390 NOTES9 For the argument’s origin, see Peter Van Inwagen, An Essay on Free Will (Oxford:
Clarendon Press, 1983).
10 Strictly, of course, sentences like “The past entails the future” are shorthand for
sentences like “True propositions about the past entail the true propositions about
the future.”
11 That Sue is obligated to do A and Sue does A entails Sue causes B do not entail Sue
is obligated to cause B. Like everything in philosophy, this example – and any other
- will be challenged by someone. If one accepts at least some versions of so-called
deontic logic, the entailments will hold. If they did hold, then this wouldn’t be what
the critic of the Control Principle says it is – an incorrect principle that is analogous
to the Control Principle.
12 That state of affairs A obtains entails that state of affairs B obtains if and only if the
proposition A obtains entails the proposition B obtains.
13 In fact, incompatibilists often hold that were determinism true, no one would, strictly
speaking, act at all, since Necessarily, if A is an action by Karen, then Karen is free
relative to performing A.
14 There are problems with this sort of claim, to be noted later.
15 Strictly, given a set of propositions for which the laws are truth conditions.
16 Or at least to approximate this goal – sufficient ingenuity can fit a lot into this scope,
and we may have left something interesting out. Fortunately, if the account is not
quite as inclusive as it might be, it remains true that the account includes many
important elements, and the inclusion of some other element of the same sort will
not make any difference to our conclusion.
17 For convenience, let “choosing to do A” be a success term – that Jane can choose to
lie only if, given that she so chooses, she lies.
18 Compatibilists sometimes suggest that actions out of character are unfree, though
this would make most heroic action unfree. There are other possible fine-tunings,
but by now the general compatibilist strategy should be clear.
19 Strictly, (HP*) reiterates what (CFa) and (CFb) stated earlier, putting the point in
another way.
20 Of course further refinements are possible; for example, perhaps in the last sentence
one should thing of “A” as replaceable only by terms describing basic actions (roughly,
actions performable without performing other actions in order to perform them).
But this formulation seems sufficient for present purposes.
21 Assumptions made “for the sake of the argument” can be replaced by other
assumptions (e.g., that persons are fully immaterial, or fully material; that causes
are simultaneous with, or are both precedent to and simultaneous with, their effects)
without affecting the force of the argument; the argument’s success as a proof does
not require their being true as opposed to some alternative. Thus the premises so
assumed could be avoided by using complex disjunctive premises; the
cumbersomeness of that procedure motivates the use of “for the sake of the argument”
assumptions.
22 Again, there are behavior-relevant states besides intentions, but here let “intentions”
represent the entire range of states that are behavior-representing and
behaviorrelevant. I will not worry here about sorting out “intention-as-
representational” and “intentional-as-end-seeking” as this is not crucial to the
argument.
23 By “entire set of causes” is meant “the set of phenomena that, given background
condition, is sufficient to yield the effect in question,” leaving aside the difficult
question as to how to mark off background conditions from causes. The argument is
compatible with a wide variety of ways of making this distinction.