24
2 The Visual World
An Evening at the Cartesian Theatre
From time immemorial men and women have speculated about seeing, and
how the power of vision reveals the world to us. The great philosopher Plato
wrote in the fourth century B.C. that light emanating from the eyes touched
objects with its rays. The rays probe objects in much the same way as we
would with our fingertips. This idea was plausible for many reasons: First, the
ancients believed that seeing is an active pursuit. Also, the ancient Greeks had
already noticed the “fire” gleaming in animals’ eyes if you surprise them at
night, suggesting there is a source of light in the eyes. Moreover, many people
believe they can “feel” on their backs when someone is looking at them from
behind.
If this scenario strikes you as extravagant, imagine how our modern scientific
account of vision would strike our ancestors! In the picture given to us by brain
science, the optical image is dismantled by the brain and reconstructed in the
form of an abstract Gestalt which is sensory rather than pictorial.
The idea of a picture in the head re-enacting what appears to the eyes has
been derisively called the “Cartesian Theatre”. This expression was famously
coined by the American philosopher Daniel Dennett. His criticism is aimed
at a theory of animal vision that is still popular among some vision scientists,
according to which the brain reconstructs a detailed copy of the visual image.
The primary visual areas of the brain map out boundary lines, detect fine differences in shading and reflectance, and use this information—it is believed—to
reconstitute a faithful copy of the external image, like a photograph.
The problem with this account is that it brings the visual subject no closer
to seeing than he was before. Instead of the panorama stretched out in front of
him, he must now discern the photograph displayed in his head. Dennett says
that this requires us to imagine a homunculus, or little man inside the subject’s
head, peering at the inner photograph. It is this inner spectator who is really
looking at the picture. But then, the homunculus requires a homunculus of
his own, and so on to infinity.
This little allegory makes it clear why it is not the visual brain’s task to reconstitute a picture. Rather, the brain assembles a profoundly transformed analysis
of what a scene portrays. Instead of producing a pictorial facsimile, the brain
creates a coded “report” of all the useful information extracted from the optical
array. Finally, by a process which science cannot explain, this information is
transformed into the sensory experience of seeing.
The function of the senses is to give animals sensory experience, and evolution has ensured that each of the senses has its own unique, very distinctive
way that it feels to us. For example, we experience smells as one thing, sounds
2 The Visual World
An Evening at the Cartesian Theatre
From time immemorial men and women have speculated about seeing, and
how the power of vision reveals the world to us. The great philosopher Plato
wrote in the fourth century B.C. that light emanating from the eyes touched
objects with its rays. The rays probe objects in much the same way as we
would with our fingertips. This idea was plausible for many reasons: First, the
ancients believed that seeing is an active pursuit. Also, the ancient Greeks had
already noticed the “fire” gleaming in animals’ eyes if you surprise them at
night, suggesting there is a source of light in the eyes. Moreover, many people
believe they can “feel” on their backs when someone is looking at them from
behind.
If this scenario strikes you as extravagant, imagine how our modern scientific
account of vision would strike our ancestors! In the picture given to us by brain
science, the optical image is dismantled by the brain and reconstructed in the
form of an abstract Gestalt which is sensory rather than pictorial.
The idea of a picture in the head re-enacting what appears to the eyes has
been derisively called the “Cartesian Theatre”. This expression was famously
coined by the American philosopher Daniel Dennett. His criticism is aimed
at a theory of animal vision that is still popular among some vision scientists,
according to which the brain reconstructs a detailed copy of the visual image.
The primary visual areas of the brain map out boundary lines, detect fine differences in shading and reflectance, and use this information—it is believed—to
reconstitute a faithful copy of the external image, like a photograph.
The problem with this account is that it brings the visual subject no closer
to seeing than he was before. Instead of the panorama stretched out in front of
him, he must now discern the photograph displayed in his head. Dennett says
that this requires us to imagine a homunculus, or little man inside the subject’s
head, peering at the inner photograph. It is this inner spectator who is really
looking at the picture. But then, the homunculus requires a homunculus of
his own, and so on to infinity.
This little allegory makes it clear why it is not the visual brain’s task to reconstitute a picture. Rather, the brain assembles a profoundly transformed analysis
of what a scene portrays. Instead of producing a pictorial facsimile, the brain
creates a coded “report” of all the useful information extracted from the optical
array. Finally, by a process which science cannot explain, this information is
transformed into the sensory experience of seeing.
The function of the senses is to give animals sensory experience, and evolution has ensured that each of the senses has its own unique, very distinctive
way that it feels to us. For example, we experience smells as one thing, sounds