88 The Debate over the Consequence Argument
The problem with proceeding this way is that the reasons for thinking that the
past is fixed for a person might vary from the reasons for thinking that the laws
are fixed for a person. Or at any rate, it should not be assumed that the reasons
are the same and that they can thus be lumped into a single claim about power
necessity.^19 True, one might try to argue from NPo and NL to N(Po&L) by way
of further steps. But van Inwagen’s elegant formulation allows one to avoid
these complications. (Maybe. Or maybe it illicitly masks them. This is some-
thing we’ll consider in the next section.)
4.5. Questioning Rule β and Seeking an Improved Version
Having set out van Inwagen’s version of the Consequence Argument, we’ll now
focus on the role of Rule β in the argument. Notice that Rule β does not appear
as a step in the argument (steps 1 through 7). Rather, it is an inference rule that
is intended to justify or make rationally permissible transitions between steps—
in this case, step 5 from steps 3 and 4, and step 7 from steps 5 and 6.
Reflect back on the formulation van Inwagen gave of the Consequence Argu-
ment in terms of it not being up to us what the laws of nature and the past are,
and how from these considerations, if determinism is true, it is not up to us what
we do now. We began the chapter by quoting this formulation. The insight that
binds these judgments—that makes compelling the transition from powerless-
ness with respect to the past and laws to powerlessness with respect to our
actions—is codified in Rule β. In this respect, Rule β is meant to offer a source
of support for an incompatibilist conclusion. It is supposed to be an intuitively
plausible principle that aids in adjudicating the dispute between compatibilists
and incompatibilists. More specifically, it is meant to show, without being ante-
cedently biased toward one philosophical position or another, that there are
reasons to move one in the direction of an incompatibilist conclusion. The
pattern of inference, then, must in some way have an independent standing as a
valid way of reasoning. (Above we offered an illustration of this with reasoning
about Calamity Sam’s hand of poker in his dealings with Diamond Jim.)
Now, if it turned out that Rule β, and relevant variants, such as Transfer, were
after all not valid patterns of reasoning quite generally—that is, if some instances
of the application of the rule yielded invalid results—this would raise the worry
that an application of Rule β itself when used in the Consequence Argument is
not innocent of theoretical bias in the debate between compatibilists and incom-
patibilists. The trouble is, as various philosophers have pointed out, Rule β
appears not to be valid, since there are instances of it that yield intuitively incor-
rect results.
A first concern is this. With the aid of a few innocent assumptions, a logical
truth, and a few simple logical steps, one can show that Rule β is committed to
agglomerativity with respect to “no one has a choice about,” or what others call
“power necessity” (McKay and Johnson, 1996). That is, Rule β makes permiss-
ible inferring from Np and Nq, that N(p&q). The problem is that there are clear