Consciousness

Chapter

Six

The unity

Crick concluded that ‘Consciousness [.  .  .] exists
only if certain cortical areas have reverberatory cir-
cuits (involving cortical layers 4 and 6) that project
strongly enough to produce significant reverber-
ations’ (Crick, 1994, p. 252). He argued that it may
be more efficient for the brain to have ‘one single
explicit representation’ rather than sending tacit
information to many different parts of the brain.
In other words, he distinguished explicit and tacit
(conscious and unconscious) information, and
thought that the unity of consciousness is real and
not illusory.

In their later work, Crick and Koch (2003) gave up
the idea that 40-hertz oscillations are a sufficient
condition for the NCC, arguing instead that the fea-
tures of a single object or event are bound together
when they form part of one temporary coalition of
neurons, and that the primary role of synchrony is
to help one coalition in the competition for con-
sciousness. EEG studies have confirmed a role for
synchrony in visual binding (Tallon-Baudry and
Bertrand, 1999; Tallon-Baudry, 2003). Catherine Tallon-Baudry and her colleagues
in Lyon showed that gamma oscillations were much stronger during feature-bind-
ing tasks, and for tasks in which participants had to hold the representation of an
object in short-term memory while searching a display, than during control tasks.
They also studied two patients who had electrodes implanted in extrastriate
visual areas during their treatment for epilepsy. When these patients were asked
to keep a visual image in mind while performing a matching task, visual areas
separated by several centimetres became synchronised, with oscillations in the
beta range (15–25 Hz). They took these results to confirm Hebb’s (1949) fifty-year-
old suggestion that short-term memory is sustained by reverberating activity in
neuronal loops.

Another experiment used a modified version of the famous image of a Dal-
matian dog hidden in a black and white meaningless pattern. Seeing the dog
correlated with an increased EEG response in the gamma band. From these
and other studies, Tallon-Baudry concluded that any stimulus elicits locally
synchronised activity in early visual areas, sufficient for coarse and unconscious
identification. These local oscillations could then be more strongly synchronised
between areas ‘to provide a much more detailed representation of the stimulus,
along maybe with the conscious experience of it’ (2003, p. 361). As so often with
neuroscientific studies, however, consciousness is brought into the discussion
only as an added extra at the very end.

Binding by synchrony does not necessarily involve gamma oscillations. In their
model of temporal binding, Andreas Engel, Wolf Singer, and their colleagues in
Frankfurt, Germany, propose that objects are represented in the visual cortex by
assemblies of synchronously firing neurons (Engel et al., 1999; Engel, 2003; Singer,
2000, 2007). For example, in Figure 6.3 the lady and her cat are each represented
by one such assembly of cells. Each assembly consists of neurons which detect

25 milliseconds

Time

The local

‘field potential’

The firing of

one neuron

FIGURE 6.2 • A simple figure to illustrate how

neurons fire in a 40-hertz rhythm.

(A 40-hertz oscillation repeats

every 25 msec.) The smooth

curve represents the local field

potential. This is a measure of the

average ‘activity’ of many of the

neurons in that neighbourhood.

The short vertical lines show the

firing of just one neuron. Notice

how, when this neuron fires, it

fires ‘on the beat’ of some of its

neighbours, represented by the

local field potential. The usual

sign convention for plotting the

field potentials has been reversed

(Crick, 1994, p. 245).