|
Steve J. wrote: Would you agree with the
following
restatement?: No model is a complete reflection of Natural
Law. A
model only encodes an aspect of it. Thus I would say
the failure of
Physics models referenced above is a
failure of the modeler to recognize the
limitations
of the model and not failure of the model to capture
an aspect
of the Natural Law.
I think there are many reasons why models fail. The fault
could be with the model, with the modeler, with the application, with the
encoding and/or decoding, and some combination of the above, which may have
synergistic effects in interaction with each other that magnify the
problems.
I also think it may be impossible to know whether any model is a
complete reflection of Natural Law/s or not.. but, this being a complex
universe, the set of relations underlying it is likely to be infinite
(containing infinite models) all of which are incomplete. In any case, one
thing is pretty clear: it is better to err on the side of caution and presume
that our models are incomplete!
So is the inapplicability of Physics-based Laws to Biological
systems purely a situation of a set of models being incomplete? Maybe. Or
maybe the models are incorrect, in some respect. It could be any combination of
the above reasons for model failure. My father's opinion was that there are
aspects of the Physics-based models that don't satisfy the Hertzian Condition,
but the side-effects are too subtle to be noticed when the models are applied to
certain kinds of systems (simple systems)...or else the side effects are
easily bearable (within the "margin of error"). He said that
atoms are complex systems and yet are considered to be in the province
of Physics. He pointed out, in his work, all the areas where the
Physics-based models cause problems or create paradoxical results. The concept
of "state" was, in his view, inapplicable to natural systems in this
universe.
Judith
PS: You asked in your other post on this subject whether I thought
Newtonian models of gravity were subject to contextual change... I'm not a
physicist, but from what I know of my father's work, I have a "hunch" or
preliminary opinion about the answer to that question; My intuition is that one
of the Laws of Nature in this universe is that the impact of context is
inescapable. This is a relational universe. What is "gravity"? It's an effect of
relations, is it not?
----- Original Message -----
Sent: Sunday, December 19, 2004 5:12
PM
Subject: Re: [ROSEN] What is Natural
Law?
Judith:
"In my view, and I daresay in my father's as
well, the
so-called "Laws of Physics" are inferential laws,
pertaining
to the models for use in specific
applications, and are only laws in that
sense. If they
are not applicable outside of those
original
applications, as in with Biological systems, then they
are
proving that they are not reflections of "Natural
Law"."
Steve:
I think I agree. Would you agree with the following
restatement?:
No model is a complete reflection of Natural Law. A
model only
encodes an aspect of it. Thus I would say
the failure of Physics models
referenced above is a
failure of the modeller to recongize the
limitations
of the model and not failure of the model to capture
an
aspect of the Natural Law. It's a case of "when you
have a hammer
everything looks like a nail." If you
unthinkingly apply a model developed
for one set of
observables to another set you will get "a failure
of
model". To be more precise it is not a failure of the
model (the
entire commuting modelling relation) but a
construction of an entirely new
Bad Model by taking
the right side of the relation from the
original
model and trying to encode another natural system
into
it.
--- Judith Rosen <***>
wrote:
> Hi Steve (and everyone else too, of course),
>
> I think I can help answer at least part of this
>
question:
> Steve J. wrote:Anyone who can provide a definition
>
of what "Law" is
> for the purpose of present discussion?
>
> I posted an excerpt from "Life, Itself" that
> discusses Robert
Rosen's
> concept of Natural Law. He was describing, in that
>
excerpt, how the
> very fact that we are discussing phenomena in
the
> ambience (and
> arguing about what constitutes proof that
we have
> figured out various
> real consistencies in the
ambience) means that
> certain things have to
> be true. The fact
that those things have to be true
> in order for us to
> be
discussing and modeling it is an embodiment of
> this truth: There
> are consistent principles underlying phenomena we
> perceive in
"the
> ambience" and these principles echo over and over.
> This
is Natural
> Law, in totality.
>
> I think the concept of
an echo is a good analogy (An
> analogy is a word
> model, so a
good analogy is a word model that
> satisfies the Hertzian
>
Condition). To model something is a form of echoing
> chosen aspects of
> it in a new system of some sort. That new system may
> be
thoughts, may
> be text (words, language), may be mathematical
>
descriptions, may be
> visual (diagram/graphs/maps/three
dimensional
> sculptural
> representation, etc)... In short, the
mode of
> modeling doesn't change
> what you're doing-- you're
still creating or using a
> new system to
> represent something
else, in some way and to some
> degree. No model is
> going to
correspond to any original system entirely
> (identical twins
>
are not the same in every way, for example) and it
> is not even
> necessary for a model to "get as close as possible"
> in its
entirety to
> that 'ideal' in order for it to qualify as a good
>
model. In fact, I
> would argue that it is not an "IDEAL" at
all.
>
> Why? Because what aspects you want to model
depend
> entirely on what
> you want the model FOR. Based on the
context of what
> you want a model
> to do, the aspects to be
included in the modeling
> relation will be
> chosen. So already
context is critical, as is an
> ability to think! A
> good model
is one where the relation of the
> inferential entailment in
>
the model to the causal entailment in the system
> being modeled is one
> of correspondence AND pertains to what the model is
> supposed
to be
> used for. Trouble can enter the picture from all
>
directions, though,
> because what if the modelers don't know what
they
> need? What if they
> don't know what they DON'T need???
What if the
> modelers don't realize
> they don't know?! This is
where the discussion turns
> to issues of
> optimality and
side-effects. However, let's assume
> that the modelers
> managed
to create models that were in solid enough
> correspondence with
> the aspects of their natural systems which had
> required the
use of
> models for some reason. Let's further assume that
> the
models were
> applied well to whatever tasks required their
>
application (another
> avenue for serious trouble to creep in). In
that
> case, the models
> would prove very useful in their
purpose/s and would
> satisfy the
> Hertzian Condition.
>
> The Hertzian Condition is contextual, too, though.
> Models
created for
> one purpose may only be corresponding accurately
(in
> the modeling
> relation) in those aspects necessary for
that
> particular purpose. If
> the model is assumed to be in
correspondence in all
> sorts of other
> ways, and applied to
different purposes, what are
> the chances that it
> will be in
correspondence in ways made necessary by
> any new context? I
>
think this is what has happened/is happening with
> Physics. Those
> models of reality satisfy the Hertzian condition
> well enough
to be
> useful when applied to the tasks they were created
> for
but the
> assumption has carried over that they can be trusted
>
to be accurate
> representations of all aspects of any natural
>
system. One of the
> things my father wrote in the notes I found after
he
> died was the
> statement; "There are no such thing as
Paradoxes in
> the natural world.
> There are only poorly created
models, poorly
> applied. What we call
> paradoxes are really a
symptom; a side-effect." They
> are a symptom
> that something is
not corresponding well between our
> models and the
> systems
being modeled.
>
> By inference, then, we can say that the
more
> applications in which a
> model satisfies the Hertzian
condition (and its
> application doesn't
> generate
side-effects), the more it is likely to
> reflect at least some
>
aspect of Natural Law. Kepler's "regularities" don't
> automatically
> fall into this category because the sun doesn't
> "rise".
That's a
> model, based on our perception. (However, it's
>
interesting that this
> is a perception that other organisms apparently
have
> modeled too! The
> sun's regular appearance and
disappearance in any
> ecosystem triggers
> the single largest
migration/behavior in organisms
> of any natural
> cyclical
phenomenon.)
>
> I don't think it's accurate to say:
>
Steve J. wrote:Natural Law as per Rosen. An
> epistemological
>
principle delineating what is open to (human?)
> cognition.
>
> He said that the establishment of true
> correspondence via a
modeling
> relation was proof that we are capable of
>
perception/cognition/communication of aspects of
> Natural Law. As you
> will see in the excerpt I posted, he said; "Natural
> Law makes
two
> separate assertions about the self and its
> ambience".
What these
> amount to (my translation) are 1.) That natural
law
> exists; and 2.)
> That we can perceive and describe aspects
of it.
>
> In my view, and I daresay in my father's as
well,
> the so-called "Laws
> of Physics" are inferential laws,
pertaining to the
> models for use in
> specific applications,
and are only laws in that
> sense. If they are
> not applicable
outside of those original
> applications, as in with
> Biological
systems, then they are proving that they
> are not
> reflections
of "Natural Law". Natural Law was what
> he defined as
>
pertaining to both self and ambience: The entailment
> in the universe
> (causality). Laws of science are our attempts to
> establish
congruent
> modeling relations and are inferential laws of
>
entailment. They have
> been generated by our "models" (whatever
those
> models may be,
> including mental imaging,
intuitional
> thoughts/pictures, and
> assumptions). The Hertzian
Condition refers to the
> "encoding" and
> "decoding"
verification processes.
>
> Does that help?
>
>
Judith
> ----- Original Message -----
>
From: Steve Johnson
> To: ***
>
Sent: Saturday, December 18, 2004 2:16 PM
> Subject: [ROSEN]
What is Law?
>
=== message truncated
===
__________________________________
Do you
Yahoo!?
Yahoo! Mail - You care about security. So do we.
http://promotions.yahoo.com/new_mail
|