experience. This predictable gap is not a problem
for perception; it is an integral part of how it has
evolved to work.
So how are gaps dealt with in perception more
generally? One idea is to investigate where the
mechanisms responsible for filling-in start to
converge with those responsible for ordinary
perception, to help identify ‘the critical stage
at which subjective visual experience emerges’
(Komatsu, 2006). The argument is that because
the two have different starting points – one
beginning with photoreceptors responding to
external stimuli, the other not – there must be
a specific point at which the two converge and
perceptual consciousness emerges. But this
again commits us to some version of the idea
that we can only experience what is represented somewhere ‘in consciousness’.
CHANGE BLINDNESS
Look at the picture in Figure 3.8 for a few moments. As you explore it you probably
make many saccades and blinks, but you hardly notice these interruptions. It feels
as though you look over the picture, take it in, and now have a good idea of what
is there. Now ask yourself this question. If the tray under the teapot disappeared
while you blinked, would you notice? Most people are sure they would.
Research showing they are wrong began with the advent of eye trackers, which
made it possible to detect a person’s eye movements and make a change to a
display during a saccade. In experiments beginning in the 1980s (Grimes, 1996),
participants were asked to read text on a computer screen and then, during their
saccades, parts of the surrounding text were changed. An observer would see
the text rapidly changing, but the participants themselves noticed nothing amiss.
Later experiments used complex pictures, with an obvious feature being changed
during saccades. The changes were so large and obvious that under normal cir-
cumstances they could hardly be missed, but when made during saccades they
went unnoticed.
This may seem very strange, but the effect is easily explained by the links between
eyes and brain. Under normal circumstances, motion detectors quickly pick up
transients and direct attention to that location. In set-ups like these, however, this
mechanism is disabled. A saccade causes a massive blur of activity that swamps
out these mechanisms, leaving only memory to detect changes. The implication
is that trans-saccadic memory is extremely poor. With every saccade, most of
what we see must be thrown away.
This research complements earlier work on trans-saccadic memory and visual
integration across blinks and saccades (for a review, see Irwin, 1991). For a long
time, it was taken for granted that the visual system must somehow integrate its
successive pictures into one big representation that would remain stable across
body movements, head movements, eye movements, and blinks. This would, of
‘Is the richness of our
visual world an illusion?’(Blackmore et al., 1995,
p. 1075)FIGURE 3.7 • A field of yellow doughnuts. Shut
your right eye and look at the
small white dot near the middle of
the illustration with your left eye.
When the page is about six to nine
inches from your face, one of the
doughnuts will fall exactly around
your left eye’s blind spot. Since
the black hole in the centre of the
doughnut is slightly smaller than
your blind spot, it should disappear
and the blind spot then is ‘filled in’
with yellow qualia from the ring so
that you see a yellow disc rather
than a ring. Notice that the disc
‘pops out’ conspicuously against
the background of rings.
Paradoxically, you have made a
target more conspicuous by virtue
of your blind spot (Ramachandran
and Blakeslee, 1998, p. 236).