160 Three Source Incompatibilist Arguments
No one is responsible for the fact Plato died in antiquity. Nor is anyone
responsible for the fact that if Plato died in antiquity, he never met Hume.
Therefore, no one is responsible for the fact that Plato never met Hume.
Call this case Plato. There are many other cases like it one might construct. It is
such cases that—at least apparently—make appealing a principle like TNR. Of
course, Fischer’s objection is not that TNR is not broadly appealing but rather
than TNR- R is not, and that the latter’s appeal really just begs the question
against the compatibilist (that is, its appeal is exhausted in an intuition support-
ing incompatibilism). But suppose it were pointed out to unbiased, impartial
inquirers who might come to see the initial appeal of TNR by way of examples
like Snakebite and Plato that when there are two paths to a certain result, the
principle does not seem to work. Suppose, for instance, that it were pointed out
that in a revised case like Two- Path Snakebite TNR fails. In Two- Path Snakebite
just prior to John’s being bitten, Leonard secretly poisoned John so that he
would die soon thereafter. Such an audience, seeing that two- path cases create
problems, might readily think that a natural move is to restrict the principle,
since so restricted, it captures their pattern of inference rather innocently when
reasoning about cases like Snakebite and Plato. If this is correct, then it at least
seems that a principle like TNR- R is not after all question- begging. If so, the
Direct Argument survives as a serious threat to compatibilism.
Thus far, we have explored the strategy of producing counterexamples to
TNR as a means of resisting the Direct Argument, and this has led to proposed
revisions such as TNR- R. A slightly different strategy, developed by McKenna
(2008c), seeks to undermine TNR or revisions like TNR- R not by adducing
counterexamples (of the form NRp & NR(p → q) & ~NRq), but by calling into
question the initial appeal of any such principle as one that can be employed in
an argument for an incompatibilist conclusion. Consider Snakebite and Plato,
which were offered as innocent examples confirming the appeal of TNR. Notice
that in each example the transfer of nonresponsibility from one fact to con-
sequences of it never in any way involves a causal chain passing through an
agent when she is acting competently and unimpaired in her exercise of well-
functioning agential resources (by standards that would be agreeable to both
compatibilists and incompatibilists alike). In Snakebite, for instance, a poisonous
snake bites John and he dies from it. There is no fact about which no one is
responsible that leads to any agent acting in any normal way at all.
With the preceding observation in place, consider the history of a putatively
free agent at a determined world as a compatibilist might see it. However the
facts of the remote past and the laws unfold, they will include a history that
results in this agent—let’s call her Ann—coming into existence. As her life
progresses, she will acquire various skills and abilities, eventually coming to be
able to deliberate and form intentions. At some point, Ann will exercise
her freedom in a normal way. From conditions that lack freedom, a free agent
will emerge, one whose actions can be candidates for assessment in terms of
moral responsibility. So consider Ann’s first free act (FFA)—free at least as the