More
posting, alas.
Kylie wishes to prove that free will is incompatible with determinism, and as an example she posits an omniscient agent who knows in advance everything that I will do. Such an agent, while fictional, is a useful symbolic stand-in for determinism in a thought experiment.
She hopes to establish, I guess, that if this agent knows
infallibly in advance that Kylie will eat chicken for dinner, then Kylie will eat chicken.
I agree! Kylie will eat chicken, in this scenario. No doubt about it.
But here is where things get slippery. Kylie, like all incompatibilists, now makes an illicit move. She purports to show in this scenario not JUST that Kylie WILL eat chicken, but that Kylie MUST eat chicken. And that is the point where the incompatibilist goes off the rails.
As I tried to show in the post that Kylie has clearly misunderstood, there is no reason to accept that Kylie “must eat chicken.” The erroneous argument condenses to this:
If an omniscient agent knows in advance that Kylie will eat chicken, then Kylie must (necessarily) eat chicken (no possible alternative, no free will for Kylie).
As I explained in the post in question, the above commits the modal fallacy, which I have invoked in different contexts. Since Kylie eating chicken is a contingent (could have been otherwise) matter, then her eating chicken can never be a
necessary truth. Only necessary truths — e.g., that triangles have three sides — can never be otherwise. There is no (logically) possible world at which triangles have four sides, for example.
As I explained in my post using a different example, the corrected argument for Kylie and chicken goes like this:
Necessarily, (If an omniscient agent knows that Kylie will eat chicken, then Kylie will [but not must!] eat chicken).
The upshot that in the presence of an omniscient agent, Kylie is free
to eat whatever she wants. That is, she has free will — she has live options!
What if Kylie chose to eat liver (ugh!) instead of chicken?
No problem!
Then we get:
Necessarily, (If an omniscient agent knows that Kylie will eat liver (ugh!), then Kylie will [but not must!] eat liver (ugh!)).
Kylie can eat what she wants! What she can’t do is evade
detection of her choice, even before she makes it, by an omniscient agent. But the choice is clearly there. Kylie’s free choice supplies the
truth grounds of what the omniscient agent knows in advance. If she chooses chicken, the OA foreknows chicken. If she chooses liver, the OA foreknows liver. That’s it!
Turning to determinism, this is the crux of the dispute between compatibilists and incompatibilists. The incompatibilist wants to say that given antecedent circumstances x, y, and z, then Kylie MUST eat chicken — no free will.
It’s the same illicit move as discussed above! The compatiblist says only (and correctly!) that given antecedent circumstances x, y, and z, Kylie will [but not must!] eat chicken. Kylie is free to eat what she wants. What if she chooses to eat liver (ugh!) instead? Then there would be DIFFERENT antecedent circumstances — a, b, and c, perhaps.
The difference between WILL and MUST makes all the difference in the world!
Please observe that (most) compatibilists agree with incompatibilists that under
the exact same circumstances, Kylie will eat chicken. The compatiblist just points out that since Kylie does not HAVE TO eat chicken, determinism poses no problem to free will.
The compatibilist correctly cashes out “could have done otherwise” to mean, “would have done otherwise, if …” If what? If circumstances had been different! Which makes perfect sense. We think in these terms all the time!
Marvin goes to the restaurant and orders a salad for lunch. In so doing he remarks to his dining companion, “I had a huge breakfast, so I’m going to have a light lunch. But I can tell you the steak is really good here. If I had skipped breakfast, I would order the steak.”
Free will and determinism, in compatible action together.