Re: Modeling relations and semantics

[Calvin Ostrum <***>]

Carlos Limarino writes:

> So, I have exactly the same problem with Tim Gwinn
> definition. Is 'encoding' a measurement or not? If it
> is a measurement, the idea they manifest "free
> creations of the human mind" (except, in the trivial
> sense that the measurement instrument was maybe
> manipulated by a human body which we ascribe the
> possession of a mind) is very dubious to me.

Maybe I should not butt in, but based on my
reading (and only that and nothing more, thus
rather a stab in the dark) I would interpret 
it like this: encodings of a natural system 
into a formal system are measurements, 
and free creations, in the sense that 
you are faced with reality just being out
there, and then you yourself decide what 
operations to perform on that reality in order
to come up with measurements that can 
then be encoded into a formal system.   The
idea is then that the formal system allows
you to make inferences which after decoding
will tell you something about the natural
system you are modelling.   Wow!  What is
so profound about this? It is just common 
sense.

Since you can make the measurements 
with the intention of encoding them into any of 
many formal systems you might have at the ready,  
in effect, it is really the formal system, along
with a general strategy for linking it up to
the natural system, that is the free creation.   
You are laying the formal system as a 
grid over reality, and it may fit better 
or worse depending upon the encoding/decoding 
system that in effect constitutes 
the "laying over of the grid".

One thing that does not appear to be
discussed is that the way that a whole
system or strategy of measurement must
hang together.   Otherwise, the 
measurements are indeed too free.
Your encoding and decoding maps, in
other words, could be arbitrarily 
gerrymandered and/or disjunctive.

The reductionistic aspect of this is also
somewhat disturbing.  Reality is reduced
to a set of human-employed models and
their fit with objectively arbitrary human 
means of perception.   Both of these
problems can be addressed by taking
a more objective view (Reading the
introduction of Thomas Nagel's
"The View From Nowhere" is
helpful in this regard).


> Are you using 'entailed' in some Rosennean notion? I'm
> afraid I can't make any clear sense of what
> 'entailment' means in Rosen's words. I asked Tim Gwinn
> about this, but could you give me a definition of
> 'entailment' in Rosen's terms? To give you an idea, I
> don't know how to look at Rosen's diagram in the book
> (Life Itself) and say the arrows he labels as
> 'encoding' and 'decoding' are 'not entailed in the
> diagram'.

Entailment on the left side is causality, on the
right side, it is inference.  Of course, in general
causality is not going to "commute" with inference,
whatever this means, precisely, since 
causality has a temporal direction, for example,
and has complete, standard "inputs", whereas
inference has no temporal direction or standard
"inputs" (you can infer about the past, or based on
disjunctive information, etc).  This is probably 
why these diagrams, so fondly copied, never 
go beyond the vague  (1) = (2) + (3) + (4) 
stage.   Anyway, to attempt to 
answer your question then, the (2) and
(4) are neither causal nor inferential, hence they
are "unentailed".  

(Think "context of justification" versus 
"context of discovery")

Of course, if you want to become more objective,
you will examine the person as a modeller, and
the processes he goes through, in which case 
his activities of choosing (2) and (4) are entailed
in both senses: they are part of the space of
causes and part of the space of reasons both, to
use Wilfrid Sellars's apt terms.
 
> I don't know if I understand it in the sense you want
> it. If the idea is that the instrument used for
> measure a physical quantity is not part of the system
> being measured, that seems rather obvious to me, but
> I'm afraid I don't see how to relate the metaphysical
> idea of the existence of different objects to the idea
> that a meter or some other instrument is a semantic
> element to a given system, since when I replace m and
> v in 'Et=1/2m*v^2' with the same numbers it gives me
> the same result, independently how I measured mass or
> velocity

Yes, this is related to the point I made before about
the arbitrariness of the human means of perception
when it comes to the models.   Philosophers of science
have been far more sophisticated about this than
Rosen in general, such as for example when they
talk about auxiliary hypotheses being necessary to
draw inferences from a scientific theory.   The
auxiliary hypothesis are part of what makes up
the encodings and decodings in this more traditional
context.  

> As you surely understand, since Rosen said:
> "I have been, and remain, entirely committed to the
> idea that modeling is the essence of science and the
> habitat of all epistemology."

It's hard to see any clear way that Rosen's
discussion of this goes beyond all the discussion
engaged in by philosophers of science in general.

> And the fact he bases controversial claims about the
> epistemological status of sciences like physics,
> chemistry and molecular biology in his notion of
> 'modeling relation', I think is very important to
> define clearly and rigorous the whole idea in order to
> determine if is coherent from a logical point view.

 Good luck!