[extropy-chat] Role of Observer is not Relevant

[Jef Allbright]

On 3/27/07, Lee Corbin <lcorbin at rawbw.com> wrote:
> On 3/27/07, Jef Allbright <jef at jefallbright.net > wrote:
>
> > To amplify Russell's remarks somewhat, there's an assumption running
> > through this thread that there can be "information" without an
> > observer, thus the talk about "information flowing" during a "causal
> > process", but not in the case of a lookup table. There's an essential
> > subjective element that's being ignored here.  Just as any pattern of
> > bits may be validly said to "be random" or to "contain information",
> > it depends on the observer.  Without specifying the observer, the
> > statement is meaningless.
>
> It's quite possible that I, or Stathis, or anyone is not deciphering this
> passage correctly, but the picture I'm getting is that to you we must
> specify an observer as a integral process of any primary process
> that we wish to describe.  If true, this is going to make for a lot
> of awkwardness.  (So I hope it isn't what you mean.)

To be meaningful, any statement necessarily entails a context, and
context is necessarily subjective; it always only represents a partial
view.  Is the following sequence random or is it meaningful? Depends
entirely on context.
"01101101011001010110000101101110011010010110111001100111011001100111010101101100"

In everyday conversation the context is specified implicitly.  It goes
without saying to the extent that we share a common view of things.

In scientific communication, every effort is made to specify the
context as explicitly and completely as practical, while recognizing
that it can never be complete.

When philosophizing about the limits of meaning this inherent
subjectivity becomes most critical, even if not apparent.  And when
the philosophical subject turns to subjectivity itself, people often
embark down the recursive rabbit hole futilely looking for an exit
that can't possibly exist.

> Instead of just describing a process, or an object, in our language L,
> must L also include an outside perceiver of said process?  Bishop
> Berkeley has returned!

Your "just describing" entails encoding within a highly organized
context as described earlier.  As a programmer, you certainly realize
that your meaningful high-level formulation must be interpreted or
compiled to a much broader sequence of machine language before it can
be executed by a much broader sequence of transistors switching,
implemented by vast patterns of charge transfer in a semiconductor
lattice (within a pattern of layers and interconnects), consisting of
atoms... expanding on its way to objectivity and meaninglessness.  The
most meaning was at the highest level of encoding, and meaningful only
to you or some similar observer.

George Berkeley's epistemology was idealist (as yours seems to be. I
recall you favor the idea of platonic existence of numbers, which is
symptomatic of the same mindset.)  Mine takes a pragmatic form.


> I might be unable to say, "the car is traveling down the road" without
> also discoursing on who is observing the car. (Now I'm sorry again
> if I'm misinterpreting you, because the following will just be more
> overkill.)  Of course this also leads to infinite regress, because how
> are we permitted to use a language L that includes a description of
> a process P and also includes the observer O of P?  That would
> clearly not be adequate, since L ought also to include O', the
> observer of P+O.
>
> Instead, the language of simple realism is vastly preferable, along
> with the realist postulate that things may exist without being observed.

Lee, you describe the process but you veer off at the last moment and
deny the implications.  Rather than compromising on "simple realism",
one can fully embrace subjectivity with no such compromise. [More on
this further below.]

As a pragmatic realist, I am saying that we achieve best results by
considering something to exist only to the extent that it is observed,
and observation is always only indirect.

I expect that you value Occam's Razor, which principle is at the core
of what I'm saying here.  William of Occam, along with mathematical
models such as maximum entropy, would strongly suggest that you
abstain from postulating entities that may exist without being
observed.


> Stathis remarks to this:
>
> > Indeed. [what???]   But it gets weird when the observer is himself
> > the product of the information, bootstrapping itself into a self-awareness.

It only seems weird when one tries to follow the rabbit hole to see
where it leads.  It's quite naturally a recursive rabbit hole (due to
the subject being inherently subjective) so it only goes deeper, but
never exits.  Although we're surrounded by recursive processes in
nature, most people never internalize the concept.  Even many (if not
most) programmers of software never get comfortable with recursion
even though it is supremely powerful and elegant as a means of
expression (when appropriate, and notwithstanding the associated
issues of finite machine resources.)

> Has there been any miscommunication here?  I get the feeling that, to
> use my description above, Stathis has reverted to talking about the
> process P, WHICH JUST SO HAPPENS, QUITE ACCIDENTALLY,
> to contain an (impotent) experiencer, who is, shall we say, only reflecting
> on certain memories and abstract thoughts and isn't perceiving anything
> outside of himself.

I found it very difficult to parse the preceding paragraph, but one
element stood out:  Rather than saying it "contains an experiencer",
can you see it as "expressing an experiencer"? Do you see an essential
difference?

- Jef