>>> Compatibilism (the thesis that free will and determinism are com-
>>>patible, ie. that both can be simultaneously true) is crud. If we
>>>define determinism as:
>>>
>>> `The theory that the state of a system at some time and the rela-
>>> tions governing the time development of the system determine, un-
>>> ambiguously, the state of the system at any later time.
>>This definition presupposes that "the state of a system at some time"
>>has meaning. According to relativity, the notion of simultaneity is
>>dependent upon the observer.
> What exactly is your point here? If "the state of a system at some time"
>does _not_ have meaning, then determinism does not have meaning (which would
>also mean that any negatively defined indeterminism does) and voluntarism
>has some major obstacles cleared out of its way. Including compatibilism.
>Which would make my day, but not, I think, yours.
I have no trouble with determinism. But I view it as a model which is
useful for making predictions, rather than as some absolute truth.
> Anyway things are by no means this simple. The success of various
>deterministic theories and predictive systems that we have indicates that
>we can _do_ things with the notion of "the state of a system at some time".
The above sentence seems to support determinism, at least in the sense in
which I view it.
>Furthermore if `system' is defined locally (so as not to violate
>any prohibitions against superluminal influences) the definition stands just
>fine, and enables the interaction of different `systems' to be
>deterministically described. (The old `surface of a light cone' trick.)
Keep in mind that the light cone depends on a particular point in
space-time which you use as the vertex. If you use the light cone in
your definition the most that you can say is that the state determines
the system up until that particular point. Once you advance further
into the future effects from outside your original cone begin to be
visible.
>>> And further is we define voluntarism (the thesis that we have `free
>>>will') as:
>>>
>>> `The theory that a person is free with respect to same act if they
>>> both can perform it and refrain from performing it.'
>>
>>In most cases, a person cannot "both" perform and refrain from performing,
>>since one often excludes the other. Perhaps you should be using the
>>word "either". But, in any case, the statement depends on the uncertain
>>semantics of "can".
>
> At the risk of being repetitve _that_ is precisely the point. `Both' is
>exactly the correct word (remember I am defining the traditional version of
>a doctrine, not defending it) since that makes clear the presence of
>multiple possible outcomes.
And at the risk of being equally repetitive, I repeat this is impossible,
unless you are talking about a "multiple worlds" view in which the world
is constantly splitting into many worlds each with a different set of
choices. Perhaps we have different meanings for the word "both". I
always thought that "both" meant both. You seem to imply that "both"
means you have a choice to pick either, which is quite different.
> One clearly _must_ exclude the other, otherwise
>there would be only one future, and hence no `decision' or `possibility.'
Again, unless you are taking a multiple worlds view, there is only one
future. Historians will eventually record what that future is, although
at present we do not know. The question of free will is not whether there
are multiple futures, but whether that future is already pre-determined,
or whether we have some way of influencing it.
>>> The question, therefore, is _which_ of the two is true.
>>
>>This insistence on either/or, this belief in the existence of some
>>absolute truth, is at the heart of your problem. There is no reason,
>>other than religious belief, to suppose that the real world works this
>>way. Rather than asking for the absolute truth, you should be asking
>>about which is a useful interpretation of the available evidence.
>>Compatiblism maintains that determinism and free will are both useful
>>and compatible interpretations.
>
> Wow! This final paragraph is so mind boggling that all I can do is ask a few questions and hope for subsequent illumination. I can identify no `point' to which I can be expected to reply/respond.
>
>(1) Please tell me _which_ religious belief you are thinking of (or do you
>maintain that they all involve some kind of `principle of non-contradiction',
I'm not suggesting any religious belief. However it is a property of
many religious beliefs that you can "know" something for which there is
no empirical evidence.
>(2) I want to have lunch. I have x$ there are two different kinds of food
>available at the canteen, each of which costs x$. The question, therefore,
>is _which_ of the two I can buy. What has this to do with religion? I
>think (in my fundamentalist mania) that when two courses of action are
>mutually exclusive, that only _one_ can be followed, that when two
>propositions are mutually exclusive that at most _one_ can be true. To
>quote _Dire Straits_:
This is all well and good when talking about formal systems. I'm saying
that it has no certain applicability to the real world. What is likely
to happen is that when you get to the canteen you find that one of the
two kinds of food is sold out, and the other one was not prepared in
accordance with your expectations. Worse still there is a sales tax
which will be added to your x$.
You can use a formal system to model reality. Then you can use arguments
such as yours within that formal system. But you are still only proving
something about the formal system, so that whether your conclusions are
valid about reality depends on the accuracy of the formal system as a
model. This can only be determined empirically.
To put it in other terms, if there is such a thing as an absolute truth
then it is unknowable. I am questioning whether it makes any sense to
presume that a truth exists if it is unknowable.
>(4) What is the `evidence' if not the effects of the way the world _really_
>is? What is an `interpretation' of that evidence if not an attempt to make
>sense of it in a way (pending your explanation of `useful') the sucess of
>which is related to the accuracy of the fit with the way the world is?
I have no essential disagreement with this. But keep in mind that that
the model on which you build your interpretation may be an imperfect
fit. The best you can hope for is that it fits the evidence within the
error of measurement. You can have no assurance that there is a perfect
fit.
>(5) Following from (4) what is a model which _does_ fit properly if not the
>`absolute truth'?
How could you ever know if a model fits perfectly, or if there even
