248 Contemporary Incompatibilism: Libertarianism
must be of a different sort from that of the causally relevant events, and on the
occasion of a free decision, the exercise of the agent- causal power must be dis-
tinct from the exercise of the causal powers of these events. The disappearing
agent objection shows that causal powers of the events are not the sort that can
provide the decision- settling control needed for moral responsibility. In fact,
O’Connor says:
We insist upon the importance of the distinction between (the persisting
state or event of one’s having) reasons structuring one’s agent- causal power
in the sense of conferring objective tendencies towards particular actions,
and reasons activating that power by producing one’s causing a specific
intention. On the view I have described, nothing other than the agent himself
activates the causal power in this way. To say that I have an objective prob-
ability of 0.8 to cause the intention to join my students at the local pub
ensures nothing about what I will in fact do. I can resist this rather strong
inclination just as well as act upon it. (O’Connor, 2009: 213)But if the propensities of the reasons and those of the agent- causal power toward
the relevant range of effects are exactly the same, then either the causal powers
of the reasons and of the agent- as-substance are identical after all, or else this
identity of propensities is an unexplained coincidence.
One might also object that if the propensities of the agent- causal power
were structured by objective probabilities, the agent would lack the control in
action required for moral responsibility in the basic desert sense (Pereboom,
2014). Suppose God created us as agent- causes whose distinctive causal power
featured two propensities, one for self- interest and the other for morality. Each
of these propensities is structured by equal and unalterable objective probabil-
ities, so that, over our lifetimes, we can expect half of our decisions when self-
interest conflicts with morality to be self- interested and half moral. Even if the
agent on any such particular occasion could choose either the self- interested
or the moral option, from the incompatibilist point of view the agent would
not be blame worthy in the basic desert sense for the overall pattern, in par-
ticular for the lifetime’s immoral half. And it would seem that if an agent isn’t
blameworthy for the pattern, she isn’t blameworthy for the instances that make
up the pattern.
It might be replied that this verdict results from the probabilities being
unalterable.^2 But suppose we change the case so that agents are in addition gov-
erned by a probabilistic law that specifies a 50 percent probability over a lifetime
that a subject agent- cause her moral conversion, whereupon 100 percent of the
subsequent actions in situations of conflict between morality and self- interest
will be moral. Even then, from the incompatibilist point of view, the agents in
this world would not be blameworthy for the pattern of immoral actions they
may perform, and thus, it would seem, not for the instances that make up the
pattern. More complex cases can be constructed, but the verdict would appear to
be the same.