Chapter
Three
The grand illusion
While you did that all the other portraits would just be face-shaped blobs. Now
you turn to the next, foveate that one, and again conclude that it is Marilyn and
that it looks just the same as the first one. Now another . . . You can make at most
four or five saccades a second. So, you cannot possibly check every single one in
the brief time it takes you to conclude – ‘all Marilyns’. You never see just one clear
portrait and a lot of blurry blobs, you see the whole detailed lot of them. How can
this be?
Could the brain be taking one of its high-resolution foveal views of the portrait
and reproducing it lots of times, as if by photocopying, over an inner mental wall?
Of course not, says Dennett. Having identified one portrait, and using texture-
detection mechanisms to identify that all the blobs are of a similar size and shape,
and finding nothing to suggest that the other blobs are not also Marilyns, the
brain jumps to the reasonable conclusion that the rest are Marilyns too, and labels
the whole region ‘more Marilyns’. This is more like paint-by-numbers than filling in
pixel-by-pixel. The reason you would notice a moustache or a silly hat is that you
have dedicated pop-out mechanisms to detect such anomalies. If none of these
are activated, the conclusion ‘all the same’ stands.
Of course, it does not seem that way to you. You are convinced that you are seeing
lots of identical Marilyns (or Dans, in our picture), and in a sense, you are. There
are lots of portraits out there in the world, and that is what you are seeing. Yet it
does not follow that there are lots of identical faces represented in your brain.
Your brain just represents that there are lots: ‘no matter how vivid your impression
is that you see all that detail, the detail is in the world, not in your head’ (Dennett,
1991, p. 355).
This does not apply only to the multiple Marilyns room, or our multiple Dans pic-
ture. When you walk along the street, you cannot possibly look at all the detail
around you, yet you see no gaps in the places where you haven’t looked. Does
the brain fill in the spaces with plausible guesses about cars and trees and shop
windows and children running to school? Does it need to?
There is a range of ideas about filling-in (Pessoa, Thompson and Noë, 1998;
Komatsu, 2006). One, known as isomorphic filling-in, is that the brain actually fills
in the details as though to complete a picture in the brain (or ‘in consciousness’).
According to Koch, ‘the brain does not simply neglect the blind spot: it paints in
properties at this location’ (2004, p. 54). Another, known as symbolic filling-in, is that
the process is more conceptual than picture-like and occurs higher up the visual
system. The most sceptical view is that there is no need to fill in anything at all.
Neuropsychologist V. S. Ramachandran reports both formal and informal exper-
iments on a range of cases (Ramachandran and Blakeslee, 1998). With normal
observers, if two vertical lines are shown, one above and one below the blind spot,
the observer sees one continuous line. The lines can be offset slightly and still seem
to form a single straight line, but if the same is done with horizontal lines they do
not line up. Missing corners are not completed, but if the blind spot is positioned
‘We depicted
consciousness as a place
peopled with small
imitations and these
imitations were the
images’(Sartre, 1940/2004, p. 5)- Do not turn the page yet. On the next page you will see an illustration. Try to look at it for just three seconds. You might
like to practise counting at the right speed first, or get a friend to time you. Then turn the page, look at the picture while you
count to three, and then turn back.
What did you see? Try to describe the picture in words before you look again.