Chapter
Nine
Agency and free will
that in the end he abandoned determinism [pp.
168–169]). The results are also compatible with
dualist interactionism (Popper and Eccles, 1977)
or with ‘the possibility that physical events are
susceptible to an external “mental force” at the
micro level’ (Libet, 2004, p. 154). Libet therefore
proposed
that conscious control can be exerted
before the final motor outflow to select or
control volitional outcome. The volitional
process, initiated unconsciously, can either
be consciously permitted to proceed to
consummation in the motor act or be
consciously ‘vetoed’.
(Libet, 1985, p. 536–537)The idea, then, is that unconscious brain events start the process of a voluntary
act but then, just before it is actually carried out, consciousness may say either
‘yes’ or ‘no’: the action either goes ahead or not. This would happen in the last
150 to 200 ms before the action. Libet provides two kinds of evidence for this
conscious veto. First, participants sometimes reported that they had an urge to
act but then aborted or suppressed the action before it happened. Unfortunately,
the neural correlates of aborted self-timed actions cannot be measured because
averaging over many trials needs a movement signal to act as the time cue. So in
additional experiments, participants were asked to move at pre-arranged times,
and then abort some of the actions, allowing the averaging to be done. These
showed ramplike pre-event potentials that then flattened or reversed about
150–250 ms before the preset time. This suggested to Libet that the conscious
veto interfered with the final development of the RP.
In this way, Libet was able to retain a causal role for consciousness in voluntary
action. He concluded that his results are not antagonistic to free will but rather
illuminate how free will operates. When it comes to morality and matters of con-
science, we can still be expected to behave well. Although we cannot consciously
control having an impulse to carry out an unacceptable action (whether rape or
murder or stealing sweets in the supermarket), we can be held responsible for
consciously allowing its consummation – or not. As Richard Gregory characteris-
tically punned it, ‘We don’t have free will, but we do have free won’t’ (1990). The
idea has since acquired more support from a study finding a similar ‘point of no
return’ about 200 ms before movement onset: before this point, the movement
can still be vetoed (Schultze-Kraft et al., 2015).
As with Libet’s earlier experiments, the debate following publication of his results
raised both philosophical and methodological problems (undated references
in this section all refer to commentaries following Libet [1985]). Eccles used the
data to support his dualist-interactionist theory, and David Rosenthal (2008) has
argued that the findings are just what a HOT theory of consciousness would
predict: a mental state is conscious only if it is the object of a higher-order men-
tal state, which you would expect to come after the decision itself. So he says
(Pre-plans) (No pre-plans) (Conscious wish)Self-initiated act: sequenceRP I RP II WSEMG–1000 –500 –20 00 msec350 ms
FIGURE 9.5 • According to Libet the sequence of
events in a self-initiated voluntary
act is as shown. Preplanning (RPI)
occurs as much as a second before
the movement. For spontaneous
actions without preplanning,
activity (RPII) begins about half
a second before the movement.
Subjective awareness of the will
to move appears about 200 msec
before the movement. Subjective
timings of a randomly delivered
skin stimulus (S) averaged about
–50 msec from actual time (Libet,
1999, p. 51).