Your entropy analogy is good - it helps disambiguate between choosing in accordance with strong pressure (what you're saying is wrong with the Christian message) and choosing against strong pressure (our fictional hero fighting against the odds). I think my post may have conflated these slightly.
Can I clarify what you're saying: are you saying that, for a given unbalanced choice (i.e. one with strong pressure in one direction), we can't know a priori how free it is until the decision has been made, and then if the person chose the "difficult" option they did so freely, but if they chose the "easy" option they did so much less freely? Or are you saying that, if the choice is known to be unbalanced, we can say a priori that it's not very free, because most people will choose the easy option and that will average out with the few who choose the difficult option?
(The answer may be obvious if I were more familiar with the maths in the entropy analogy, but I hadn't actually met that formula before.)
I think there are logical difficulties with the former. Suppose you have a smoker trying to quit. Clearly it's more difficult for him to resist a cigarette than it is for a lifelong non-smoker. But if he successfully resists and you congratulate him on his free choice against the odds, you can't also say, if he gives in, "It was inevitable, he wasn't free, he couldn't have chosen any differently." Surely both possible outcomes have to be as free or unfree as each other? (I don't think free/unfree is the same thing as easy/difficult.)
If it's the latter, that makes sense, but I think it's a statistical statement which relies on there being multiple instances of the choice. (You can encode 1000 flips of your biased coin in fewer than 1000 bits, but you still need 1 bit to encode 1 flip, regardless of whether it came up the expected way or the other way.) So it depends whether you think every choice that every person has to make is unique. I think the case is arguable either way, but if they do turn out to be unique then I don't think the statistical principles can be applied, in the same way that you can't use gas-diffusion laws to predict the motion of a single particle.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
timeasmymeasure
no subject
Date: 2008-05-29 03:18 pm (UTC)Can I clarify what you're saying: are you saying that, for a given unbalanced choice (i.e. one with strong pressure in one direction), we can't know a priori how free it is until the decision has been made, and then if the person chose the "difficult" option they did so freely, but if they chose the "easy" option they did so much less freely? Or are you saying that, if the choice is known to be unbalanced, we can say a priori that it's not very free, because most people will choose the easy option and that will average out with the few who choose the difficult option?
(The answer may be obvious if I were more familiar with the maths in the entropy analogy, but I hadn't actually met that formula before.)
I think there are logical difficulties with the former. Suppose you have a smoker trying to quit. Clearly it's more difficult for him to resist a cigarette than it is for a lifelong non-smoker. But if he successfully resists and you congratulate him on his free choice against the odds, you can't also say, if he gives in, "It was inevitable, he wasn't free, he couldn't have chosen any differently." Surely both possible outcomes have to be as free or unfree as each other? (I don't think free/unfree is the same thing as easy/difficult.)
If it's the latter, that makes sense, but I think it's a statistical statement which relies on there being multiple instances of the choice. (You can encode 1000 flips of your biased coin in fewer than 1000 bits, but you still need 1 bit to encode 1 flip, regardless of whether it came up the expected way or the other way.) So it depends whether you think every choice that every person has to make is unique. I think the case is arguable either way, but if they do turn out to be unique then I don't think the statistical principles can be applied, in the same way that you can't use gas-diffusion laws to predict the motion of a single particle.