The Intelligent Eye

AutorGregory, Richard Langton, 1923-
PealkiriThe intelligent eye / R. L. Gregory
IlmunudNew York [etc.] : McGraw-Hill Book, c1970
ViideGregory, Richard L. 1970. The intelligent eye. New York: McGraw-Hill Book.

In a delightful paper, 'What the Frog's Eye Tells the Frog's Brain', by Lettvin, Maturana, McCullon and Pitts, several specific pattern-receptive mechanisms are identified. The eye responds to movement, to changes of illumination and to what we may call 'rotundity'. A small black shadow is signalled strongly and serves to evoke the fly-catching reflex.(Gregory 1970: 22)
Rotundity is "sphericity: the roundness of a 3-dimensional object". Sphericity is yet another interesting term which might come handy when discussing the semiosphere, as in I doubt whether such disparate sign-processes are ultimately grouped into sphericity. In any case, this quote is noteworthy in my mind because it almost answer a question I pondered about when I reminisced about Battlefield 2; namely, if staying immobile on a roof with a sniper could be pulled off in real life. The answer is most likely no, because even if you don't show movement, the change in illumination, or simply being "a small black shadow" could give you away.
In this book I propose to consider the inner 'logic' of perception. The main argument is that perception is a kind of problem-solving. Pictures are regarded as a remarkable invention, requiring special perceptual skills for seeing them, leading to abstract symbols and ultimately to written language. By considering the perception of objects represented in pictues and symbols (including the pictograms of early languages) I hope to show that our most abstract thinking may be a direct development of the first attempts to interpret the patterns in primitive eyes in terms of external objects. (Gregory 1970: 31)
This general formulation of his argument is also stated on Gregory's wikipedia page. Generally it seems that this kind of thinking can also be applied to perception of bodily signals: noticing significant action, in this sense, is a kind of problem-solving (a statement not far off from what is propounded in some popular books about body language).
The retinal images do not have the double reality of external pictures. We do not see them both as patterns and as representing something else. We read aspects of reality from the patterns in the eyes, but we do not also see our eyes' images. (Gregory 1970: 32)
Ah! This was exactly what I was arguing for in my critical comment towards J.G. when she claimed that perception is "coded". Perception is not coded in signs in themselves: they do not have a double reality. When I see the cheese right in front of me on the table I'm not seeing signs of cheese, I'm seeing cheese. So while Gregory says that pictures are important because they show absent things, the pictures themselves are not absent, their referential content (object) is.
Object-recognition is simplified by the fact that most familiar object are largely redundant. Faces have two eyes - and so but one needs to be seen. If there is an eye - there will surely also be a nose. If a head - somewhere near a body, legs and feet. Indeed we could make no sense of close-ups on the cinema screen if we did not make such associated facts about familiar objects. (Gregory 1970: 86-87)
Automatically brings to mind Yu. Lotman's discussion on a young lady who upon first seeing a film found is gross and disgusting because the heads seemed to her decapitated.
When sensory information is used for guiding action, it must control movements which are appropriate to the position and sizes of the surrounding objects. Sensory information must be scaled appropriately to the external world, for control and prediction to be possible. Similarly, the readings of instruments cannot be in arbitrary units but must, at least ultimately, be tied to familiar objects. Some measures are direct (such measures as lenght with rulers) while others are indirect (such measures as temperature with a thermometer). (Gregory 1970: 98)
Firstly, action control and mapping of external environment I should investigate more. Secondly, the note about instruments being exact made me think if a ruler with arbitrary/random measures on it would be a good prank. And thirdly, I'm not so sure about Gregory's use of direct and indirect; rather it should be a matter of degree. His direct is visual, mine would be corporeal. This would suppose that the body is an instrument on it's own, but this debate it too complex to get into at the moment.


Post a Comment