Free Will A Contemporary Introduction

Seven Views of Contemporary Compatibilism 191

as she was not caused to endorse and then sustain unsheddable values or princi-
ples by means that bypassed her capacities for critical evaluation (1995: 193).
Mele argues for his historical thesis by invoking pairs of examples in which
one agent does not have an objectionable history and the other does, where it
seems that only this factor and no other explains our intuition that one agent is
free and responsible (or instead autonomous) and the other is not. Mele’s most
frequently discussed case is that of Ann and Beth (1995: 145; 2006: 164–5).
Here is our compressed retelling of the cases:

Ann is an extremely productive college professor who is by stipulation auto-

nomous. She devotes her life almost exclusively to her work. Beth is an

equally talented colleague who has many interests, not just her work. Beth’s

dean wants Beth to be like Ann. So he hires a team of psychologists capable

of “new­ wave brainwashing” (1995: 145) to capture Beth in her sleep and

without her knowledge eradicate her hierarchy of values and install instead

a hierarchy that is just like Ann’s. Beth turns out to be in the relevant

respects a “psychological twin” of Ann (145). Afterwards, Beth begins to

act just as Ann does. The question is, when Beth acts, does she do so auton-

omously? Since Beth is just like Ann, and since, by hypothesis, Ann acts

autonomously, if we judge that Beth does not act autonomously, it must be

due to her unusual history.

Mele’s proposed diagnosis of what explains the difference is just as set out in the
preceding paragraph: Beth fails to satisfy a negative historical condition. When
she acts in ways that please the dean, her history includes (rather than lacks) a
process of value- acquisition that bypasses her ability to acquire and then sustain
her values under her own steam.
In Free Will and Luck (2006b: chapters 6– 7 ) Mele revisited his 1995 compat-
ibilist proposal, defending it against various incompatibilist arguments. In doing
so, he shifts to writing in terms of freedom and moral responsibility rather than
autonomy. He makes short work dispensing with the Consequence Argument
and the prospects for leeway freedom (2006b: 138). As discussed earlier (Section
5.3.1), Mele defends Frankfurt’s argument, and so, as it bears on his compatibil-
ist proposal, he advises a source compatibilist position (2006b: 138).
What about arguments for source incompatibilism? Mele replies (138–44) to
Pereboom’s Four­ Case Manipulation Argument.^5 One of his objections involves
a softline reply in response to Pereboom’s Case 1, arguing that Plum in Case 1 is
“out of the control loop” (142), and so his agency is simply not engaged. Recall
that we considered such a reply (Section 7.4) and argued that Pereboom had the
resources to reply in a way that addresses Mele’s objection but still retains the
force of the manipulation case. But set this dispute aside. Note that Mele’s
historical proposal is illustrative of how historical compatibilists can appeal to
principled resources for adopting a softline reply to a range of manipulation
arguments (a point we note above in Section 8.1). If, for instance, the case of
Beth were to be imported into a manipulation argument, supposing Beth’s