HOW THE BRAIN MAKES POTENTIAL 173In the 1970s, this view was encouraged by experiments in which elec-
trodes were inserted (in anesthetized animals) near cells in the ret ina or
parts of the “visual” brain. Th e cells seemed to show sensitivity to one of
a range of simple visual features, such as light spots, lines, or bars, shone
into the eye. Th ese activating features seemed to get more complex the
“higher up” in the visual system the electrodes took recordings from. For
example, some cells seemed more sensitive to angles, movement in a
specifi c direction, or even whole objects. Th e tacit assumption is that
each level in this cascade comprises some pro cessing rules built into the
circuitry for building the simple features into complex ones. An internal
model of the external world is thus created from converging features.
It sounds simple, but there are many reasons such feature detection
alone cannot create the functional image required. One of these is that
the world is populated with objects that are three dimensional. Th ese
become collapsed as two- dimensional images on the ret ina. And the
same two- dimensional image can be created by a variety of diff er ent ob-
jects/features. So the model implies a great loss of visual discrimination
(the opposite of what actually happens).
Another prob lem is that the two- dimensional image would be grossly
distorted on the back of the eye, because the eye is hemispheric (imagine
watching photos projected onto a bowl- shaped screen) and the image
is upside down. Moreover, the receptor surface is not a continuous sheet
(like camera fi lm) but an aggregate in the form of millions of light-
sensitive cells. So the visual fi eld is really detected in the form of a cloud
of light spots.
Yet another prob lem is that the image of any single object is never still.
We will be moving toward and/or around it; it moves to diff er ent places
at diff er ent distances (with changes in apparent size); it may rotate, move
behind other objects, become partially hidden, and so on. Th e movements
of the image also have to be distinguished from the apparent motion cre-
ated by ordinary eye movements. Th is is further exacerbated by the fact
that the eye is constantly moving, from side to side in its socket up to ten
times a second. Th ere is simply no stable image to rec ord.
In other words, the image of every object is under constant spatiotem-
poral transformation; the cloud of light points is in constant motion. And
This content downloaded from 139.184.14.159 on Tue, 17 Oct 2017 13:54:00 UTC