158 Three Source Incompatibilist Arguments
features of the case appears to give the advocate of the Direct Argument room
for a reply.
Consider first the fact that the example involves moral responsibility for con-
sequences. A defender of the Direct Argument might point out that moral respons-
ibility for consequences is always in some way derivative of moral responsibility
for action (or omission), and it is not surprising that a Transfer NR principle not
restricted to actions (and omissions) could be easily discredited by examples like
Avalanche (Pereboom, 2001). What matters, however, is whether nonresponsibil-
ity for past and laws, and nonresponsibility for their leading to an agent’s acting as
she does, lead to nonresponsibility for those very actions. A case like Avalanche
leaves this basic source incompatibilist worry entirely unaddressed.
A critic of the Direct Argument might reply by constructing examples con-
cerning responsibility for basic actions, such as decisions, that have the same
structure and so would threaten a version of Transfer NR restricted to actions
(e.g., Fischer and Ravizza, 1998: 157–8). The resources for doing so are, as it
turns out, familiar. Consider this Frankfurt- type example:
Imagine that Betty deliberates and decides to destroy the enemy basecamp
by T3. Unbeknownst to her, there is a nefarious demon who will cause Betty
to so decide if at the very moment Betty actually decides she were not to do
so. As it happens, Betty decides on her own.
This has the same structure as a counterexample to a pertinent version of
Transfer NR as applied to actions:
- Betty is not morally responsible for the presence and the intentions of the
demon. - Betty is not morally responsible for the fact that, if the demon is present,
she, Betty, will decide by T3 to destroy the enemy camp. - Betty is morally responsible for deciding by T3 to destroy the enemy camp.
True, critics of Frankfurt’s argument who dispute the coherence of this style of
example will not be persuaded by this counter- reply. But set that aside here.^15 It
seems that such an example has the right structure to show that pertinent ver-
sions of a Transfer NR principle, even ones restricted to basic actions (and omis-
sions), are invalid.
Now consider the fact that the examples employ two- path cases. An advocate
of the Direct Argument might object that the examples differ in a dialectically
relevant sense from the way in which, if determinism is true, an agent’s putative
free acts and omissions are ensured. McKenna (2001) and Eleonore Stump
(2000) have developed this way of resisting Fischer and Ravizza’s criticism of
the Direct Argument. Here is how McKenna puts the objection:
[If] determinism is true, then the manner in which the facts of the past and
the laws of nature entail one unique future is not analogous to the manner in