246 Contemporary Incompatibilism: Libertarianism
York occurs. So at least we can say that by causing D, the agent brings about G
as a logical consequence of causing D.
But even on Chisholm’s bolder view, it wouldn’t be the agent’s causing G
that fundamentally explains why G is an event for which she is responsible, but
rather the fact that G embeds the responsibility- conferring relation, her agent-
causing D at t. The crucial lesson is independent of the additional causal claim.
On the agent- causal libertarian’s view, what’s metaphysically fundamental is
that the agent substance- causes the decision, and even if the more complex event
of type G that embeds this agent- causal relation turns out to be uncaused, it fea-
tures, as a component, an instance of causation that supplies the control required
for moral responsibility.
10.8. Agent Causation and Rationality
Another issue for agent- causal libertarianism concerns whether it can incorp-
orate the influence of reasons on action and decision. On Donald Davidson’s
widely accepted account, reasons influence an action by causing it. Given that
reasons are analyzed as agent- involving events, and that the influence of reasons
on action is causal, actions would be event- and not substance- caused, at least in
this respect. Actions for which agents are morally responsible are typically
rational actions, and so such responsible actions would in one respect need to be
event- rather than substance- caused. This objection poses a challenge to an
agent- causal account of morally responsible action.
Clarke (2003) and O’Connor (2000) have each proposed accounts that
address this objection. On Clarke’s integrated account of agent causation, an
action for which the agent has responsibility- conferring control will be caused
by two distinct factors, a complex of beliefs and desires, which includes reasons
for action, and the agent- as-substance. This two- stream account allows Clarke to
endorse Davidson’s causal view of reasons for action and to accept agent causa-
tion. A crucial question that arises for this account is whether the agent- as-
substance as cause of a decision and the causing of this decision by the
belief–desire complex are sufficiently integrated for the substance- causing to be
rational. An intuitive way to think of agents as rational is as influenced by
reasons in making decisions. So one might ask: On Clarke’s account, is the
agent- as-substance influenced by the belief–desire complex that constitutes her
reasons, or is she not? If she isn’t, then the account appears vulnerable to an
objection Galen Strawson (1986) raises, that the agent- as-substance is non-
rational in her causing of the decision. If the agent- as-substance is influenced by
the reasons, then either this influence is causal or non- causal. The non- causal
option conflicts with Clarke’s Davidsonian position on the role of reasons. But
he seems to want to deny the causal option insofar as it involves reasons causally
influencing the agent- as-substance. He says, for example, that “a substance
cannot be an effect” (Clarke, 2003: 158). Yet there would seem to be a signi-
ficant motivation to hold that if agents are substance- causes they can be causally
influenced by reasons.