dennisf@teklabs.UUCP (08/29/83)
Recent discussion of free will has prompted me to bring up Donald MacKay's argument, which I've not seen yet on the net. First, as a parenthetical comment, I see little hope in arguing the possibility of free will from Heisenberg Uncertainty since: 1. Brain function occurs at a macro level, and apparently does not require quantum explanations any more than satisfactorily complete physical explanations of other common physical phenomena. 2. Freedom of the will is recognized by us as conscious agents and it is at the level of agency rather than the molecular level of organization that freedom of the will is normally attributed. For all we know, molecular or sub-atomic randomness would reduce our freedom, which requires reliable brain operation, not enhance it. Now for MacKay's argument: this argument is based on logic, not brain physiology, and deals with what would follow from the assumption that our conscious experience correlates with the activity of our cognitive mechanism (or, loosely speaking, our brain). MacKay shows that even if our brains were as mechanical as cash registers, it would still be incorrect for us to assume that we were *not* free in a certain precise sense. That is, if by freedom we mean that there cannot exist predictions of our willful behavior which we would be both correct to believe and incorrect to disbelieve, then MacKay's argument shows that such freedom is simply a logical fact. Suppose some super-scientist were able to make entirely reliable predictions on the basis of brain science, etc. of our mental activity that we associate with freedom of the will. He makes the predictions and watches ... and every time, they are correct. Now, what would be the logical status of those predictions if he were to offer them to us. Would we be correct to believe them just as he was correct to believe them when he kept them to himself? For the super-scientist they are necessarily true, but for us they would not be since these predictions fail to take into account all of the data; namely, they fail to account for our believing them. So that once we believe such a prediction, we invalidate it as necessarily true. We end up believing something which, upon believing it, becomes false. Ah, you say, if the super-scientist were to take into account our believing the prediction before he offers it to us, then when we believe it, it *will* necessarily be true for us also. Yes, but we then would not be incorrect to disbelieve it, for in disbelieving it we fail to satisfy one of the assumptions upon which the prediction was based, namely, our believing it. The "you" predicted in that case would not be the "you" who disbelieves it. Therefore, there cannot exist a prediction of our choices which is inevitable for us; that is, is true for us whether we believe it or not, if only we knew it. The truth of the prediction depends upon whether we believe the prediction or not, and that choice is up to us. So if you accept this definition of freedom of the will, it is simply a logical fact that we are free, and it has nothing to do with the determinateness or not of our brains. You see, the confusion comes in assuming that predictablity-by-an-observer is logically the same as inevitability-for-the-agent described by the prediction. For more on this argument, see Donald MacKay's book *The Clockwork Image* or his more recent one: *Brains, Machines, and Persons*, the first put out by IVP, the second, I recall, by Eerdmans. Dennis Feucht Tek Labs
laura@utcsstat.UUCP (Laura Creighton) (09/01/83)
Can I get another word in? I have no idea about an eternal Is, but I do know that there are a lot of funny things that happen in the brain and big problems with any theory that tries to explain what consciousness is. Qualifications: This is not my field. It is, however, my fathers. I will relay all comments to him to see if he wants to say anything. He may not, however, since he personally does not think that religion should be provable, and therefore avoids such arguments generally. 1: You can not talk about a physical model of consciousness without getting into the nitty gritty details of the chemistry of the brain. 2: Some chemicals in the brain produce very noticable results in very small quantities. 3: The motions of the molecules of the chemicals in the brain as they attain chemical equalibrium can only be described in terms of probabilities. 4: Thus, unless our understanding of chemistry, or probabilities change (possible, I guess, but highly unlikely at this date) one can only talk about the probability of neuron X being stimulated. 5: Thus, you can not build an entirely deterministic model of consciousness even if you could map the value of the firings (and the intensity of the firings) of certain neurons into their results. (this is an oversimplification of what one would actually have to do) As far as I can see, the determinists have had it. This does not mean that the free will people have emerged victorious, but I personally have trouble thinking of a middle ground between 'free will' and 'determinism'. Entirely random behavior would be one, but THAT doesn't correspond to the world I recognise. Perhaps at some level things are deterministic but at a lower level it is random. I find the boundary condition very hard to think about, however... laura creighton utzoo!utcsstat!laura
larry@grkermit.UUCP (Larry Kolodney) (09/01/83)
From Laura Creighton:
3: The motions of the molecules of the chemicals in the brain as they attain
chemical equalibrium can only be described in terms of probabilities.
4: Thus, unless our understanding of chemistry, or probabilities change
(possible, I guess, but highly unlikely at this date) one can only talk
about the probability of neuron X being stimulated.
5: Thus, you can not build an entirely deterministic model of consciousness
even if you could map the value of the firings (and the intensity of
the firings) of certain neurons into their results.
(this is an oversimplification of what one would actually have to do)
The reason you can only talk about the molecules in the brain in terms
of probaility is because it is impossible for us to observe with any
accuracy the individual molecules. Thus, we talk about the 'average'
molecule. This rule is made absolute by the uncertainty principle,
which says that no matter how good our measurement instruments are, we
can never get a high resolution of both momentum and position. Without
those data, we cannot predict the value of those data in the future,
except on a probabilistic basis.
That doesn't mean the path of an individual molecule is not
predetermined, it just means we can never predict it. THus, for all
intents and purposes, the path IS probabilistic. BUt the actuality of
the matter can never be know. (anyone care to flame about the
difference between perceived reality and reality?)
--
Larry Kolodney (The Devil's Advocate)
{linus decvax}!genrad!grkermit!larry (until Sept. 8)
(ARPA) lkk@mit-ml (after sept. 1)