29 September 1997
Thank you for your letter of 13 September responding to my comments on your notes on unit 1 of The Possible World Machine, and also for your notes of 19 September on unit 2. This letter has missed my 'seven-day deadline'. The time I spent working on the Pathways Web site has left me with a back-log of letters! I expect I shall catch up soon; sorry for the delay.
On the question of varieties of possible worlds, it is important to stress the logical difference between the 'world' of this or that person's beliefs and the actual world, or the way things are irrespective of what this or that person believes. One might say that, since no two persons' beliefs are exactly the same, we all inhabit different belief worlds but the same actual world.
Seen in this light, the claim that, in certain circumstances, the belief that magic is effective can be sufficient to give rise to a world where magic is actually effective emerges as. a pretty radical claim. I would have to fill in quite a lot of anthropological detail to make such a claim plausible. It is not simply a matter of spells 'working' by auto-suggestion. That is not magic but merely psychology.
— Do continue to respond to my responses. You've certainly got into the swing of things!
Now to unit 2:
1. 'It is hard to make a similar case for free will.' I assume that what you mean is that it is far easier to find evidence that our decisions are influenced by our past, than it is to find evidence that we are not influenced, for the simple reason that the belief that there is no influence may be due to ignorance. Even if as the result of exhaustive statistical analysis I found no discernible bias or pattern in my choice of lunch time beverage (say, I always choose Coca Cola, Lilt or Tango) it does not follow that today's choice was not causally influenced by factors I am unaware of.
2. Right. I cannot prove that a certain action is not determined, but, equally, I cannot prove that it is, once one grants that the thesis of determinism is not necessarily true. 3. 'I'm surprised you handed that twenty pound note into the police station.' 'What? You should have known me better than that!' — We do expect others to be able to predict our actions in such cases; yet there is no implication that in doing them we are somehow 'less free'. However, more needs to be said here. You can say that despite the argument against free will, the belief in free will remains necessary; but if it is actually false that we have free will, then that necessary belief is merely a necessary illusion.
4. I don't see anything inconsistent in your description of a 'world of total determinism'. You have merely outlined a theory of punishment based on deterministic lines. 'Responsibility' is a notion defined in terms of a 'freedom of choice' possessed by someone in their right mind, not under coercion etc. (You punish in order to 'alter the controls' or else to put a dangerous person out of circulation.)
5. I don't see this. The 'compatibilist' view just outlined finds room for a distinction between free/ unfree agents or actions in terms, say, of a person's ability to respond to criticism of his or her actions. There is a sense of 'he could have done otherwise' which is compatible with determinism. Now, if determinism is true, then our innate knowledge is determined (say, by evolution). Perhaps this means that we are born with a certain character or dispositions. Then it's sheer bad luck if a weakness in my character leads me to a life of crime. — So what should we conclude from that? (I may have missed the point here.)
6. I admit I do feel uneasy about the soul. The argument against free will is a dilemma: If determinism is true then we're all machines. If determinism is false then we're all roulette wheels. Does belief in the soul present a third overlooked alternative in terms of this argument? I would argue that the soul, if it exists, has a nature, and according to that nature either the relation between input and output — or experiences and decisions taken — is deterministic or it is not deterministic. What other possibility is there? Tell me.
7. My chess computer has 'free will'. When choosing an opening, or how to respond to an opening I have played, it does not always make the same move but selects among a group of alternatives using a weighted randomising device. (The choice doesn't have to be weighted, but like a good chess player the computer chooses risky or unusual openings only occasionally, to 'spring a surprise'.) — What does this show? Simply that it's no good speculating how a deterministic universe could 'programme' free will unless you can explain how this would differ from simply saying that the human brain has evolved similar randomising devices. (As it surely must: otherwise you would be stuck in the supermarket like Buridan's ass with your hand poised equidistant from two identical tins of baked beans.)