Consciousness

• seCtIon tWo: tHe BRAIn
  specific features of the objects, such as
  colour, movement, or orientation. The rela-
  tionship between these features is then
  encoded by temporal correlation among
  the neurons. This means that neurons
  representing one object fire in synchrony
  with each other, but out of synchrony with
  neurons representing other objects at the
  same time. This means figure and ground
  can be segregated and individual objects
  represented without confusing them. The
  model allows for many kinds of syn-
  chronised neural activity apart from that
  based on oscillations.

Engel and his colleagues reviewed many

studies showing that synchronisation

between cells occurs widely in both per-

ceptual and motor systems. They conclude

that arousal and selective attention involve

enhanced synchrony in the relevant pop-

ulations of neurons, and that ‘temporal

binding may indeed be a prerequisite for the access of information to phenom-

enal consciousness’ (Engel et al., 1999, p.  133). In their view, synchronisation is

necessary for consciousness but not sufficient because information must also

enter short-term memory. This brings the theory closer to GWT.

The theoretical and empirical basis of binding by synchrony remains controver-

sial. Some experiments have failed to find any relationship between synchrony

and binding of moving patterns or static visual objects (e.g. Thiele and Stoner,

2003; Dong et al., 2008). Some researchers have also argued that the theory is

incoherent in being focused on early stages of cortical processing, even though

both neurological evidence and the perceptual facts of binding suggest that

it must be a high-level computation; and that the architecture of the cerebral

cortex lacks the mechanisms needed to decode synchronous spikes and to treat

them as a special code (Shadlen and Movshon, 1999). Others have even sug-

gested that none of this is needed because the ‘binding problem’ does not exist

(Merker, 2013).

These attempts to solve the binding problem seem related to the unity of con-

sciousness, but although they explain how percepts are unified, we may still

wonder how this relates to subjectivity. Some conclusions, like that of Engel and

colleagues above, are ambiguous about correlation and causation, for exam-

ple when they conclude that ‘at least at early stages of sensory processing, the

degree of synchronicity predicts reliably whether neural activity will contribute to

conscious experience or not’ (Engel et al., 1999, p. 146, our italics). Some of their

phrases also imply Cartesian materialism  – including talk of information having

access to phenomenal consciousness (e.g. 1999, pp. 133, 141, 144). Although they

are trying to explain the unity of phenomenal awareness (subjectivity, or ‘what it’s

like’) in neural terms and without magic, they seem to imply that information is

Assembly 1

Assembly 2

FIGURE 6.3 • Engel and colleagues’ model of
temporal binding. The model
assumes establishment of coherent
representational states by temporal
binding. The model assumes that
objects are represented in the
visual cortex by assemblies of
synchronously firing neurons. In
this example, the lady and her cat
would each be represented by one
such assembly (indicated by open
and filled symbols, respectively).
These assemblies comprise
neurons which detect specific
features of visual objects (such
as, for instance, the orientation
of contour segments) within their
receptive fields (lower left). The
relationship between the features
can then be encoded by the
temporal correlation among these
neurons (lower right). The model
assumes that neurons which are
part of the same assembly fire in
synchrony, whereas no consistent
temporal relation is found between
cells belonging to different object
representations (from Engel et al.,
1999, p. 131).