Re: Models and ecosystem balance/change...

[Steve Johnson <***>]

Judith Rosen:

"If I left my potted citrus trees outside on my deck
tonight, they would be badly damaged or killed by the
cold. The internal models don't adapt and don't
"learn".

Steve Johnson:

Though I agree with the spirit of the comment I think
it is not an accurate example of the failure of model
in the sense of "Anticipatory Systems". Adaptation is
a broader concept than learning. If I were placed in
the middle of Antarctica with no food or clothes I
would have a pretty good model of what is going on but
I could still not adapt because of the limitation of
my physical body not of the limitation of my capacity
to build models of the environment. I would propose
the example of the butterfly flying towards a light
bulb as an example of model failure. Its incorrect
model causes it to die not the limitation of its body
- it could have chosen to fly in another direction.
Sorry for nitpicking.

   



--- Judith Rosen <***> wrote:

> Dan, I think I understand your uneasiness with what
> is undeniably a 
> simplification process that is required for the
> scientific creation of 
> models-- particularly where natural ecosystems are
> concerned (although 
> the potential for problems can be substantially
> reduced by having a 
> theoretical underpinning of complexity rather than
> the machine 
> metaphor/reductionism).
> 
> However, the fact that models are limited is a fact
> that even living 
> organisms cannot escape. Anticipatory System Theory
> postulates that 
> living systems are anticipatory in that they have
> models of 
> themselves, of their environment, and of more than
> one time-scale 
> encoded into their organization-- and this
> information is utilized to 
> generate an "anticipatory mode of system control".
> This gives living 
> organisms a huge edge in maintaining stability over
> a purely reactive 
> system, but it's not foolproof. Changing global
> climate naturally 
> leads to the death of species and corresponding
> change of whole 
> ecosystems, as we see in the fossil record. While
> these internal 
> predictive models seem to encode that "change is a
> constant", there 
> are ranges within which the models can predict and
> outside those 
> ranges, the models are useless. If I left my potted
> citrus trees 
> outside on my deck tonight, they would be badly
> damaged or killed by 
> the cold. The internal models don't adapt and don't
> "learn".
> 
> This is what led my father to say the behavior of
> living organisms is 
> guided/constrained by "models" in the first place,
> knowing full well 
> that models are abstracted and simplified versions
> of the systems 
> being modeled. The fact that you can "confuse"
> and/or "fool" a plant 
> is exploited by humans all the time, for endless
> reasons including 
> profit. What's the U.S. revenue generated by
> Poinsettia sales over the 
> holidays? Where did those plants evolve? What is
> their normal blooming 
> time? How does the plant trade induce the colorful
> display to coincide 
> with Xmas?
> 
> So, while there is certainly potential for danger in
> the creation of 
> models, it's not a danger peculiar to science or one
> that only humans 
> deal with; it's part of life in general. We have all
> sorts of models 
> encoded within us, as well, some of which seem to be
> vestigial from an 
> earlier time when our environment was different from
> what it is now. 
> Hence, we have an appendix (as one example) and
> don't know what the 
> heck it would be useful for. Obviously, the fact
> that our evolution 
> included a changing diet is part of the information
> we have encoded 
> into our organization... but how far can it change
> and still maintain 
> the integrity of the system?
> 
> This is how I look on the epiphany Atkins stumbled
> onto. The current 
> low-carb movement is a return to eating within our
> encoded models, for 
> a lot of us. Sometimes called "The Maker's Diet" or
> a Biblical Diet, I 
> think it goes much father back than that. The
> epidemic of type two 
> diabetes is a symptom that a diet high in processed
> carbs is too far 
> outside the ranges we can tolerate. We are also
> learning that 
> biodiversity in soil ecology has a direct effect on
> the nutrient value 
> of food grown in it and chemical intensive farming
> techniques are 
> damaging that ecology. Humans have an edge in
> dealing with the 
> limitations of our models: We can (hopefully) use
> our intelligence to 
> learn and deliberately modify our behavior to match
> our internal 
> models.
> 
> Judith
> 
> 
>   ----- Original Message ----- 
>   From: Howard Pattee
>   To: ***
>   Sent: Friday, December 10, 2004 8:00 PM
>   Subject: Re: [ROSEN] Which part of the brain (or
> subject-object 
> system) is conscious?
> 
> 
>   At 02:14 PM 12/10/04 -0500, Dan wrote:
>   >Howard,
>   >
>   >This is all well and good for a formal system and
> for a
>   >static model. But I think if you try to extend it
> to into the
>   >material realm and a continuous and changing
> modeling
>   >process, this view itself needs a complement.
> 
>   HP: This necessity for placing a cut between the
> model itself and 
> what is
>   being modeled has nothing to do with he scope or
> time frame of what 
> is
>   being modeled. Whatever system you are modeling
> requires that the 
> model is
>   not identical with the system itself, but a model
> of the system. 
> Complex
>   systems require more than one model, but the
> models must still be 
> separate
>   from whatever is being modeled .
> 
>   >Dan:  When we consider that the observer only
>   >can observe with aid of a continuous stream of
> energy
>   >and material from the environment (air, water,
> food, etc.)
>   >- including likely interactions with the
> objectified system
>   >he/she observes so the object system and
> observer's life
>   >support cannot be disentangled - and that also a
> continual
>   >stream of energy and materials that have been
> altered in
>   >quality, quantity and configuration (waste and
> other
>   >products, impacts), it seems to me that the
> epistemic cut is
>   >fully bridged, or healed.
> 
>   HP: In this example you are only moving the cut so
> as to include 
> more of
>   the ecosystem. Your model is now about a much
> bigger system that 
> will
>   require many more measurements (encodings), but
> the larger model 
> will still
>   be separated from the system by these encodings.
> All you have 
> 'healed" is
>   the previous cut made for the simpler system.
> 
>   >Dan: So not only is the location of the cut
> arbitrary, but I think
>   >the cut itself is arbitrary or a choice too when
> one considers
>   >that there is an alternative.
> 
>   HP: I don't see any alternative. The position of
> the cut is all that 
> is
>   arbitrary. You have not shown how the concept of
> "model" can evade 
> the
>   modeling relation as Rosen represents it.
> 
>   Howard
> 



        
                
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - You care about security. So do we. 
http://promotions.yahoo.com/new_mail