154 Three Source Incompatibilist Arguments
contend that being a mediated rather than an ultimate cause of action is sufficient
for freedom and responsibility. It would just be that the nature of the mediated
cause is rich enough to capture the many features of agency marking off persons
as sophisticated creatures by contrast with other beings whose resources for
causal efficacy in the natural order are by comparison more impoverished.
Second, a compatibilist might instead reject the second premise of the argument
and attempt to offer a compatibilist- friendly account of ultimacy. Such a com-
patibilist could grant, as premise 1 contends, that an important condition upon
acting freely is that one act from resources of her own agency that she had some
hand in shaping for herself, but that this is consistent with a deterministic history
of a sort that falls far short of a condition like U2.^11
In our estimation, what the preceding assessment of the incompatibilist
version of the Ultimacy Argument shows is that for incompatibilists to advance
a compelling version of it, what they need is a substantive, non- question-
begging, defense of a proposal for what ultimacy is such that it would help
render both premises 1 and 2 of the argument true. This, it seems, would require
a different style of argument that would yield a conclusion for a thesis like U2.
If so, then such a source incompatibilist defense would really rest on this distinct
argument. As readers will soon learn, in our estimation the best candidate argu-
ment to do this work is a Manipulation Argument for Incompatibilism—which
we take up in Section 4.
Before moving on to the next argument for source incompatibilism, we pause
to clarify something that should be of use to many studying this literature. It is
important to distinguish between philosophers who are committed to an ultimacy
thesis, such as U2, and philosophers who are committed to some version of the
Ultimacy Argument. Robert Kane (1996: 73–8), for instance, is committed to an
ultimacy thesis, and so is Derk Pereboom (2001: 4; 2014: 4). But neither is com-
mitted to arguing for a thesis by way of an Ultimacy Argument. As it happens,
both instead appeal to a manipulation argument.
7.3. The Direct Argument for Incompatibilism
We turn to the Direct Argument, first advanced by Peter van Inwagen (1983).
The Direct Argument for incompatibilism is so named, as van Inwagen
explained, because it aims to establish an incompatibility between moral
responsibility and determinism without taking the indirect route of trying to
show that something necessary for moral responsibility—such as the freedom to
do otherwise—is incompatible with determinism (van Inwagen, 1983: 183–4).^12
In this way, it aims to secure moral responsibility’s incompatibility with deter-
minism directly. In deference to van Inwagen’s intentions, we will follow his
lead and forgo any attention to the freedom condition for moral responsibility in
setting out the argument.^13
The Direct Argument’s structure mirrors the structure of the Consequence
Argument, but instead of transferring powerlessness over the past and the laws
to powerlessness over their consequences, it transfers nonresponsibility for the