the limited number and speed of our
hands and arms (and even of our speech, though that’s rather
a special case).Otherwise, I don't see howthe line of argument applies to the considerations
at hand. That’s why I reframed it in terms that seem
to you to be “off by a country mile.” I don’t see a
direct application of degrees of freedom in a single
loop to a speculation that alternative lower-level
loops are always capable of being found or created
as means of controlling at a higher level when the
initial means of control are not working.
No. The single loop isn't the source of the problem, butfinding alternative means of control that don’t create a new
bottleneck when coupled with the competing controller’s
operation, is the single-loop’s solution to the problem.
Lower-level supporting loops may, in principle, always be
imaginable, but can they also survive the other bottlenecks.
In your car versus where people want to go example, the
solution does not involve conflict unless the use of the car
or the alternative happened to make the route impassable for
the other. This situation may not be so kind in other cases
of conflict, particularly internal ones such as are the
realm of concern for MoL.Failure to find or createalternative lower-level means, I called ignorance,
though it might also be called failure of
imagination. If in fact you are applying the notion
of degrees of freedom to the range of freedom to
employ alternative means,
which, I hope you now see, I am not,
then by saying that thedegrees of freedom there are limited you are merely
denying my premise that the creativity of control
systems is not finite, at least not in any obvious
way.
I'm not denying it. I'm simply saying that there are“envelope” limits to which any such creative solution must
conform. There is a finite supply of tools (muscles) and
time in which to use them, and if the creative solution
needs unavailable tools or too much time, it won’t work. I
previously said that in less concrete language.Martin
On Fri, Jul 13, 2018 at 10:07 PM Martin > > > Taylor <csgnet@lists.illinois.edu > > > > wrote:
[Bruce Nevin2018-07-13_18:15:28 ET]
I am inclined to thehypothesis that conflict is always a consequence
of ignorance.
Could you elaborate on this, emphasising the“always”? To me it sounds as though you are saying
that all conflict is caused by incorrect perception,
and that therefore the degrees of freedom limitation
usually invoked as the reason for conflict is
actually just a side-effect of some misperception.
Or are you saying that the ignorance involved is a
failure to perceive that it might be possible to
control by using some different, possibly as yet
unlearned, lower-level perceptions?Martin