Re: Free Will

Omega (omega@pacific.net)
Thu, 30 Jan 1997 01:04:15 -0800


John K Clark wrote:

It seems that the nature of our disagreement is becoming a bit clearer.
Let's see if this can at least identify where we disagree, and why I
disagree with your definition of free will.

1. DIFFERING DEFINITIONS:

> > I agree that your definition is clear and precise (at least in
> > meaning):
>
> Thanks, meaning is what counts. My definition is "A being has free will if
> and only if it can not predict what it will do next".
>
> > The problem is that many things (as in a digital slot machine) can
> > be random (in the broad sense) because of Turing non-predictability,
> > while yet being totally deterministic.
>
> I agree and said as much in my post, but I don't see what the "problem" is,
> at least not for me. You say that a determinist universe has no room for what
> you call free will, not even with Turing non-predictability, so the only
> thing you have left is events without a cause, that is randomness. Sounds
> like a rather poor sort of free will to me, not much worth having.
>
> > The classic definition (in my own words) of free will in philosophy,
> > theology (ackh), and culture at large, is: The ability to select or
> > to not select an action
>
> Then we most certainly have free will, I select things and make decisions a
> thousand times a day.
>
> > free from constraints imposed by external circumstances,
>
> Free? The term you're trying to define is FREE will, I fear we're getting a
> little circular here. I gave you my definition and it is not circular.

It is not in anyway circular, it only seems circular because I didn't make it
clear that my definition of "free" means (and only means) "acausal". I know
that you consider this:

> a rather poor sort of free will

but it is my unambiguous definition. You yourself have accepted this as some-
thing which has meaning in reality, both in your comments on what random may
mean, and in your own statement the other day which said:

"I know of no law of logic the demands that every event have a cause."

> > or by other agencies (such as divine will) that might impose
> > additional constraints not found in the external circumstances.
>
> Vague. How do you distinguish between " additional constraints" and "external
> circumstances"? It all seems like exactly the same thing to me, even divine
> will is just part of the external circumstances, just part of the problem to
> be solved. I gave you my definition and it is not vague.

Again this is the same issue. No vagueness was intended, because by "addi-
tional constraints", I meant ALL constraints that one can semantically define
(as would be consistent with the meaning of acausal). I had hoped to make
this clear in the part that followed the one you just quoted where I described
the evolution of philosophical thought on this subject following the rise of
Newtonian science.

> > one has free will if they are an acausal agent within some realm
>
> Then we have free will, I certainly cause things to happen, but it's too
> broad for my taste, my definition is just a subset of that.

A further clarification here. By "acausal", I mean "acausal" in principle,
not just "functionally acausal" because of Turing-unpredictability. With
this in mind, your definition can not be viewed as a subset of mine, since:

1. My definition requires acausality (leaving aside for now any other
additional things that might also be required such as self-referencing).

2. Your definition requires self-referencing (which I believe to be ontolog-
ically undefined) and unpredictability.

If your definition of unpredictability includes acausality, then our defini-
tions are those of two partially overlapping sets. If it doesn't (and thus
excludes acausality) then our definitions are totally disjoint sets.

I agree that the self-referencing will ultimately be important, but I am not
using it at this time because I feel that no ontological basis has yet been
defined for the concept of self-referencing. I am not nit-picking this issue
just for the fun of it, but because I believe that establishing the onto-
logical basis for self-referencing requires us to penetrate into an area of
the ontologically unknown that we currently perceive as acausal behind which
we may find both a truer form of acausality, and the ontological basis for
self-referencing.

It is an analyis that I doubt will work if we broaden the meaning of the
free-will to include the Turing style unpredictability, as we would then be
confusing mathematical semanitics with physical causal principles; a problem
that would be analogous to dragging the subject of orbital chaos into a
discussion of the principles of gravitation.

> > Where something is predictable, then free will does not exist.
>
> Not according to my definition. I may be able to predict your behavior
> perfectly (provided I don't tell you what it is) and you can do the same with
> me, but we both have free will because we can't predict our own behavior.

The phase "Where something is predictable, then free will does not exist"
is the current operative principle by which MY definition of free will may be
tested against reality. That it is false by YOUR definition is not any big
suprise as our two definitions quite different. The statement here is simply
irrelevant to our discussion until we have an agreed upon definition.

While we're on the subject of the acausal:

> > Or, through some form of trancendent acausality in keeping with the
> > first meaning, some limited form of free will does exist
>
> Trancendent acausality? I have no idea what that means.

Actually it was more redundant than anything, as trancendent causality would
be a form of causality which exists outside of time as we know it. From
some (but not all) metaphysical viewpoints it is synonomous with acausality.
I should have not used the term since the only time it would not be redundant
would be when it is wrong (especially true considering the fact that I don't
even have a solid opinion on the metaphysical distinction between the two).

2. REGARDING YOUR DEFINITION OF FREE-WILL:

2.1. FREE WILL AND TURING:

> > Either free will does not exist at all, in which case this new
> > definition is wholly wrong because nothing that it called free will,
> > was in fact such.
>
> According to my definition, free will defiantly exists, absolutely no doubt
> about it, Turing proved it.

But you definition of free will is based upon the type of inescapable ran-
domness that Turing proved exists (unless it's just simply a self-asserted,
truth) so what could possibly be more circular than to turn around and say
that Turing proved free will as given by this definition?

2.2. FREE WILL AND COMPLEXITY:

> > wondering how much we really know of ourselves.
>
> I don't wonder I know. We know very little about ourselves and will never
> know much more, that's why we'll always have free will. Someday we or our
> offspring will be far smarter than we are now, but the thing they will try to
> understand, themselves, will be far more complex.

Here you say complexity is what will keep our offspring from losing their
free will in the face of their greater intelligence.

> > yet this is what complexity based free-will theory tries to tell us
> > is possible.
>
> I don't think complexity theory has anything to do with it and never said it
> did.

And, here you say complexity theory has nothing to do with it.

Not only do I find distinction between the "theory" of complexity that I
mention and the "actuality" of complexity implied by your above statement
to be somewhat of a fine hair to split, but according to M. Mitchell Waldrop
in his book _Complexity_ on page 234, he describes how complexity researcher
Christopher Langton refers to Turing's halting algorithm while explaining
the behavior of class III and class IV cellular automata.

In point of fact, the type of Turing unpredictability you are talking about
is specifically one of the things that complexity theory studies when it does
this kind of work with cellular automata.

2.3. FREE WILL AND SOLIPSISM:

> > the "actuality" of free will that I'm talking about would (according
> > to your above commentary) change strictly as a result of changes in
> > one's knowledge regarding self and/or environment.
>
> Correct.
>
> > or worse yet solipsism
>
> My definition does not demand solipsism, neither does it disprove it, and I
> don't know of anything else that can do one bit better in that regard.

Here you deny that your definition is in anyway related to solipsism.

> > Consciousness and free will may or may not exist in actuality
>
> There is one consciousness that I know for sure exists, my own. I can't be as
> certain about yours or anybody else.

Yet here you make a statement that could perfectly be described as solipsistic
according to the number one definition of the word in my dictionary.

2.4. SELF-REFERENCING:

> > it leads to local definitions of truth
>
> No. I can always predict what I am going to do or I can not, and that is true
> hear and there and everywhere.

> 1) I predict that in 2 minutes or less I will be inside my house.
> 2) I find that my key does not work.
> 3) My "prediction" was not a prediction at all, it was wrong.
> 4) I have free will.
>
> Of course, sometimes we can make correct prediction about out behavior, and
> sometimes we feel like we've fallen into a rut and are turning into robots.

The problem I see here is that statement 4 is just an assertion that could
be made without even bothering with statements 1-3.

I can see the meaning behind your definition, but since your definition of
free-will is obviously that which a self-referencing entity makes about
itself, it is hard to see how it could ever be wrong. Equally difficult
to see is how it could ever be useful for anything.

2.5. GODELIAN INCOMPLETENESS:

Since the inevitableness of self-referenced Turing unpredictability is really
just another another way of saying that all systems suffer from Godelian
incompleteness with regard to the subject of their own behavior, how can we
turn around and then claim it to be some kind of causal principle by which an
entity would effect its "will" within the world?

I see two possible answers to this question:

a. Your definition of free will is so mutant that it no longer even claims to
be something that represents a causal principle by which an entity would
effect its "will" within the world, and thus amounts to a tautology that
expresses nothing more than the Godelian incompleteness of a self-referen-
cing entity regarding its own behavior. Tautology definitions:

1: Needless repetition of the same sense in different words.
2: A logical statement which can never be false.

b. Or you really are claiming that Godelian incompleteness is some kind of
causal principle by which an entity would effect its "will" within the
world, thus bringing Cartesian dualism back from the grave, in drag.

3. REDEFINITION OF WORDS:

> > Unfortunately your definition is the logical negative of the classic
> > definition because when stripped of the self-reference (which I will
> > get to shortly) it reduces to an operative principle which says:
> > ? Where something is not predictable, then free will does exist.
> > It is a well established fact in formal logic that the negative of a
> > logical statement has a different meaning from the original, and
> > furthermore derives no support as to its validity from the original
> > statement.
>
> Please note, I said "if and only if", A being has free will if and only if
> it can not predict what it will do next. If A = B then it is also true that
> not A = not B, thus because it means exactly the same thing, I am of course
> perfectly happy with the following statement: A being does not have free will
> if and only if it can predict what it can do next.

The meaning of your "if and only if" is quite clear, but that doesn't change
the fact that taking the logical negative of the classic definition changes
the meaning; something which your next statement shows is your intent anyway:

> > This redefinition thus leads a couple of problems: The first is
> > that it changes the the established meaning of a term which goes
> > back millenia in human culture.
>
> I sure hope so, because the established meaning of the term is, well...,
> it has no meaning, its not even wrong, its gibberish.

Unfortunately, the established meaning is not so incomprehensible. I'll even
bring in self-referencing to try to close the gap between the two meanings by
presenting the classical meaning in a somewhat narrower way along with yours:

1. Classical: A self-referencing locus of the acausal.
2. Yours: The state of self-referenced unpredictability.

I think it is unreasonable to say that the classical meaning is not: both
meaningful and distinct from yours. I'm sure a great many people would see
it as having a crystal clear meaning.

In any case, having an agreed upon meaning for words is already a difficult
enough problem in our society without people going around redefining words
on the fly. Considering what this kind of redefinition does to the signal
to noise ratio within our society's communication, there really is no excuse
for not coining a new word or phase for capturing a meaning which is as
distinct as this.

> > >John:
> > >Yes, I've found that nearly everybody I meet on the street has
> > >exactly the same opinions I do.
>
> > Omega:
> > This is not at all my experience.
>
> Holy cow, what a revaluation! Let me write this down before I forget it.
> The people who sit next to me on the buss may not have exactly the same
> opinions I do about Anarchy, Atheism, Nanotechnology, Uploading or Turing's
> solution of the Entscheidungproblem and its relationship to free will.
> Thank you for the insight you have given me about my fellow man.

You were the one who mentioned "everybody I meet on the street" in the
first place, but again, that doesn't change the fact that the redefinition
you espouse is being hyped by one very well defined and fairly narrow
segment of our culture to the detriment of the general ability of our
culture to communicate within itself.

-- 
In the Ecstatic Service of Life -- Omega