Re: anticipatory - Daisyworld

["John M" <***>]

Hi, Tim
(Daisyword, below) and Judith (also to her 9/25 post  "time, annticipation")

I found (superficially) my objection to the idea of an 'anticpatory'
homunculus in
R-complexities as explained by both of you (and I think others, as well).
Here is my position:
nothing anticipates, nothing prepares for "not yet" functions to be
anticipated.
I can use the word in the *other* direction only and will come back to the
'anticipatory systems'  in this sense later below.

The way I CAN look at anticipation is BACKWARDS:
 if a functional change does occur, it is dependent on structural
(functional) precursors(!)  that ANTICIPATE the possibility (potential) to
the function that occurres. E.g. the desiduous maple-tree what J. mentioned
in her post,  drops its leaves IN ANTICIPATION of the cold weather-to-come,
for survival:
it does survive, because such occurrence is a result of its dropping the
leaves,
so for the survival the desiduous behavior is anticipatory (a functionally
necessary condition).
A gene is anticipatory for the development of an organ. Tim will hate the
other expression I would use as a synonym: a "precondition" for something to
occur. Not that the maple "knew" about the snow coming soon, (no thinking
homunculus), but its survival IS preconditioned by dropping leaves, as being
anticipatory to it in advance.
*
Now I come to my view "without the homunculus" who(?) KNOWS what to
anticipate and prepares the complexity for it.

In the wholeness (unlimited functional changes) of nature every possible
variation occurs, good and bad (an implied condition also to my position in
the
"parasitorical" vs. "symbiotic" dichotomy of earlier).  Some are "good"
(i.e. feasible for "survival" in our views), some are not (i.e. "bad" - in
our narrow view). The former help to proliferate, the latter instigate
extinction.
The variants evolve in a sensitivity to REACT in 'cases' - (e.g. maple
tree) - to the shortened daylight or(and) the reduced temperature etc. - by
a "disease" of the leafs to drop. The leafless maple is less sensitive to
frost and wind. Some further variations of those had already the function of
a revitalization when the conditions change to more viable (we call it:
springime). In the millions of variants, a certain type (a good hit in
evolution-speak) survived and proliferated in a measly N million years and
we call it a maple tree.  Now we (evolutionists, biologists, philosophers
and other learned reductionists) look at today's maple tree and say: it
developed the characteristics FOR SURVIVAL. A static view from the snapshot.
And WE  assign anticipatory capabilities to something in a 'try and err'
infinite game, because we do not see the "unsuccessful"  i.e. the died out
variations. So "nature is smart", "God had its plans", the "evolution
strives towards 'better' and 'more complex' formats".

This (my) position is definitely deterministic and recognizes the 'unlimited
variants in change' = the process of existence. It also recognizes the
variations from which survival can be 'self-selected' into the kinds we see
NOW and draw our (scientific) conclusions upon: in a static view of the
dialectic dynamism.
(I learned a dialectic view -believe it or not -  in the Marxist-Leninist
seminars I was forced to participate in. They meant it politically and
economically, I extend it now into my 'complexity-thinking).

So my definition of an 'anticipatory' structure, process, function,
arrangement etc. is the one WITHOUT which a process cannot occur.  The
PROCESS has to anticipate it... as a necessary conditon for it to occur. For
feasibility.

This is the way I interpret Judith's words:

>An example of an anticipatory system's behavior that my father used with me
is useful in the discussion of feedback controls in a mechanistic system
where it relies on an "error signal" and in an anticipatory system where the
changes are made before they will be needed, based on an "internal
predictive model" (a phrase that isn't what it sounds like, as Tim has
pointed out,...)<

An "error signal" is definitly anthropomorphic, unless we emphasise Judith's
correct quoting as having been included 'in a mechanistic system' - where
such may  indeed be built in (as the designed mechanical homunculus). See
below.

((Things are different in 'human minds involving' social etc. organizations,
where the complexity of the thinking components (people) may involve errors,
both in anticipation and in ways to anticipate (right or wrong ones). But
that goes into an again slippery definition of the nonexisting 'levels' of
complexity. ))

Please consider for the "error signal" what Tim quoted earlier from RR:
> "Anticipatory control differs radically from the more familiar styles of
> control based on feedback. Feedback is a purely reactive style of control
> based on the notion of error signal. Intuitively, the behavior of a
>feedback system has already deteriorated before any corrective action can
be >taken; the degree of feedback correction imposed is determined by the
error
>signal.
> An anticipatory system clearly does not employ an error signal in this
>sense and thus may begin to institute a control action before any
deterioration
> has occurred." [RC 325-326]

Yes, feedback is 'purely reactive' but in preceeding generations and it
destroys the (error-involving) variant-avenue, freeing the 'niche' for the
variations with more applicable conditions. So the a posteriori observed
"change" shows the application of the "error-anticipation" in, what we *now*
consider in the (static) observational image of the early generations.
And it is "very scientific" to call it "adaptation".
I think this is in the spirit what RR expressed in the last par of Tims
quotation.

I find sometimes Rosenspeak difficult to follow (his language is not
sufficiently "mothertonguish" for me to properly capture the word-twists)
and so I feel like paraphrasing. This is not an argument 'against', rather a
clarification for myself, which may be in a different shade from the
understanding of some others.
However, I feel like a Rosen-ally, appreciating his comprehensive
pioneering - composing an extended system, what I did not do (and probably
will not ever).
Please consider that my thinking does not come from a biologic view and is
not based at all on a -quantiz/ed/able- formalism. I went from generalities
(nature-complexity of Bogdanov, mind-gestalt of Köhler, implicate and
wholeness of Bohm, strong criticism of the Santa Fe formalistic ways, and
others in recent times) into some fundamental agreement I found in RR - at
least what I read ABOUT his work. I am just planning to look into Ian Smuts'
complexity as an early comparison (1924). There are also two (post-Rosen?)
German systems (econo-behavior and a novel energy-concept, - Ewes and
H.Hass) I find worthwhile to evaluate. So I try to keep "objective" and
"free" - within my own logic and philosophical (natural view-related)
acceptance.

With best regards

John M

----- Original Message -----
From: "Tim Gwinn" <***>
To: <***>
Sent: Wednesday, September 24, 2003 10:47 PM
Subject: Re: Daisyworld / N-body

> Hi JohnM,
> One of the characteristics that distinguishes a mechanistic feedback
system
> from self-an anticipatory system is that the feedback system relies on
some kind
> of "error signal". Since I happen to be reading one of Rosen's unpublished
> works "Rosennean Complexity" (that Judith is now publishing and selling
>via her website) at the moment, I will quote from that:
>
> "Anticipatory control differs radically from the more familiar styles of
> control based on feedback. Feedback is a purely reactive style of control
>based on the notion of error signal. Intuitively, the behavior of a fedback
> system has already deteriorated before any corrective action can be taken;
>the degree of feedback correction imposed is determined by the error
signal.
> An anticipatory system clearly does not employ an error signal in this
sense
> and thus may begin to institute a control action before any deterioration
> has occurred." [RC 325-326]
>
> So, based on the predicted future of the internal predictive model, an
> anticipatory system can generate control actions _now_ instead of waiting
> until some future time _after_ the system will have deteriorated.
Certainly,
> time plays a central role in both feedback and anticipatory controls. I'm
> not sure if this fits in with your view of anticipation as "a time-related
> view where some feedback-informations are visioned as 'earlier' events".
>
> Hope this helps,
> Tim
>
>
> > -----Original Message-----
> > From: ROSEN Forum [mailto:*** Behalf Of John M
> > Sent: Wednesday, September 24, 2003 5:34 PM
> > To: ***
> > Subject: Re: Daisyworld / N-body
> >
> >
> > Tim, here is my idea from my stubborn ignorance on the contents
> > of your and
> > Victor's learned posts about Gaia, the anticipator.
> > I liked Lovelock's Daisy-s (I read it ~15 years ago) and it gave
> > me a strong
> > credit for Gaia. With the other (your) side I confess: I glanced only
> > briefly in the 'Anticipatories' and could not swallow it.
> > Then I reconsidered it some 6 times and found that "AHA", I only
consider
> > time a (reductionist?) crutch to our understanding the world, (I
> > had sweats
> > to think up concepts in a timeless vision), so "anticipatory" and
> > "postsipatory" are only differences in a timed view. Just like cause and
> > effect. I made my peace with RR's anticipatorialism by putting aside the
> > concept as a time-related view where some feedback-informations
> > are visioned
> > as 'earlier' events. Not too practical, but it gave me the peace
> > of mind. Of
> > course I may have completely misunderstood RR.
> >
> > The Gaian feedback-series I consider as the same, but perfectly
> > followable.
> > If I look at it without the timing factor, I am right in the middle of
> > anticipation (also without timing).
> > Don't forget that in my narrative the universe is a timeless
> > fulguration and
> > only in the inside do we identify time and space for our comfort.
> >
> > So I resign from this topic with anticipation of a good resolution
between
> > the two of you. Then I may know which part of the agreement to join.
> >
> > Best regards
> >
> > John Mikes
> >
> > ----- Original Message -----
> > From: "Tim Gwinn" <***>
> > To: <***>
> > Sent: Wednesday, September 24, 2003 4:00 PM
> > Subject: Re: Daisyworld / N-body
> >SNIP precedents