Chapter
Nine
Agency and free will
the thalamus) is stimulated. As with the cortex, long trains of pulses are required
for neuronal adequacy, but unlike in the cortex, a primary evoked potential also
occurs as it does when the skin is touched. The backwards referral hypothesis
makes a clear prediction: that stimulation to medial lemniscus should be referred
back in time to the start of its train of impulses, even though stimulation of cortex
is not. In this final experiment, Libet again asked participants about the relative
subjective timing of different stimuli. As predicted, he found that if a skin stimulus
came at the same time as the start of a train of pulses to the medial lemniscus,
the participants felt the two simultaneously – even though the train of pulses
was only felt at all if the stimulation went on long enough to achieve neuronal
adequacy.
What should we make of these findings? In general, critics have not agreed on
serious weaknesses in the methods used or the specific results. The ideal way to
be sure is to repeat the experiments, but medical advances mean that operations
to expose the brain are now very rare. So the experiments are unlikely ever to be
replicated. We are probably best, then, to assume that the findings are valid. The
real controversy surrounds how to interpret them.
Libet’s own interpretation is his ‘time-on theory’ of consciousness. This has two
components: first, that consciousness can occur only when neural activity contin-
ues long enough for neuronal adequacy (usually about 500 ms), and second, that
activity with a shorter duration can still be involved in an unconscious process or
converted into a conscious one by increasing its duration. He suggests that atten-
tion may work by increasing the excitability of certain areas so as to lengthen the
duration of activity and so achieve the time-on for consciousness (Libet, 2004).
On his view, unconscious processes really do ‘become conscious’ when neuronal
adequacy is achieved. He says that ‘when the duration of repetitive similar acti-
vations of appropriate neurons reaches a certain value, then the phenomenon of
awareness emerges’ (2004, pp. 58–59).
This theory provides an answer to the question we focused on in the last chapter:
what is the difference between conscious and unconscious processes? According
to Libet, the difference is whether neuronal adequacy is reached or not. To com-
pare this with one contrasting example, when Milner and Goodale suggest that
processing for perception in the ventral stream leads to consciousness, while dor-
sal stream processing for action does not, Libet (1991) argues that the important
difference for consciousness is not the brain areas where the processing occurs,
nor what kind of activity it is, nor what it leads to, but only whether it continues
for long enough.
Libet also makes some much more controversial suggestions. In particular, he
claims that the evidence for backwards referral raises problems for materialism
and the theory of psychoneural identity (i.e. that consciousness and neural activity
are the same thing). He even considers ‘the possibility that physical events are sus-
ceptible to an external “mental force” at the micro level, in a way that would not be
observable or detectable’ (Libet, 2004, p. 154). Roger Penrose (1994a and 1994b)
also believes that the phenomena uncovered in these experiments challenge
ordinary explanations and demand reference to nonlocality and quantum the-
ory. Similarly, Karl Popper and John Eccles claim that ‘This antedating procedure
does not seem to be explicable by any neurophysiological process. Presumably
it is a strategy that has been learnt by the self-conscious mind’ (1977, p. 364). In
DID MY THOUGHTS
CAUSE THIS ACTION?‘when the duration [. . .]
reaches a certain value,
then the phenomenon
of awareness emerges’(Libet, 2004, pp. 58–59)