···
On 2019/02/17 4:02 PM, Bill Leach
( via csgnet Mailing List) wrote:
wrleach@cableone.net
…
Very interesting perception Martin.
BTW, where I lost you is in the
second paragraph near the end. I don’t have time to ponder it
just now but hope to later.
I will comment that in the case for
sentient beings, or at least for humans, they are fully capable
of controlling for perceptions that have nothing to do with
either procreation or survival (either their own or someone
else’s).
bill
On 2/17/19 1:00 PM, Martin Taylor (
via csgnet Mailing List) wrote:
mmt-csg@mmtaylor.net
[Martin Taylor 2019.02 17.14.14]
To my way of thinking, what Bill Leach says about reorganization
makes good sense. Rather than critique it, I would like to put
it in a larger context. The context in which we typically think
of reorganization is that the perceptual control hierarchy is
“inside”, and the intrinsic variable system (whatever it may be)
is “outside”, as is the external environment. A big, wide,
multi-variable feedback loop links the inside and the outside in
such a way that organisms that keep functioning long enough to
have descendants will transmit their feedback loop organization
in more copies than the organization used by those that are less
good at surviving. The loop “control hierarchy - environment -
intrinsic variables - control hierarchy” is self-stabilizing, in
an evolutionary sense (of it’s huge multiplicity of strands,
some do not go through the environment at all. The braids of the
loop that do go through the physical environment are likely to
be the most stable. The “internal loops” can work usefully only
so long as their results help the organism to live in the
external environment.
In a larger context, everything that is within the skin of a
living entity is “inside” and the rest of the world is
“outside”. This “rest of the world” includes other living
things. There is a feedback loop between “inside” and “outside”
here, too. This loop incorporates the first one in the same way
that control of a high-level perception incorporates control of
lower level perceptions that contribute inputs to the higher
level perceptual function. Like the first loop above, this one
can be partitioned, this time into loops through other living
things and loops that include only non-living things. The loop
strands through other living things are like the internal loops
between the perceptual control hierarchy and the intrinsic
variable part of the “outside” in the first kind of loop. The
structure of the loops through other living things can vary, but
only insofar as they help the organism to survive long enough to
propagate that structure through their descendants.
In this context, the "other living things" has the place that
“the intrinsic variables” had in the first feedback loop. We
have a name for it in this context: “the ecology”. I tend to
think of the intrinsic variables also as being an ecology, which
includes the interactions of our microbiome, our cellular
structures that we call “organs” and so forth. Both are dynamic,
full of little feedback loops and big ones. too. Both matter to
our survival, individually and collectively.
In a yet larger context, we can bring "other living things" into
the “inside”, perhaps in stages, starting with “other people in
our community”, “other people”, “other animals”, etc. but
eventually bringing all other living things into the big tent.
Here we have the feedback loop simply between the living
“inside” and the physical “outside”. The ecosystems of the
inside are continually reorganized over billions of years by the
feedback loop between the living inside and the physical
outside. If the Grand Reorganization process works, life will
continue. When it doesn’t. life will stop. Just as the
reorganization feedback loop between intrinsic variables and the
perceptual control hierarchy creates ever more complex
perceptual structures to control, so the Grand Reorganization
has been creating ever more complex ecologies, and the “Middling
Reorganization” that separates living things from the physical
environment has been creating ever more complex life-forms (and
a greater variety along with increasing complexity)
The mechanisms of these reorganizations at different levels
presumably differ, but functionally, are they not all the same?
I am reminded of Richardson’s frequently quoted ditty from
nearly a century ago: “* Big whirls have little whirls/ That
feed on their velocity/ Little whirls have little whirls/ And
so on to viscosity* .” The relationship is the same between
levels of eddies, between levels of “inside-outside”, and
between levels of the perceptual control hierarchy, always until
we get to the most microscopic level, where things change.
Always the littler item is powered by the greater, its structure
supporting the structure of the greater in a different kind of
feedback loop.
None of this suggests how reorganization is implemented at any
of these levels of inclusion, but the notion of evolution as a
weeding-out process is very similar to the idea of “use it or
lose it” in individual skill in perceptual control. I find that
suggestive.
Martin
On 2019/02/17 5:50 AM, Bill Leach (
via csgnet Mailing List) wrote:
wrleach@cableone.net
My
first comment on reorganization is that I believe that, within
mammals at least, the system that would comprise what we would
call the reorganization system is undoubtedly vastly more
complex than any of us would suspect.
For those of us old enough in the engineering world to have
worked with pneumatic and mechanical control systems,
reorganization would be defined as what the engineer does to
the system when properly functioning components fail to
produce proper control. In those days there were few changes
that could be made to an existing system. That is, without
introducing new hardware, reorganization was limited to such
things and input and/or output gain, and mixing in derivative
and integral signal components. Mechanical control systems
were even more limited in what could be changed without new
hardware. The addition of additional hardware would of course
also be a potential component of a reorganization strategy
those an engineer would be quite reluctant to make such
changes until convinced beyond any doubt that doing so was the
only possible solution.
In today's engineered control system world, with the huge
amount of computer power, measurement, data collection and
storage, and analytical power available, control systems are
able to analyse their own performance and often times do a
much better job of reorganization than the engineer of old
could achieve. In spite of occasional catastrophic failures,
these systems are improving at what is now a staggering rate.
Because much of what formerly was always achieved with
physical hardware in an engineered control system is now
achieved using software these more advanced systems can create
entirely new control loops as needed as well as change the
existing connections between control loops.
I believe that much of what modern engineered control system
design have might well exist in living control system though
implementation would likely be almost unrecognizably
different. Despite what many claim about the brain and nervous
system being a like a digital computer, I think a computer
comparison is only valid if you are talking about an analog
computer. And in that respect the brain and nervous system is
so far advance as an example of an analog computer that we
don’t have anything that even remotely emulates it.
I also believe that Bill's random (or at least pseudo-random)
change is certainly a part of that system. His work on
modelling e-coli behavior using a random walk resulted in a
correlation that is hard to deny.
With respect to what I often see referred to as intrinsic
systems (which I take to mean internal systems such a blood
pressure, blood chemistry composition and the like), I believe
that any sort of random changing is very restricted if allowed
at all.
Of course such random changes in the face of sustained failure
to achieve satisfactory control might well be the explanation
for those unexpected deaths that occur where the medical
community is absolutely baffled as to why the person died.
"Survival of the species" is conclusion for a more specific
term, at least for mammals, which is that each individual
living being has a biological system that has a reference for
surviving. Even for those that commit suicide their intrinsic
systems were and will continue to try to survive until they
fail completely.
So, I see the idea that reorganization for intrinsic system is
limited based on the idea that reorganization methods that
occurred where the individual died are ultimately removed from
the gene pool. I think that idea is further supported by
education studies that have shown that the attitudes and
experience of up to 5 generations of ancestors have a
statistically measurable effect on the studied individual.
While I know of no studies that expand that observation into
other areas I don’t think it is completely unreasonable to
presume that such effects could well exist for other aspect of
control.
As many discussions on the net have pointed out the
reorganization system must have a perception of the magnitude
(and probably direction for at least some systems) of error
present and a reference for the maximum allowed error. It is
also reasonable to presume that there also must be some
integrator function so that large but short duration error
spikes do not drive the reorganization system into action
prematurely.
Again as suggested on the net, there is also likely a
hierarchy of change methodologies so that changes gradually
become more drastic as the error either increases or is
sustained for longer periods of time. And I think it is
likely that the references for these are highly dependent on
the specific system(s) that would be acted upon.
I also suggest that one of Rick's demonstration programs shows
that directed reorganization clearly exists and that the same
demonstration shows us that for any person performing the demo
for the first time, will not be using a directed method when
the first overwhelming error happens.
bill