Re: ontological levels

["Tim Gwinn" <***>]

Roberto,

See below.

>
> Tim,
> see below a couple of short comments.
>
> (skip)
> > My concern is that any system under study, such as a "bacterium" or a
> > certain "social phenomenon", are subjectively defined.
>
> Let us consider the bacterium case (the question of "social phenomenon"
> is much more intricate)
> I cannot imagine that you really believe that the bacterium's
> dimensions, structure and properties
> are "subjectively defined". They are not a product of our phantasy,
> imagination, or whatever.
> They are there, in themselves, and we only try to capture as many as
> possible of their properties.


I think we might be using the phrase "subjectively defined" in two different
ways. By subjectively defined, I do not mean to imply phantasy or delusion
or the like. Rather, 'subjective' in the sense that any given collection of
observables to be taken as the system under study will be a collection that
is of *our choosing*. Again, I am not implying that any of these observables
are somehow imagined or illusory. In the case of the bacterium, it feels
quite natural to consider the collection of those observables that arise
within the spatial dimensions of what we consider to be the boundary of the
form or structure as being "the system" - which we define as "bacterium".

However, to the extent that the system thus defined has ongoing interactions
with (what is now defined as) that system's environment means that what we
have called 'system' (in this case, 'bacterium') is not innately isolable.
And therefore our demarcation of the edges of that system (and, accordingly,
the demarcation of sets of observables) is an abstraction. This is not a bad
thing - we cannot do science without making this kind of demarcation of what
we propose to study and model.

My question is: what grants one system definition preference over another?
In this case, why not choose a larger set of observables that extend beyond
the cellular wall, for example, to some other spatial point? Certainly,
there are generally good pragmatic reasons (e.g., mathematical tractability)
for choosing to define a system's boundaries in accordance with some
observed structure(s); but how does use of that criteria (system boundary =
structural boundary) tell us something ontologically fundamental? (I don't
mean that entirely as a challenge - I really feel I am missing some basic
point.)

I suppose another way to ask it is this: what allows us (or directs us) to
assert that the existents (the object "out there") have some innate 'system
nature', if you will, that coincides with our human manner of systematizing
the available observables (which we presume to derive from those existents)?

For Rosen, as I understand him, he considered observables as our basic
starting point. We posit that there is a world of existents behind those
observables, but those existents are essentially Kantian noumena for us -
not knowable directly. Starting then from only these observables, it is we
as humans who systematize them in our minds and in our models. It is
plausible to impute back to the world of existents the relations we
apprehend between observables. If we considered our system definition as a
kind of relation between the observables comprising that system, then it
would seem we could likewise impute that relation (that "system-ness") back
to the world of existents. However, since the choice of which observables
comprise a system is our choice, then I do not feel that this relation of
"system-ness" is a basic property which can be imputed back to the world of
existents.

As Rosen says:
        "The notion of system-hood is at that same level of generality [the notion
of set-ness in mathematics] and plays the same role in our management of the
ambience. As noted, it segregates things that "belong together" from those
that do not, at least from the subjective perspective of a specific self, a
specific observer. These things that belong together, and whatever else
depends on them alone, are segregated into a single bag called system;
whatever lies outside, like the complement of a set, constitutes
environment.
        "The partition of ambience into system and environment, ***and even more,
the imputation of that partition to the ambience itself as an inherent
property thereof*** [my emphasis], is a basic though fateful step for
science. For once the distinction is made, attention focuses on system.
Systems and environment are thenceforth perceived in entirely differnt ways,
represented and described in fundamentally different terms." [LI p. 42]


Regards,
Tim