298 Revisionism and Some Remaining Issues
to be locked and impassable, it also seems that I cannot deliberate about
which door to open – or even which door to decide to open. (Nelkin, 2011:
130; cf. Kapitan, 1986: 247)But I am neither certain that I will open door #1, nor that I will not, and the same
for my opening door #2, so in this situation the agent satisfies our representative
epistemic condition. And here is the connection to a belief in determinism. If I
believed determinism and its consequences, then in any deliberative situation I
would believe that all but one option for what to do was closed off—that all but
one would be “locked and impassible.” Given that in van Inwagen’s example I
cannot rationally deliberate about which door to open, believing determinism
and its consequences would also seem to preclude rational deliberation between
options. A compatibilist account would need to explain why rational deliberation
is impossible in the two- door case, but still possible for the deliberator who
believed in determinism and its consequences.
In response to this challenge, Tomis Kapitan (1986) argues that on the prefer-
able compatibilist position there are two epistemic conditions that the deliberator
must satisfy (Pereboom, 2008a, 2014 agrees and fine- tunes Kapitan’s proposal).
One of these conditions expresses an epistemic notion of openness for what to
do. The other specifies the epistemic condition on the efficacy of deliberation.
For example, the compatibilist might require that to rationally deliberate
about whether to do A or B, the deliberator must believe that if as a result of her
deliberating about whether to do A or B she were to judge that it would be best to
do A, then she would also, on the basis of this deliberation, do A; and similarly
for B (Pereboom, 2008a, 2014). This condition is not satisfied in the two- door
case, because the deliberator does not believe for each door that she would open
it as a result of her deliberation- based judgment that it would be the best one
to open. So the deliberative efficacy condition gives us the result we want about
this case.^8
The debate between deliberation- incompatibilists and deliberation-
compatibilists continues to be lively, with interesting and strategic moves made
on both sides.
12.4. Experimental Philosophy and Free Will
The free will debate is in part an empirical affair, and this has given rise to
experimental work—X- phi—on free will. X- phi employs social science tech-
niques, most prominently surveys and analysis of their results, to discern peo-
ple’s intuitions and judgments about a range of philosophical topics such as free
will and moral responsibility, intention, personal identity, and knowledge.
Important X- phi work on free will has been done on the issue of beliefs about
compatibilism and incompatibilism,^9 on error theories about folk intuitions on
this topic,^10 on manipulation arguments,^11 and on whether our criteria for holding
morally responsible are the same across cultures and across types of situations.
This last issue is the X- phi area we will examine by way of illustration.